Well, here are some pics of my build of Bruce's concept TPU. Gotta love the "Export to Reality" button in my CAD software! :D ;)
I hope to fire this baby up next week; the only major item left outstanding is the boost converter for the pulses and the HV bias.
For now, enjoy. :)
P.S. If anyone here wants to have a copy of the FPGA firmware that I wrote to generate the pulses, just let me know. you can get a copy here: http://www.overunity.com/index.php/topic,2894.0.html I made it so that I can tweak pretty much anything I want to with respect to phase, pulse width, timing, etc.
Quote from: eldarion on July 27, 2007, 12:58:39 AM
Well, here are some pics of my build of Bruce's concept TPU. Gotta love the "Export to Reality" button in my CAD software! :D ;)
I hope to fire this baby up next week; the only major item left outstanding is the boost converter for the pulses and the HV bias.
For now, enjoy. :)
P.S. If anyone here wants to have a copy of the FPGA firmware that I wrote to generate the pulses, just let me know. I made it so that I can tweak pretty much anything I want to with respect to phase, pulse width, timing, etc.
Hi Eldarion,
The coil looks good. Can't wait to see what happens. I also can't wait to see that controller of yours at work on the Bob Boyce TPU in the future.
I can't believe I never even asked you what you were using for your controller. So I take it you can write programs for microprocessors. You are indeed the man we have been looking for. That will make designing a proper controller so much easier, from what a couple of the guys have mentioned before to me.
Happy Days! :)
Bruce
Hey man,
The TPU looks sweeet :-). How big is your controller board? Can is it small enough to fit into your ring? if not you may run into major problems syncing up the controller to the coil once that high speed field forms.
Just a few days ago me and a colleague got the bright idea to see how much of a relativistic effect we could 'predict' from the rotational speed of the TPU. In my case, the fundamental resonant frequency of my coils was about 480 kHz. Now, this is a very bad approximation because we basically converted the angular speed to a linear speed and used a formula from special relativity to calculate the time shift that would result.
I can't remember what the numbers were exactly but it came out to be somewhere in the neighborhood of 276 nanoseconds difference for every one second of time. At first glance, this sounds like no big deal but from an electronics perspective, that would throw off the frequencies by more than a quarter wavelength in my case. I imagine that the effect would probably be a little worse in reality since the field is 'technically' in an accelerated reference frame, so we would have to use general relativity to predict the effects (I'm not even going to go there, lol).
But any who, my point is that if you create a field that can spin at over 30 million RPMs (in my case), there is bound to be some kind of relativistic interactions. So ultimately, we all need to keep our junk in the ring, lol. I'm guessing that wouldn't be to hard with the micro controller if you can make a small board for it to mount on inside the ring.
God Bless,
Jason O
love the rings
great work
looking foward to see the controller
wer
It would seem that I forgot to post a pic of the controller. Here you go! :)
All the wires dangling off of it go to my power supply and scope; they are there only for testing the pulses before I hook it up to the coils. The board in the back is a delta-sigma ADC; the MOSFET board to the right contains MAX627 MOSFET driver chips and IRF640s.
I'll clean up the source a bit and then post it if you like, but I will warn you: it's all written in Verilog and VB .NET. This is no ordinary microcontroller! ;) It's actually an FPGA, meaning that I create logic using a hardware definition language. I have been programming microcontrollers for the past 5 years or so, but once I got a taste of the power of an FPGA I didn't want to go back. :D
My only concern is that the board is a bit too fat to fit in the center of the coil; I wonder if I can mount it above the coil in the center? Or up on end in the center?
Eldarion
Hi Eldarion,
Nice board you have there! I am planning on learning to program both microcontrollers and FPGAs. Which do you think is best to start learning first?
God Bless,
Jason O
Quote from: Jdo300 on July 27, 2007, 01:51:59 PM
Hi Eldarion,
Nice board you have there! I am planning on learning to program both microcontrollers and FPGAs. Which do you think is best to start learning first?
God Bless,
Jason O
Hmmm...that's a tough call. In many respects, for this type of work (TPU controller), knowing how to program an FPGA would be more useful, in my opinion. You could try reading through these excellent tutorials on Verilog: http://www.asic-world.com/verilog/veritut.html
The reason that an FPGA is better here is that you can design the logic circuits to work all in parallel, versus a standard microcontroller, which can only run tasks sequentially. Sequential execution is very bad when you must generate signals with a resolution in the tens of nanoseconds!
Hope that helps,
Eldarion
@Eldarion,
OK -- I stand corrected. FPGA does look like an 'of the shelf' solution for a suitable controller. The Spartan 3 range looks perfect, although as you say a bit big. I presume that they have embedded versions once you have a design worked out?
I dont want to get into learning yet another programming system -- but if it's fairly easy for you then I'll look at going this way.
@Jason
Wo -- you are so right about the relativistic effects and timing problems... more than you can know. Your calculations should be at the *very* low end of the scale when things get working well. I hadn't expected anyone to notice the relativistic differential problem just yet - excellent work!
Mark S.
Well, I just finished design and construction of the $20 bias supply! It provides 150-200V from a 13.8V source, depending on load. Plus, it only draws about 5 watts!
I have rescaled the meter in the picture to read in volts x10. As you probably have guessed by now, I am a bit crazy about metering and instrumentation--if something goes wrong, I want to know exactly where with a glance, not have to spend the next 30 minutes probing around with a multimeter... ;D
Eldarion
Quote from: eldarion on July 28, 2007, 09:44:39 PM
Well, I just finished design and construction of the $20 bias supply! It provides 150-200V from a 13.8V source, depending on load. Plus, it only draws about 5 watts!
I have rescaled the meter in the picture to read in volts x10. As you probably have guessed by now, I am a bit crazy about metering and instrumentation--if something goes wrong, I want to know exactly where with a glance, not have to spend the next 30 minutes probing around with a multimeter... ;D
Eldarion
Eldarion,
Nice work man, would you mind drawing up a little schematic for us of your bias supply?
Thanks
Quote from: tao on July 29, 2007, 04:53:43 PM
Quote from: eldarion on July 28, 2007, 09:44:39 PM
Well, I just finished design and construction of the $20 bias supply! It provides 150-200V from a 13.8V source, depending on load. Plus, it only draws about 5 watts!
I have rescaled the meter in the picture to read in volts x10. As you probably have guessed by now, I am a bit crazy about metering and instrumentation--if something goes wrong, I want to know exactly where with a glance, not have to spend the next 30 minutes probing around with a multimeter... ;D
Eldarion
Eldarion,
Nice work man, would you mind drawing up a little schematic for us of your bias supply?
Thanks
Sure! Here you are...
Hi Eldarion,
I have 3 x ali cases for the dds 20 boards I have built.
Hammond 1455 series
http://www.hammondmfg.com/pdf/1455L1601.pdf
Like you, I have been wondering how to get the electronics into the centre of the TPU.
I hope to put two cases back-to-back and the mosfet driver and mosfets under that.
I will need to extend the controls using some clear tubing and chop sticks or something to press the buttons.
Also using a really big 15" TPU may help things.
Regards
Rob
Quote from: MeggerMan on July 29, 2007, 06:10:22 PMAlso using a really big 15" TPU may help things.
Yeah, I think with the amount of circuitry I have here that may be the only thing that can help! ;D
Here is a pic of the finished TPU, all hooked up but for the collector to its load. I still have to figure out a way to limit the current drawn by the magnetic bias coils, as when I hook them up to the battery on the right (all in series), the jumper leads get extremely hot in a few seconds. Given that there is only 0.29 ohms of resistance in those coils, I am probably drawing nearly 41.3 amps (410 watts)!!! That is a surefire way to not get overunity performance! :D
Any thoughts?
P.S. I won't be firing this up tonight; I'll save that for tomorrow when I have a clearer head. Given that there is almost 150V floating around in this thing, I am liable to get zapped. :P
Eldarion
@Eldarion
You should only need a few hundred milliamps at most on that magnetic bias winding. You can use a simple PWM to drop the current . You really want the ability to dial the strength of that magnetic bias anyways.
Bob
Quote from: Bob Boyce on July 29, 2007, 11:05:01 PM
@Eldarion
You should only need a few hundred milliamps at most on that magnetic bias winding. You can use a simple PWM to drop the current . You really want the ability to dial the strength of that magnetic bias anyways.
Bob
Bob,
Thanks for the PWM idea; I had not thought of that. I will stick another MOSFET on my driver board for variable bias control, since I have to re-do the high voltage bias anyway.
I take it the magnetic bias can be pretty weak, and it does not have to be steady on, i.e. it can (should?) be pulsed?
Eldarion
Hi Eldarion,
The PWM is a good idea, but make sure that you use a low pass filter to smooth out the spikes that would result from the switching circuit. This could be done simply with a large capacitor in parallel with the DC output and a filter choke to smooth out any HF hash you might have on top of the cap. Though it the cap is big enough, it usually can get rid of most of this without the choke. Having a pulsed bias does cause interesting effects but it also may complicate the functioning and tuning of your coil.
I also attached two possible ways for you to put an electrostatic HV bias on the output coil without the HV supply draining into the load. I tested out the single coil one and it *seems* to work although you can't measure a voltage difference across the coil since there is no current flow (a volt meter and scope both need to see a current flow in order to read the voltage potential across something). So ultimately you would want something like an electrometer to really observe this, but It may work.
As for the second one, I know that this will work because I've done some experiments with Otto's 50-turn coil where I used a two-wire collector with one being grounded and the other scoped. I found that when there was an electrostatic difference of potential that the collector wire picked up much greater spikes when I pulsed the 50-turn coil. But by far, the best example is the ECD. If you look at the diagram closely, you'll see that the collector loops are simply two capacitively connected wire loops. The load bulb was then just hooked across the loop that was charged positive and the loop that was charged negative... A lot of people missed this tidbit of info when then assumed that the ECD was just a DC-DC step up converter. Anywho, try these two setups out. I'm confident in the second one, but not so sure about the first. Let me know what happens though :).
God Bless,
Jason O
By the way, as another note. The reason that the bifilar biasing coil works is the same reason that Eric Dollard was able to wirelessly charge a capacitor off of RE emitted from the light bulb in his video:
http://video.google.com/videoplay?docid=-6461713170757457294
The bigger the potential difference, the more RE that can be captured. Thats my thought.
God Bless,
Jason O
Jason,
I hooked one of my collector coils up like you described in the bottom picture of your post and ran the TPU with a scope across a 22-ohm resistive load, but all I got was a small amount of hash and the resistor stayed stone cold, regardless of frequency. Putting the generator into the center of the TPU (well, sorta on top of the TPU in the center ;D) did not change anything.
I read your "Phase Relationships and Harmonics" document, and I think I'll implement those suggestions and try again. My pulse sequences are way off of what is suggested there.
BTW, I re-built the bias supply, and now it can generate 150V under significant load from the pulse generator.
Thanks for your help, and I will defnitely keep trying to get this working!
Eldarion
Bob,
Is the attached pulse sequence correct for the rotating 3-phase "small hurricane" device?
All the pulses would be 160V in amplitude; I simply staggered the amplitudes here so that the sequence would be easier to see. The green sequence would go to Primary Coil 1, the red to Primary Coil 2, and the blue to Primary Coil 3.
Thanks!
Eldarion
More like this. Blue is primary 1, Red is primary 2, and Green is primary 3
Bob
intrupt the pulces and reinduse them on the third haromic and repet them n the pulse and intrup again. that it does't have gain in the current in ammps . like petto bismole lol and on farts lol , we need even flow of curret with no farts (amps )Ã, hehehehe make it simple volts .(milie amps) we'er growing them not makeing them !
@BB,
To get this train of pulses with 120 degrees of phase we could have one function generator.
Pulsed into a decade counter with 4 outputs used and the 4th linked to the reset.
1st output = blue
2nd output feed to a flip flop, output = red
3rd output feed to a divide by n set to 4 (or two flip-flops in cascade), output = green
Therefore in relation to the input frequency of f:
blue = f/4
red = f/8
green = f/16
If you need blue to pulse at 35,705 Hz just set f = 148,820 Hz
Red will be 17,852.5 Hz
Green will be 8,926.25 Hz
Simple to setup and scope.
Regards
Rob
Quote from: MeggerMan on August 02, 2007, 09:09:13 AM
@BB,
To get this train of pulses with 120 degrees of phase we could have one function generator.
Pulsed into a decade counter with 4 outputs used and the 4th linked to the reset.
1st output = blue
2nd output feed to a flip flop, output = red
3rd output feed to a divide by n set to 4 (or two flip-flops in cascade), output = green
Therefore in relation to the input frequency of f:
blue = f/4
red = f/8
green = f/16
If you need blue to pulse at 35,705 Hz just set f = 148,820 Hz
Red will be 17,852.5 Hz
Green will be 8,926.25 Hz
Simple to setup and scope.
Regards
Rob
Quite right, very easy to set up, but not adjustable for slight phase variations between phases. Even assuming that we are able to get our primaries precisely located in exactly the right places, a perfectly balanced 3 phase field rotation is not going to put out much like that. That is where a null in the response is, and minimal output occurs.
If it were a case of just needing simple 3 phase output, then we would just need to drive all 3 primaries with a common 3 phase motor controller. As I am told quite often by those that do not understand the precision phase requirements, I could just buy a 3 phase motor controller chip. When I ask them if a chip like this provides precision phase control, of course the answer is no.
Bob
Bob,
Thanks for the diagram! That helps quite a bit.
I should be able to adjust the outputs of my controller to match that pattern exactly.
Eldarion
Hi Bob,
Yes, you quite correct, it's not flexible at all. In fact all you can change is the frequency.
To be able to change the phase will require either a dds chip with 4 outputs with phase adjust or 3 dds chips phase change each one and sync them with a soft reset, which I am not sure if it is possible.
The 4 output chip looks like a go-er but it is expensive, but the nice people at Analogue Devices will send you two samples if you ask nicely.
Phase adjustment using a 4 output DDS AD9959 seems a sensible solution, what are your thoughts?
Regards
Rob
Hi Bob,
Thank you very much for the phase timing diagram. It cleared up alot of haze. Nothing like a visual aid to assist.
If I understand what you are saying though, about the "precision phase control", is that every coil will be slightly different, and the phase will need to be tuned from the open ends of each of the primaries, to get the tuning exact or precise, for "that" set of primaries on that coil?
Eldarion and I have decided to go the six primaries route, when we get to that point. We look forward to seeing the Hex controller you said you did not mind releasing to us. It should give some good ideas, or provide an "exact" replication for some of the guys who want to do that. Our cores will be arriving soon, and I for one am very excited to wind a proper coil.
@ Bob and All
I have spent enourmous amounts of time, redoing the .PDF. Editing it and making it far more professional in appearance as well as much easier to find information based on subject, rather than post. It will be a couple of weeks before it is ready for everyone. Dr. Mark is providing some nice artwork and we are adding a cover. It will also give me time to add the rest of the upcoming photos from the remainder of Bob's build. I have also removed information not confirmed by Bob, so as not to confuse. This .PDF will be something Bob can use in the future for any other replicators he chooses to work with.
Also, with Bob's blessing, I desire, when it is finished, to get a copy to Mannix, to get to Steven Mark. It could be just the thing to encourage the group UEC that controls his tech to also release more information.
Happy Days! :)
Bruce
If you were UEC, would you release more information? Would you want people all over the world making their own power and who knows what else?
Quote from: Grumpy on August 02, 2007, 07:12:24 PM
If you were UEC, would you release more information? Would you want people all over the world making their own power and who knows what else?
I can smell the lawsuits now! :D
You're probably right...no corporation could ever release this information--they would be sued and driven out of business.
Back on topic, I noticed in Bob's diagram that the frequency mix creates groups of pulses, but that every other group is missing only the third pulse. I wonder what effect that causes in the TPU?
This also underscores the need for exact 0?, 120?, 240? timing, as without it, those pulse groupings will not be created.
Eldarion
Bob,
I have one other question for you. What is the recommended peak amplitude for the three primary frequencies? For some reason, I was under the impression that it was supposed to be pulsed high voltage, but I was wondering if it might work with as low as 12 volts?
Thanks,
Eldarion
Well, here is the output on the combined test channel of my controller. We should definitely see some effects here! ;D
Thanks again Bob for the diagram; my pulses looked nothing like this before.
Quote from: eldarion on July 30, 2007, 03:08:45 PM
Jason,
...BTW, I re-built the bias supply, and now it can generate 150V under significant load from the pulse generator.
Eldarion
Eldarion, What did you change to improve you Bias Power Supply? What diameter is that little toroid, what's it made of, and where did you get it? Thank you very much for sharing.
BTW - I am wondering why you guys are not favoring Coarse and Fine Phase Adjust knobs with Jason's RC Delay shown here: http://www.physicsforums.com/showthread.php?p=1392746#post1392746 Thank you also for the Xylinx Spartan thread at http://www.overunity.com/index.php/topic,2894.0.html. I will study this in the future. A fine accomplishment you should be proud of. Yes I see it does Phase also.
Humble
Quote from: wcernuska on August 03, 2007, 01:07:01 PM
Quote from: eldarion on July 30, 2007, 03:08:45 PM
Jason,
...BTW, I re-built the bias supply, and now it can generate 150V under significant load from the pulse generator.
Eldarion
Eldarion, What did you change to improve you Bias Power Supply? What diameter is that little toroid, what's it made of, and where did you get it? Thank you very much for sharing.
In order to improve the output power of the bias supply, I integrated it onto the MOSFET driver board so that I could use the FPGA to generate the exact timing required to generate maximum voltage from that toroid.
The toriod itself was scavenged from a computer power supply; it is about 1 inch in diameter if my memory serves me correctly. Sorry I can't tell you much more about it. You can use almost any type of inductor for this application, as this is just a simple switching power supply. In fact, you probably do
not want to use the same toroid, as maximum power output occurs at a frequency of 1KHz--very annoying! :D
Hope this helps,
Eldarion
Hi Eldarion,
I have been looking at phase shifting and there does not seem to be any easy way around this.
I have tried to simulate an op-amp circuit but I could not see any phase shift at all.
Your circuit seems to solve all the problems in one go.
What is your upper frequency range and what increment can you achieve at this frequency?
It would be nice to take the sine wave output from a function generator and phase shift it into 3 outputs that are 120 degrees apart with the possibility to phase shift them +/- 20 degrees say but I cannot see how this can be done.
May have to wait for Bob to show us his little circuit.
Regards
Rob
Quote from: MeggerMan on August 03, 2007, 07:54:49 PM
Hi Eldarion,
I have been looking at phase shifting and there does not seem to be any easy way around this.
I have tried to simulate an op-amp circuit but I could not see any phase shift at all.
Your circuit seems to solve all the problems in one go.
What is your upper frequency range and what increment can you achieve at this frequency?
It would be nice to take the sine wave output from a function generator and phase shift it into 3 outputs that are 120 degrees apart with the possibility to phase shift them +/- 20 degrees say but I cannot see how this can be done.
May have to wait for Bob to show us his little circuit.
Regards
Rob
Rob,
I have not tested the upper frequency limit, but I would say it might be around 1MHz (it does depend on the loading of the MOSFETs--higher loading and currents cause the MOSFETs and drivers to overheat at those high frequencies). The step precision is 20ns for all parameters, including phase adjustments. The MOSFETs are limiting the upper frequency; the FPGA's output can go up to 50MHz.
The only caveat is that my system outputs pulses, not pure sine waves. I am currently thinking that pulses are the way to go, but I could be mistaken! ;D We'll see when I hook this thing up to a core and start 'er up.
I am getting schematics together like Bruce asked me to; I'll post them in my other thread along with a software update when I finish them.
Eldarion
@Eldarion,
Yes it is pulses we need, nice square ones, I only mention sine waves as they can be used to work out the phase with.
I think I have a solution using the AD9959:
Output the 4 outputs into divide by 'n' chips and use the phase control to setup the offsets.
All 4 outputs are set to the same frequency, are all in sync and the divide by n chips do the exact frequency division.
The AD9515 chip is a delay adjust and divide by. (f/1 to f/32)
Only snag is that the delay adjust is 10ns max. in 16 steps.
Otherwise you could connect 3 x AD9515 to a single function generator.
Regards
Rob
Hello All,
For those wondering about the phase shifting problem. I have been working on this on for quite some time. At this point, I haven't learned how to program a micro controller or an FPGA so I have been figuring out how to do it with discrete components.
Below I attached a PDF circuit diagram of a simulated control circuit that can implement both the pulsed and rotational modes. It also includes precision phase and pulse width control. I have already tested and proven the entire circuit in the simulator and have built part of it with only one channel (have to finish it yet). This circuit should do what we need. and you only need three ICs for and a single function generator input for the control logic. The phase shifting and pulse width control is accomplished with two two RC circuits.
It will work great but keep in mind that the pulse width and phase shift will have to be adjusted every time you change the frequency. But this shouldn't be a major problem once you find the resonant frequency of your coils. For those of you who may have Multisim, I also included the simulation file also. Please let me know what you think.
God Bless,
Jason O
P.S. Note that the Schmitt triggers are not absolutely necessary, you can replace them with regular NAND gates. Also, I'm using all CMOS logic so I can drive the MOSFET drivers on 9V in my actual circuit.
Well, I thought I might as well give an update here.
I have been fiddling with my controller and a decently wound 3-phase toroid I scrounged up from a computer power supply.
I am beginning to think that Earl was right--no matter what I do with my power FETs and driver chips, they suck down tremendous amounts of input power, and develop terrible waveforms across the primary coils of the toroid.
So, I think I'll give the zero input power folks a chance here. ;D I can easily build a board with three of Earl's 10ns pulse generators on it that will plug into my controller instead of the power MOSFET board, and we'll see what happens. I tend to think you guys are correct, as Bob has been strongly emphasizing that extremely short, clean pulses are required. I guess the primaries are more of a "catalyst" to the reaction than anything else!
Eldarion
Hi Eldarion,
A year ago if someone had read to me what Earl is saying today, I would have thought Earl was totally crazy, out of his mind. It is extremely difficult for a traditional brain to wrap itself around RE. The idea that you can have a catalyst with close to no input power is a tough nut to accept. Half of me is still fighting it. Becoming a self-taught Aether Engineer is not easy. I'm trying the best I can and hoping that I am really not crazy.
Regards, Earl
Quote from: eldarion on August 10, 2007, 12:45:52 PM
Well, I thought I might as well give an update here.
I have been fiddling with my controller and a decently wound 3-phase toroid I scrounged up from a computer power supply.
I am beginning to think that Earl was right--no matter what I do with my power FETs and driver chips, they suck down tremendous amounts of input power, and develop terrible waveforms across the primary coils of the toroid.
So, I think I'll give the zero input power folks a chance here. ;D I can easily build a board with three of Earl's 10ns pulse generators on it that will plug into my controller instead of the power MOSFET board, and we'll see what happens. I tend to think you guys are correct, as Bob has been strongly emphasizing that extremely short, clean pulses are required. I guess the primaries are more of a "catalyst" to the reaction than anything else!
Eldarion
Wow is it quiet in here lately! :D
While waiting for the cores to arrive, I have been engaging in several experiments to determine the fundamental principle behind the TPU. From what I can tell, Bob is exactly correct when he says that an EM interference pattern is responsible for the effects. I never really doubted what Bob was saying, but I had to prove it to myself anyway. ;D
From what I can tell, the primaries are simply antennas that radiate to create the interference pattern, and as such, they should be:
a.) Open Ended
b.) Driven with fast pulses
c.) Precise! We may be dealing with GHz interference patterns here.
I found this applet that can simulate 3 source plane wave interference.
http://www.falstad.com/ripple/
Try it! Check "stopped", select 3 source, single frequency interference, and set the speed of the simulation as low as it will go. Crank the resolution up all the way. Arrange the three sources in a triangle pattern (like Bob's TPU viewed from above) and click clear waves. Then uncheck "stopped" and watch!
On the several times I did it, a rotational field was actually set up! It was really cool to watch! It was even unstable (the direction of "rotation" would randomly reverse, etc. Just another thing lining up perfectly with Bob and SM's comments... :))
I am still not fully certain of the mechanism at work behind the HVDC bias, but I think the magnetic bias may help to set the preferred direction of rotation for the vortex--it may stabilize and enhance the reaction all at the same time.
Just some thoughts...
Eldarion
EDIT: I also wanted to mention that since the primary coils are open-ended, they could be acting as 1/4 wave resonators that "amplify" the signals from the pulse generator. Connecting them the conventional way would destroy this effect...
EDIT2: For those wanting more information on quarter wave resonators, here is a good link: http://www.ttr.com/corum/
EDIT3: Take a look here: http://www-personal.umich.edu/~mssd/jed/experimental.html Someone (Bob?) mentioned that the TPU is a magnifying transmitter and receiver all in one...
Bob, I hope you don't mind me reposting this here, but I think it is of critical importance! :)
Quote from: Bob Boyce link=http://www.oupower.com/phpBB2/viewtopic.php?p=14773#14773
Hello Gary.
You are correct, if you drive any single winding on that toroid with an AC waveform, you will not see anything but single phase AC output on all the other windings. Where this phase factor comes into play is, instead of driving with AC, you drive each of the 120 degree coils in phase with very sharp and narrow unipolar negative DC pulses. The energies that are in phase will be additive. Now apply a nominal 160 VDC positive bias (through a high value resistor or LPF ) at one end of the single winding that is wound 360 degrees. At the other end of that winding, use a DC blocking cap, and apply that signal to a load (120V light bulb works well). Now as you are applying pulses and bias, start to shift phase very so slightly of phase B and phase C and observe the effects. You will not need to adjust but a fraction of a degree. If you do not see the effects, then your pulses may not be of sharp enough rise/fall times, or there may be too much of an impedance mismatch between the drive and the transformer. Oh, all applied potentials are in reference to earth ground. You may want to compare the results with and without the 160 VDC bias, reversing bias and applied drive potentials, ect.. and note the observed differences.
The other mode is rotational with 120 degrees of phase differential between A to B and B to C, then vary B and C a fraction from that. But I highly suggest you avoid that until you have a control loop set up that can allow you to shut it down very quickly. It can go out of control suddenly and avalanche if conditions happen to be just right.
Bob
I also highly recommend everyone go and read this thread carefully:
http://www.oupower.com/phpBB2/viewtopic.php?t=1609
A tidbit from that thread:
Quote from: Bob Boyce
The material is important for a several reasons.
1. The material directly impacts the highest frequency that can be applied.
2. The higher the frequency used, the greater the energy density.
3. Since there are bias fields applied, we want a material that will not saturate.
Laminated iron and ferrite fail all 3 of these criteria. Air cores require too much copper for a given frequency. The low permeability of a powdered iron mix is best. So the chosen mix plays a large role in how much copper you will have to wind on it. Since we don't want overlapped layers for any of the windings.
There are many ways to tap into this LEM pie. Almost all share a common thread, ultra fast pulses of pure electric potential
I'm not sure how SM managed to get his TPU to work with no core! At the very least, it was not as efficient as it would have been with a core...
Everything makes sense now--thanks again, Bob! ;D
Eldarion
Quote from: eldarion on August 15, 2007, 08:40:38 PM
I found this applet that can simulate 3 source plane wave interference.
http://www.falstad.com/ripple/
Try it! Check "stopped", select 3 source, single frequency interference, and set the speed of the simulation as low as it will go. Crank the resolution up all the way. Arrange the three sources in a triangle pattern (like Bob's TPU viewed from above) and click clear waves. Then uncheck "stopped" and watch!
On the several times I did it, a rotational field was actually set up! It was really cool to watch! It was even unstable (the direction of "rotation" would randomly reverse, etc. Just another thing lining up perfectly with Bob and SM's comments... :))
I am still not fully certain of the mechanism at work behind the HVDC bias, but I think the magnetic bias may help to set the preferred direction of rotation for the vortex--it may stabilize and enhance the reaction all at the same time.
Hi eldarion,
Tried, but could not get a rotation.
Three points echoing into the centre then radiating out from the centre, but no noticable rotation.
Can you give all the settings for each drop down and slider.
Regards
Rob
Quote from: MeggerMan on August 16, 2007, 06:00:47 PM
Quote from: eldarion on August 15, 2007, 08:40:38 PM
I found this applet that can simulate 3 source plane wave interference.
http://www.falstad.com/ripple/
Try it! Check "stopped", select 3 source, single frequency interference, and set the speed of the simulation as low as it will go. Crank the resolution up all the way. Arrange the three sources in a triangle pattern (like Bob's TPU viewed from above) and click clear waves. Then uncheck "stopped" and watch!
On the several times I did it, a rotational field was actually set up! It was really cool to watch! It was even unstable (the direction of "rotation" would randomly reverse, etc. Just another thing lining up perfectly with Bob and SM's comments... :))
I am still not fully certain of the mechanism at work behind the HVDC bias, but I think the magnetic bias may help to set the preferred direction of rotation for the vortex--it may stabilize and enhance the reaction all at the same time.
Hi eldarion,
Tried, but could not get a rotation.
Three points echoing into the centre then radiating out from the centre, but no noticable rotation.
Can you give all the settings for each drop down and slider.
Regards
Rob
Rob,
I'll get a couple of screenshots together soon--I'm currently winding a test coil.
Since we cannot introduce any phase shift in the simulator, you may have to place each of the three souces slightly off from a perfect triangle.
Eldarion
Hey Eldarion,
Great finds there on OUpower.com. Thanks for passing them along to us. That just helped me focus on what is what here. I have been meandering around in several different areas of experimentation but knowing that it's all about the short, fast pulses, helps keep things in focus for me. Speaking of which, I need to ask Bob about his Hex controller. I'm seriously thinking about going the FPGA route but seeing how he is driving the coils will be helpful.
God Bless,
Jason O
Quote from: eldarion on August 15, 2007, 08:40:38 PM
Wow is it quiet in here lately! :D
While waiting for the cores to arrive, I have been engaging in several experiments to determine the fundamental principle behind the TPU. From what I can tell, Bob is exactly correct when he says that an EM interference pattern is responsible for the effects. I never really doubted what Bob was saying, but I had to prove it to myself anyway. ;D
From what I can tell, the primaries are simply antennas that radiate to create the interference pattern, and as such, they should be:
a.) Open Ended
b.) Driven with fast pulses
c.) Precise! We may be dealing with GHz interference patterns here.
I found this applet that can simulate 3 source plane wave interference.
http://www.falstad.com/ripple/
Hi,
I played with it and I came up with just a setup of 2 single oscillators
with 2 different frequencies.
If you set this up, like in the attached pics,
that there is a beat frequency between the 2 oscillators
(could be coils)
there is a big changing zone between the 2 coils, where
you could place a load coil and
thus induce a huge changing magnet field.
the second picture I attached, shows an interesting pattern,
as this could be used to get a unidirectional motion
from 2 coils or toroids connected onto a platform,
so to violate Newton?s third law of motion.
As the radiation pattern of attraction and repellation
is now asymmetrical, this 2 coil device should be
able to move into one direction.
What do you think about this ?
Quote from: hartiberlin on August 17, 2007, 08:23:39 AMWhat do you think about this ?
VERY interesting! ;D
I wonder how much power a setup like this would take to manifest the effect?
Quote from: eldarion on August 17, 2007, 01:02:02 PM
Quote from: hartiberlin on August 17, 2007, 08:23:39 AMWhat do you think about this ?
VERY interesting! ;D
I wonder how much power a setup like this would take to manifest the effect?
If you drive the coils with one wire, it won't cost you anything except the power to drive the control logic ;D All we pay for is the oscillation. The induced fields in space do all the work for us :).
God Bless,
Jason O
From TPUs to starships, very nice Stefan :)
It's fun playing around with simulators isn't it? However, we need to realize they often don't tell us the whole story.
What's missing in the Hungarian and other concepts of interference is what happens at the SOURCE and in general SPACE.
Yes, waves add in space and energy seems to increase in one place and decrease in others, so if you place your receiver in one "high" intensity spot, you might get 0.001% of the TOTAL transmitted energy. Place the receiver at a "low" field intensity spot and receive perhaps 0.000001% of that total. A relative difference of 1000 times.
Do you see what I'm saying? It's not just about the concept of interference and energy increasing, but about how much energy your inputing in the system. The transmitters puts out a lot more energy in space then one little isolated spot. For those that browsed gn0sis.com I had a few MATLAB simulations of pulses adding and showed the energy was constant when integrated across the whole space. :( (but I was exicted for a few days before that) :)
Then I mentioned the source. One wave from one transmitter affects the other, and a Z matrix is often computed for phased arrays or multiple antennas close to each other. The interactions can be quite complex in multi source systems.
However, the TPU seems to be simplistic and trap the waves in a circular ring. Now it should be simpler to analyze since its a 1-D setup as far as the waves are concerned, they just go round and round passing one transmitter then the other in a seeming LINEAR path. Do the pulses go BOTH ways, as in clockwise and couter-clockwise? I don't know, the trick I belive is to make them go one way, by using DIRECTIONAL Couplers.
BEP, you mentioned you have a sloted antenna on your roof, and you're generating more power? I hope you understand what GAIN and DIRECTIVITY is all about. Just because your sloted high gain antenna transmits more power in one direction then others, and has a very narrow beamwidth, that doesn't mean its now creating more energy by this interference concept. I'm an antenna design engineer and have tested lots of antennas in the chambers to extract gain and pattern and I never received more energy out then was transmitted, perhaps I should stop using standard physics ;D
EM
All,
Just a quick status update here. As Jason knows, my aging Spartan 3 board finally kicked off, and I had to pull in a replacement to allow the experiments to continue. So I am now using a Spartan 3E board as shown; this has significantly more features, including two onboard DACs and two onboard ADCs, not to mention that oh-so-cool LCD screen. :)
Best of all, this board is the same price as the Spartan 3 -1000 I was using ($149 USD).
The core in the picture is what I am currently working with. I remembered a couple of days ago that I had an AMCC-4 nanocrystalline core left over from some MEG experiments that I had been doing, so I just had to try it! :)
This core is simply amazing in terms of performance, permeability, etc. The only problem is that I have to drive it four-phase, since there is no way to do symmetrical three-phase. This has required some software updates, but nothing too major.
Attached is a waveform with all four of the MAX627 driver chips connected to the driver coils, but with all open-ended except one. That interesting spike and the even more interesting DC offset only came about when I hooked one of the primaries open end to ground. I have tried hooking up the other primaries free ends to ground as well, but as soon as I do the input current spikes and I have to shut down to protect the MAX627s. Time to break out the MOSFET driver board again, I think. ;D
This measurement was taken with a 12V DC bias source, a 0.22uF DC blocking capacitor, and a 470-ohm load resistor. The scope leads were placed directly across the resistor, and the input power was 0.25 watts.
Eldarion
EDIT: I forgot to mention that the windings on the core are all 1:1; i.e. each winding is comprised of 10 turns of 16ga. copper magnet wire. So, that spike in the scope shot (~10V) is not due to transformer step-up...
To those interested in my odd experiments: ;D
The AMCC-4 core bombed. I was getting a mere 10% of the energy I put in out, and to top it all off, two of my windings shorted out. I consider this attempt abandoned, as a high permeability core seems to completely destroy the energy conversion process.
On the plus side, I did find out why. On this page: http://sound.westhost.com/xfmr2.htm (a recommended read for anyone working on this project) I found that an powdered iron core transformer is a wideband transformer, used where there is significant DC bias, and it does not saturate easily at all.
I'll see what I can scrounge up as far as powdered iron toroids are concerned, and give that a shot while I wait for the real core. :)
Eldarion
Well, I finally found the needle in the haystack! In an old DC power supply I came across a large powdered iron toroid, quite possibly a Micrometals mix 26 from what I can tell!
Attached are a couple of pics. I still need to "beautify" the secondary, and wind the primaries, but I think this might actually work! :) Also, I will be seeing if I can preload a magnetic bias with permanent magnets, as I think this might be safer than using a bias winding--we will see. ;D
Here is a page with the exact toroid I scrounged up:
http://www.coilws.com/toroidal/Vicor_choke.php
Eldarion
Quote from: eldarion on August 19, 2007, 07:10:29 PM
That interesting spike and the even more interesting DC offset only came about when I hooked one of the primaries open end to ground.
We had a simallar experience when I was testing things at Grhams Gundershans lab. We got the strangest resonance shifts that took many seconds to establish -- with what Graham said was overunity behaviour... but we only got this when one of the open primary coils was connected to ground through the signal generator! ... in turns out that there is a varistor in the signal ground protection circuitry. So I would encourage people to add non linear elements grounding open coils to their list of things to test...
We got the weirdest behaviours driving just 3 of 4 primary coils with the fourth coil open. We were driving with two signal generators. One driving a pair of opposite coils in seris -- but open ended. The second generator was driving a thir primary coil and we were probing the fourth which was grounded at one end (the other end open) via the signal generator. When driving with several frequency combinations we got what appeared to be overunity signature in reverse -- the inability to force resonance at one frequency. No matter how slowly we swept through the resonance peak the output would begin to rise exponentially and then stick and skip past the peak. The implication was that on resonance there was a shift in the speed of light that always caused the external signal to drift of resonance... If the drive signal originated within the device then the hope is that we could have tuned stably on (or more closely at least)resonance and maintained overunity output.
it's all about control -- not brute force!
mark.
Mark,
Thank you for your insights! Sounds like you got some interesting results.
Can I ask a couple of questions?
Was this setup toroidal, and was it air core or some other core? I assume the windings were 4-phase?
And there was a varistor between one end of the "output" coil and ground?
Also, do square waves work or just sine waves?
If square waves will also work, then my controller may be able to keep this thing in resonance...
Thanks!
Eldarion
Hi Eldarion,
We were using a very large toroidal ferrite core -- 6 inch diameter and 1" square cross section. They are a high performance custom core Graham had made up.
Graham had one that was wound with 4 single layer toroidal coils. The 4 coils were identical and covered 90 deg of the toroidal arc. I'll have to refer to photographs to check the number of turns -- somewhere about 50 turns in each coil at a guess.
Driving with square waves or sine waves gave the same result -- because we were driving the coils at resonance -- so the resultant wave was sine regardless of the drive. The coils were all driven as open ended transmission lines.
I brought two of the cores back to Australia with me to continue the work.
On driving wave forms: -- I always tune for resonance with a square wave. It's easier to see if you are exactly on resonance when driving with a square wave and you will get as good or better drive with a square wave. Of course the resultant resonant waveform is always going to be a sine wave at resonance regardless of the driving wave form... although you do want to drive with a higher impedance than the natural impedance of the resonant circuit... unless you are intentionally driving a seris resonant circuit in which case you want to drive with as low as impedance as you can.
Normally, once you have found the resonant frequency you can change to a sine wave drive with little or no change. However I have some resonators that will only work when driven with a square wave -- thats due to a very low impedance resonator being driven by a lightweight drive coil. I have never found the reverse -- a resonator that will only respond to a sine wave and not to a square wave -- practically and theoretically that just never occurs. From an efficency standpoint a square wave drive will always be more efficent if you have your impedances matched.
Mark,
Thank you for the explanation. That helps a lot!
I can't wait for the first test of a real coil wound on the correct core... ;D
Eldarion
Well, I may be getting closer...I have never seen waveforms like this before!
It would appear that the primary coils should have non-negligible inductance, as far as I can tell. Also, some current flow through the drive coil(s) (very little in my experience) seems to be required.
This waveform came about with 500ns pulses of 13.8V. As you can see, the resuting voltage at the switched coil lead (the lead connected to the MOSFET drain) is far larger and more interesting.
I will wind three "input" coils with ~30-40 turns each and see if the same waveform crops up.
Has anyone else seen these waveforms? I can see what appear to be the Barkenhausen jumps at switchoff, but I have no explanation for the sawtooth waves. My power supply was clean as a whistle during the entire test.
Eldarion
@ All
Well I came across some magnet wire here at the ocean, on the last leg of my vacation and had to begin to wind. (I have 6 pounds of 20 gauge on my porch of my farm, delivered by UPS, back home.)
It took a bit to get the hang of winding. Bob is right when he says that bee's wax is your friend! ;D I got the hang of it, and it went fairly quickly. The other winding will be easier, now that I have the hang of it. I probably over waxed it, but the windings aren't going anywhere! ;)
It took me 1 hour 44 minutes for this one winding. Wound the Magnetic bias longitudinal winding first. Counter clockwise, starting from the top to the bottom. 48 turns. Labeled the leads. Tomorrow I will wrap the top. From what Weri tells me, the inside core is the tough one.
@ Eldarion
A "proper" Bob Boyce coil is coming soon! A "real" coil fitting of that controller! ;D Then we will see what sort of waveforms you generate! ;)
Warm regards to all,
Bruce
Hi Eldarion, hi All,
The gate oscillations on the scope photo appear to me to be excessive.
The attached image shows the way to perfection, if you can't reach perfection,
then at least know what it is and try to come as close as possible.
The gate charging current does not come from the power supply; it comes only
from the by-pass capacitors on the FET driver.
I can not stress enough how important it is to keep total loop lengths as small
as possible.
Hi Earl,
I thought XR7 were high quality caps.
I have just order a delay line device for creating a programmable pulse width:
DS1023S-100+
http://datasheets.maxim-ic.com/en/ds/DS1023-DS1023S.pdf
10ns to about 250ns in 256 taps (1ns steps).
Whats more, it can also be used to offset the input pulse by the programmed delay time, so you could use it for fine phase adjustments.
These come in a nice and small SOIC package but they are expensive, about 9.50 GBP each, Digikey seem to sell them a lot cheaper than this, but for me to purchase from Digikey I need to order 100 GBP worth of parts to get the free postage.
Regards
Rob
Earl,
Just so you know, since the gate drive was not the "primary" signal I was measuring in the picture above, I used an old crummy makeshift probe on that signal. The oscillations are far less when I use the correct probe to measure that signal!
I have extensively bypassed the gate driver chips with 0.1uF ceramic disks directly connected to the power supply pins (I mean zero lead length here ;)) This is in addition to the heavy electrolytic filtering I have on the power supply input.
However, if this is not adequate, I would welcome other suggestions! ;D
Eldarion
Hello All,
Just over two hours winding time last night, for the top longitudinal winding. It was a bit tougher. Sore thumb.. ;D
Edit: I wound CCW, winding from the inside to outside, 26 turns on this one.
Hello All,
Another 2 hrs. and the bottom winding is wound. This was wound CW, *NOT* CCW, because when flipped, the CW will become CCW. It was wound from the inside core to the outside perimeter CW at 29 turns. More wax on the core, before starting, made the task a bit easier. Next is the difficult part... the inner core winding! ;D
I think I will use Wer's idea of the spray adhesive, along with bee's wax to make it work. My bees wax is in small little chips, making smearing it on, rather easy.
After each winding, that individual winding is taped. When all windings are completed, I will tape the entire core. And then it will be time for the secondary.
I also included a picture of the sunset over the Gulf of Mexico, taken Thursday evening, aboard a ship, simply because it was so beautiful and I thought some would enjoy it! ;)
Cheers,
Bruce
hello Bruce
your coil looks good 8)
i did not use any wax spray only :)
got one primary done will wind another one to night after work ;D
be talking later
wer
@Rob,
by high quality I mean low series resistance and usable into the VHF / UHF / range and beyond.
XR7 allows high C values in small space and are good for general
low-frequency by-passing, but are not desirable for coupling in low-noise amplifiers. I have seen noise figures jump from 1.5 to 9 dB because of noisy, lossy dielectric. A 330 pF SMD cap with low-loss dielectric is always nice to have in the parts box. XR7 is inherently lossy and noisy, which for many applications is acceptable. If you want to deliver Amps in nsec, I recommend
330 pF NPO or COG
1 to 10 nF
10 to 100 nF
all in parallel, all SMD, for FET driver by-pass.
@Eldarion
An oscilloscope with AC coupling and good HF scope probe (compensated while watching square wave) will show voltage sag on by-pass capacitors during gate charging.
What does the gate voltage look like when using a good probe?
Regards, Earl
Quote from: MeggerMan on September 12, 2007, 09:47:28 AM
Hi Earl,
I thought XR7 were high quality caps.
[snip]
Regards, Rob
Hi Earl,
I would post a pic of the gate waveform with a good probe--if I had a good probe! :-[
The best probe that I have is a 1x cheapie. I am looking into getting some better x10 probes--something that I've put off for far too long!
I assume the gate waveforms are OK because I can scope across the FET output, and when the FET turns on the voltage drops straight down to zero--very little slope from what I can see. Also, the FETs are stone cold, even at high frequencies.
I know the above observations will not replace good scope measurements, but that's all I can do at the moment...
Eldarion
Hello All,
Well, I have just finished the dreaded inner core! ;D It took about 4 hours, with a lunch break about halfway through. 48 turns CCW, winding from the top to the bottom.
Taped everything up real tight, afterwards, and tomorrow I will start on the secondary.
Ended up using only wax for the center. Perhaps my next core, I will try a different adhesive.
Cheers,
Bruce
If you absolutely have to, you could build your own 10:1 probe.
http://zone.ni.com/devzone/cda/ph/p/id/17
shows how - along with the equation R1C1 = R2C2.
One of the caps is a small variable to do frequency compensation while watching a square wave.
It could be that making C1 fixed with C2 variable is more practical.
Searching ebay for 10:1 probe or 10x probe might turn up something?
Regards, Earl
Quote from: eldarion on September 15, 2007, 12:49:38 PM
Hi Earl,
I would post a pic of the gate waveform with a good probe--if I had a good probe! :-[
The best probe that I have is a 1x cheapie. I am looking into getting some better x10 probes--something that I've put off for far too long!
I assume the gate waveforms are OK because I can scope across the FET output, and when the FET turns on the voltage drops straight down to zero--very little slope from what I can see. Also, the FETs are stone cold, even at high frequencies.
I know the above observations will not replace good scope measurements, but that's all I can do at the moment...
Eldarion
Hello All,
Well, the secondary is wound. How long did it take? I lost count! ;D Winding it didn't take very long, but "straightning" and "spacing" and "waxing", well, let's just say that it took a while. ;)
The first picture is of the secondary (16 gauge, teflon coated solid, silverplated wire) taped with #35 electrical tape. Over this I applied the motor tape. It wrinkled, but still makes a "solid" surface for the last windings, that of the 6 primaries. (To be wound with 20 gauge teflon coated, solid, silverplated. The red wire in picture)
Happy winding to all,
Bruce
(I still have to wrap another core after this one is finished! My fingers are sore thinking about it.)
Hello All,
I first marked the core for all SIX primary placements. ;)
And then I wound 28 turns, waxed and taped my first of six primary segments. I am using 20 gauge, solid, teflon coated, silver plated wire for the primaries. Winding is from left to right, CCW.
I hope to do one a night, over the course of the next five nights.
Cheers,
Bruce
Coil is looking excellent, Bruce!
Are you going to precisely space the primary windings like you did to the secondary? (Not sure if that's really neccessary, just thought I'd ask.)
Attached is a (blurry, sorry!) screenshot of the current controller, displaying fake data of course. ;) As you can see, I am planning to make the controller able to handle two cores at once for increased power output...I still have a lot of work to go...
Eldarion
Quote from: btentzer on September 23, 2007, 01:17:47 AM
Hello All,
I first marked the core for all SIX primary placements. ;)
And then I wound 28 turns, waxed and taped my first of six primary segments. I am using 20 gauge, solid, teflon coated, silver plated wire for the primaries. Winding is from left to right, CCW.
I hope to do one a night, over the course of the next five nights.
Cheers,
Bruce
Tried posting this last night but my browser kept crashing when I loaded pages here...
Bruce, hard to tell here on this small laptop screen, but that sure looks CW to me. It looks like the camera is closest to the finish end of the winding, and the winding at that end is exiting at the top, When I tried clicking on the image to save so I could open it in a program and blow it up, the image would not download. So I can't be sure. Remember, the way to determine direction of wind is by looking at it from the point of view of CCW wave fronts passing through the centerline of the coil from start to end. The windings should be corkscrewing CCW as you are passing through that coil from start to finish.
Bob
Quote from: eldarion on September 24, 2007, 01:44:13 AM
Coil is looking excellent, Bruce!
Are you going to precisely space the primary windings like you did to the secondary? (Not sure if that's really neccessary, just thought I'd ask.)
It is neccessary, but the spacer diameters would be different due to the thinner wire.
Quote from: eldarion on September 24, 2007, 01:44:13 AM
Attached is a (blurry, sorry!) screenshot of the current controller, displaying fake data of course. ;) As you can see, I am planning to make the controller able to handle two cores at once for increased power output...I still have a lot of work to go...
Eldarion
Yea, I can't make it out on my tiny laptop screen.
I am almost ready to order the first batch of boards for my new HexController layout. It has 6 channels of output for driving primaries, and 2 channels of output for driving bias. I guess it could also be used to drive a pair of 3 primary units, but I think that might be pushing it a bit ;-)
Bob
sorry Bruce
that is cw like mine was
go back and look at the two cores i posted
hope that your secondary is not wound that way
@ Bob
waiting on the hex controller files
wer
@Bob,
On the primary, the input would be on the left, where the lead is on the bottom and then corkscrews CCW and ends on the right, which would be open ended. The wire is wound CCW, from left to right and Corkscrews CCW from left to right.
The only way one could see it as CW is if the input is on the right and I were winding the corkscrew clockwise from right to left.
So, looking at it right to left it is CW, but I thought the timing and input had the field traveling CCW, which would be left to right??
EDIT: I will take it down to the longitudinals if you say it is CW. So, I guess I want to verify, that for the Secondary and the Primaries, we want to wind CCW winding from Right to Left?
And then the far left lead on each primary is the open ended lead?
For a CCW rotation field core, drive end of an open ended coil is on the left, IF you are facing the toroid with the winding directly in front of you (not on the far side away from you). A CCW winding has the start lead of the coil (on the left) going away from you, over the top, down the inside, back towards you on the bottom, up the front, and over the top to the right of the starting point. Continue winding in that direction. The finish end lead of the winding will exit towards you from the beneath the bottom of the core to the right. Hope this helps.
As you energize the coil from the left, the potential shock wave quickly travels down the coil from left to right, pushing the ferroresonant field along inside the core in a CCW direction, imparting some spin to that field. The direction of that spin is directly related to the direction of wind. We want that spin to be CCW (looking towards direction of wave travel, not looking at the wave coming at you) impressed upon the CCW rotating field.
The bias winding(s) increase or decrease this spin by aiding or fighting. The field direction is down through the hole in the center, and up around the outside perimeter, once all of these individual little micro-fields come into alignment.
hi Bruce
it took me a little time for it to sink in
think of it like a lefthand treaded bolt
when you look down the threaded end the threads goe ccw just like on my picture of cores
thanks just trying to help ;D
wer
@ Bob,
Thank you, that does indeed help. I will wind the exact opposite of what I have done! LOL Well, I will consider it good winding practice anyway, for this next go around.
@ Wer,
Thanks! It does take a bit to wrap ones mind around it. CCW I hope! ;D
Well, I unwrapped my two primaries and untaped, to my secondaries, to find that I had wound that correctly according to "B" of this graphic provided by Wer off of his thread.
So I have retaped and will remark and rewind primaries! All is well again in happy winding land! ;D As long as I did not have to redo the secondary, I am thrilled!
@ Wer
I added your graphic to the continually updated .PDF, if you don't mind. And reposted it here. ;)
Bruce,
Good to hear all is well again! ;)
I have also had a minor success here--I have improved the control software and stabilized the output. So, I thought, why not hook up the largest powdered-iron core toroid I have to it and try it again? I only have one primary coil wound on this particular toroid, but I might just add the others as time permits...this is the same toroid that I posted earlier wrapped in black tape with copper wire.
This time, when I fired up the primary on 10KHz 500ns pulses, I got a HUGE output waveform on the secondary. And it looked quite odd, almost a very jagged triangle wave at ~200+V peak-peak. The control circuit couldn't have been drawing more than a few hundred milliamps at 13.8V (this was with the HV bias supply shut down). So I went to touch the output wire to make sure that I wasn't just seeing some sort of phantom capacitive voltage rise...surprise! I got an almost instant RF burn from the output, and the magnitude of the output signal didn't decrease much at all.
So I am finally getting a pretty good power transfer through the toroid. I have no idea what the actual output is, but at least all my input power isn't just "vanishing" into thin air like it did before...and the odd waveform has me intrigued...
Eldarion
I need to correct a mistake I made in my previous post.
Where I said the pulse width was 500ns, it was actually six times larger at 3000ns, so I was drawing considerably more power than I had originally thought. I still get power transfer with the correct pulse width of 500ns, just not as much voltage on the output and the signal becomes a bipolar sawtooth wave.
That's what I get for not reading the data on my controller screen before posting... ::)
Oh, and by "considerable", I mean 0.55 watts. ;D
Eldarion
Well, Bruce's excellent coil came today! ;D
I have attached a picture of the coil with a 9v on top of it for scale--this thing is huge and
heavy!
Also attached is a picture of my preliminary hookups and the waveform I got across a load resistor with a DC blocking capacitor in the secondary circuit, just like Bob described earlier.
As you can see, not much yet, but I've had the coil for about an hour. I'll see if I can get anything interesting to happen later on. :)
This is how I will be going about initial testing of the circuitry and the coil:
QuotePostPosted: Mon Aug 13, 2007 10:46 pm
Hello Gary.
You are correct, if you drive any single winding on that toroid with an AC waveform, you will not see anything but single phase AC output on all the other windings. Where this phase factor comes into play is, instead of driving with AC, you drive each of the 120 degree coils in phase with very sharp and narrow unipolar negative DC pulses. The energies that are in phase will be additive. Now apply a nominal 160 VDC positive bias (through a high value resistor or LPF ) at one end of the single winding that is wound 360 degrees. At the other end of that winding, use a DC blocking cap, and apply that signal to a load (120V light bulb works well). Now as you are applying pulses and bias, start to shift phase very so slightly of phase B and phase C and observe the effects. You will not need to adjust but a fraction of a degree. If you do not see the effects, then your pulses may not be of sharp enough rise/fall times, or there may be too much of an impedance mismatch between the drive and the transformer. Oh, all applied potentials are in reference to earth ground. You may want to compare the results with and without the 160 VDC bias, reversing bias and applied drive potentials, ect.. and note the observed differences.
The other mode is rotational with 120 degrees of phase differential between A to B and B to C, then vary B and C a fraction from that. But I highly suggest you avoid that until you have a control loop set up that can allow you to shut it down very quickly. It can go out of control suddenly and avalanche if conditions happen to be just right.
Bob
(Emphasis mine on things I think are important)Obviously I will not keep it in that low-power mode, but I would like to see the above actually happen and get a feel for how the coil behaves before I try to get the finicky, dangerous high-power output running.
One other thing--I am currently set up to test this in my basement. If anyone thinks that this will prevent the effects from manifesting, please speak up! :)
Eldarion
Well, I did some calculations, and I am seriously beginning to doubt the 500ns pulse width.
Here's why:
The PWM3E system has a minimum output pulse width of 11000ns (11us).
That same system has a 500ns minimum rise time into the gates of the MOSFETs alone.
The impedance into the primary coils at 2MHz (equivalent to a 500ns pulse) is about 4 kilohms!
There isn't even enough time to build up a tiny magnetic field with a 500ns pulse.
I am getting zero output right now at 500ns.
I will be testing larger pulse widths later on today.
Eldarion
Ok, here are the preliminary results:
4.5W in, 1.9W out. This was rectified, filtered DC, pulse width 11000ns, F1 10KHz.
However, this setup gave me an odd dependence on the HV potential--changing this potential had no electrical effect, but this happened instead:
The core gives off a small audible signal when running, in this case it sounded like 10KHz. When I connected the HV potential circuit and turned it on, this audible signal doubled in apparent intensity. It also displays a very strong correlation with the voltage of the HV potential, far stronger than that tiny oscillation on the output signal I noted before.
I will be trying different output circuit configurations to see if I can couple into this effect at all and get an electrical output power increase. For all I know, I am outputting tons of longitudinal energy and don't even know it, but the core just might have "known" and started to sing in response! ;)
Eldarion
This is interesting from D9.pdf:
QuoteThe objective here is to have the complex waveform generated by the electronics produce voltages of about 25% of the main power supply voltage at the electrolyser. In other words, if an inverter is being used and its output rectified to produce about 155 volts of pulsing DC, then the toroid transformer secondary should generate about 40 volts.
The output from the electronics board is about 13.8 volts when driven by a vehicle?s electrical system, so to step that up to about 40 volts requires a step up of 2.9, which means that the secondary winding needs to have 2.9 times as many turns in it as the primary winding does. So divide the number of turns in your secondary winding by 2.9 to calculate the number of turns in each of the three primary windings. If you had 140 turns in the secondary, then there would be 48 turns in each of the three primary windings.
We only have 28 turns on each primary!
Hmmm....
Eldarion
Quote from: eldarion on October 12, 2007, 10:45:38 AM
This is interesting from D9.pdf:
QuoteThe objective here is to have the complex waveform generated by the electronics produce voltages of about 25% of the main power supply voltage at the electrolyser. In other words, if an inverter is being used and its output rectified to produce about 155 volts of pulsing DC, then the toroid transformer secondary should generate about 40 volts.
The output from the electronics board is about 13.8 volts when driven by a vehicle?s electrical system, so to step that up to about 40 volts requires a step up of 2.9, which means that the secondary winding needs to have 2.9 times as many turns in it as the primary winding does. So divide the number of turns in your secondary winding by 2.9 to calculate the number of turns in each of the three primary windings. If you had 140 turns in the secondary, then there would be 48 turns in each of the three primary windings.
We only have 28 turns on each primary!
Hmmm....
Eldarion
Wow, finally able to read and respond here now that I am on a broadband WiFi connection ;-)
I didn't write that D9.pdf document.I never gave exact turns for anything. It was always stated as a ratio, 25% of the HV bias applied to the water, and that ratio ONLY pertains to the drive necessary to get reaction of water to absorb LEM. I think Patrick took absolutes from some cores I wound in the past where I posted turns counts for a particular core and application. I had never wound a T650-52 before, so there had been no data to base turns count on for that particular core.
You're looking at the wrong source of info. As I had stated elsewhere, the hydroxy stuff has certain requirements that must be met in order for water to be entrained and respond to LEM. The energy application of this technology is not limited by these constraints. It is actually much simpler to tap LEM for energy, just harder to get LEM > TEM conversion to take place without water. Any further questions please post in a thread that is not loaded with images so i can respond from dialup at home ;-)
Keep at it, you should be able to get some positive results with few changes to the setup you are using.
Bob
Hello all,
I just got off the phone with Bob a half hour or so ago. He was very helpful! ;D I now have a much better understanding of the energy converter technology...
From what we can tell, my output stages are not able to drive the coil properly, hence my lack of results I have ordered, at his recommendation, a matched set of four IGBTs, part # MG400Q1US41(EP). I have also ordered some UCC37322 MOSFET drivers to make absolutely certain that I have the correct waveforms into the IGBTs. ;)
I will post schematics of the new output driver design after I have finished and smoke-tested it!
Also, something else interesting. The secondary outputs pure LMD, and cannot be used to close the loop without conversion from LMD back to TEM. Where you should be getting TEM output is on the longitudinal pancake windings, and that should be DC and in overunity quantities (here on Earth, anyway :))
Eldarion
Quote from: eldarion on October 13, 2007, 06:30:42 PM
Hello all,
I just got off the phone with Bob a half hour or so ago. He was very helpful! ;D I now have a much better understanding of the energy converter technology...
From what we can tell, my output stages are not able to drive the coil properly, hence my lack of results I have ordered, at his recommendation, a matched set of four IGBTs, part # MG400Q1US41(EP). I have also ordered some UCC37322 MOSFET drivers to make absolutely certain that I have the correct waveforms into the IGBTs. ;)
I will post schematics of the new output driver design after I have finished and smoke-tested it!
Also, something else interesting. The secondary outputs pure LMD, and cannot be used to close the loop without conversion from LMD back to TEM. Where you should be getting TEM output is on the longitudinal pancake windings, and that should be DC and in overunity quantities (here on Earth, anyway :))
Eldarion
@ Eldarion,
That is very, very good news! ;) Identifying the problem is good news indeed. You may want to post what you were using, compared to the new set of four IGBT's you just ordered and why the one set should work, while what you have, does not. This may help some of the replicators reading this. (and will also satisfy my own curiosity! :) )
Good job Eldarion, we look forward to Thor 1.2!
Warm regards,
Bruce
Hey Eldarion,
It's great to hear that you got your problem solved!! I'm excited to know that you were seeing some results from the high voltage that you applied to your coil as well!
I'm still working out the final pieces of my control circuit. I'm working out the details for a card that will measure the output voltage/current from the DC longitudinal windings to control the output from the coil. Once I finish with this, I'll be ready to make the PCB designs and have it built. You're doing excellent work man. Keep up the great work!
God Bless,
Jason O
Quote from: btentzer on October 13, 2007, 10:20:44 PM
@ Eldarion,
That is very, very good news! ;) Identifying the problem is good news indeed. You may want to post what you were using, compared to the new set of four IGBT's you just ordered and why the one set should work, while what you have, does not. This may help some of the replicators reading this. (and will also satisfy my own curiosity! :) )
Good job Eldarion, we look forward to Thor 1.2!
Warm regards,
Bruce
Hi Bruce,
I had already posted the schematic for Thor 1.0's output stage on my controller thread here:
http://www.overunity.com/index.php/topic,2894.msg52205.html#msg52205
I will also be posting the schematic for the new Thor 1.1 output stage on that thread, but not before I smoke test it. ;)
Basically, it looks like the IRF510s I was using could not switch fast enough, thereby causing a tremendous loss of power and
lots of ringing in the primaries, which completely destroyed the energy conversion process. Bob has informed me that the output, when shorted, should be a flame-like discharge, even without the HV bias connected (just shorting the two secondary terminals together). There should also be a choke on the +13.8V side of the primaries to prevent hash from getting into the power supply. Note that this choke has to also allow some HF power through it in order to get power to the primaries, so throwing a massive 50 henry choke in there will completely destroy the effect. ;D
Thanks to Bob for taking the time to explain this stuff to me! :)
Oh, if anyone in this group would like to take advantage of my controller once everything has been finalized, I will be offering the control software for free to the members of this group...
Eldarion
EDIT:Just realized this was posted while I was typing this reply...
Quote from: Jdo300 on October 14, 2007, 12:12:36 AM
Hey Eldarion,
It's great to hear that you got your problem solved!! I'm excited to know that you were seeing some results from the high voltage that you applied to your coil as well!
I'm still working out the final pieces of my control circuit. I'm working out the details for a card that will measure the output voltage/current from the DC longitudinal windings to control the output from the coil. Once I finish with this, I'll be ready to make the PCB designs and have it built. You're doing excellent work man. Keep up the great work!
God Bless,
Jason O
Thanks! ;D I wish you well in your controller design! (I will be doing a similar thing; measuring the longitudinal power and controlling the reaction with that information).
Also, a note to all. Please, Please do NOT connect your scopes to the output, especially when the load is connected!!! Bob had someone blow up a 75W bulb on the output with very little power on the input (it turned into a crude arc light for a while); just imagine what would have happened to his scope had it been connected! :o :-X Besides, the secondary outputs pure longitudinal energy; your scope
will not detect it. Believe me, scoping everything is a very hard habit for EE types to break. Just don't make it an expensive one, too. ;)
Thanks for the scope tip, I happen to be one of those EEs who scopes everything.
The IRF510 only has 100V breakdown, which is probably much too low. It is avalanche rated so at something like 120V drain to source, it will avalanche and short drain to source. This will happen shortly after turn-off as the coil voltage tries to shoot to many hundreds of Volts, even up to 1kV, assuming about 12 VDC supply. An IRF820A / IRF830A / IRF840A would be a better candidate with 500V rating. Depending of the back EMF, 800V or 1000V FETs might have to be used. IGBT are good, but can be a little slow. Do some Inet research on turning IGBTs on and off, they are not the same as MOSFETs.
Regards, Earl
PS, you were using a FET driver for the IRF510 weren't you?
Quote from: Earl on October 14, 2007, 06:12:36 AM
Thanks for the scope tip, I happen to be one of those EEs who scopes everything.
The IRF510 only has 100V breakdown, which is probably much too low. It is avalanche rated so at something like 120V drain to source, it will avalanche and short drain to source. This will happen shortly after turn-off as the coil voltage tries to shoot to many hundreds of Volts, even up to 1kV, assuming about 12 VDC supply. An IRF820A / IRF830A / IRF840A would be a better candidate with 500V rating. Depending of the back EMF, 800V or 1000V FETs might have to be used. IGBT are good, but can be a little slow. Do some Inet research on turning IGBTs on and off, they are not the same as MOSFETs.
Regards, Earl
PS, you were using a FET driver for the IRF510 weren't you?
Hi Earl,
Yes, I was using a driver (specifically the MAX627) for the IRF510s. When I designed the output circuit so long ago, I didn't have all the information from Bob, so I made guesses. Turns out I derated the MOSFETs too far, going for switching time instead of voltage handling capability.
Bob has stated that these particular IGBTs can be driven by the UCC37322s, and furthermore that they are good up to 5MHz, so I think I am good there. ;D Looking at the datasheet confirms that the worst-case rise/fall time is 0.5uS, which is about 2MHz, but that was measured at full load. We will not be operating these things anywhere near full load, so they will switch much faster.
Eldarion
All,
It's probably time for an update here. As you know, I am working on a custom SMD circuit board for the controller while I wait for the bricks to arrive. I should have the bricks in hand by Friday, and have already received the UCC27322P driver chips.
A warning to all: When purchasing items off of Ebay with PayPal, do not use your checking account! It will take over a week to go through (the reason that I do not have the bricks already.) :o I had to call up the surplus store directly and use a credit card instead.
Hopefully things will get more interesting here on the weekend! ;)
Eldarion
Well, the IGBTs finally came! :)
One of them is shot (drawing 1.5A when "off"), but the other three checked out OK. That one will be going back to be replaced. ;)
I went ahead and hooked everything up and ran some frequency sweeps. I was using a 40W bulb for a load, and surprisingly it lit, albiet dimly, at 5000ns pulse width, drawing about 30W from my power supply. As the frequencies were swept around, it grew brighter and dimmer, although it never approached normal 40W brightness in my short test. Disconnecting my scope ground from the signal generator caused the bulb to become slightly brighter, although the signal generator should still have been grounded through the CRT cable.
Now to the reason the test was so short: I had the bulb connected directly to the secondary winding, no HV potential or anything. I wanted to see what would happen if I tied one end to Earth ground. When I did so, the bulb nearly extinguished (why, I am not sure), and the current draw shot way up on the input. Surpisingly, the bulb stayed barely lit, fluctuating a lot, and the core made terrible sounds. On the next power-up, the MOSFET drivers blew, probably from all the HF HV coupled onto every wire near the core. Unfortunately, I have not yet received my MAX627 replacements, so further testing will have to wait until I can build a board for the new UCC27322 chips or I receive the MAX627s.
As an aside, these IGBTs are amazing. Even when the MOSFET drivers failed (which would have detonated my MOSFETs), they stayed stone cold. Scoping across the primaries revealed far cleaner waveforms than before, which is probably why this test showed a bit more of the interesting stuff Bob has been mentioning.
I think I can safely assume that the reason this test was under-unity was because of the lack of the HV potential. Bob has mentioned that the system is under-unity with a potential below 11.5VDC, and since my potential was at 0VDC, these null results are easily explained.
I am not sure why grounding only one end of the bulb did such nasty things...any ideas here would be appreciated. ;)
Eldarion
Both sides of the secondary must be kept isolated from ground at all times. the only path to ground must be through the load ;-)
Bob
Quote from: Bob Boyce on October 19, 2007, 11:06:56 PM
Both sides of the secondary must be kept isolated from ground at all times. the only path to ground must be through the load ;-)
Bob
Hi Bob,
Thanks for confirming this. :) I think I may have had far greater power flow into ground than I had through the light bulb based on the sparks I saw...maybe it was overunity after all! I still have to get used to the peculiar aspects of the longitudinal energy coming out of that secondary.
I assume I can connect one end of the pancake winding to ground, as this is conventional DC TEM?
All,
Unfortunately I did not have a chance to check the longitudinal windings or anything before my driver board blew. I have finished designing the Thor 1.1 output driver board and have sent it off to the fab; I should have it in a couple of weeks. This board should be much more stable and robust than my crazy hand-soldered wiring jobs. ::) Bypass capacitors within 2mm of each MOSFET driver, oh yeah! ;D
Signing off for now,
Eldarion
Eldarion,
ringing is usually caused by
leakage inductance
leads too long
current loops too long
a toroid usually has low leakage inductance.
most people have trouble understanding that a circuit operating at 67 kHz, still needs to be built as if it was operating at hundreds of MHz, even 1 GHz, if there are pulses with fast rise and fall times. Circuits with fast edge transistions MUST USE VERY SHORT LEAD LENGTHS. By short, I mean ZERO mm. Since zero mm is not possible, do every thing in your power to approach zero as close as you can. Spare no effort.
Notice that your gate voltage has ringing that even goes negative.
Remember my image of current loops. Look at this image again and ask yourself if all 3 of your currents loops have a length close to zero?
Is the distance between the ground of the driver IC and the IGBT source = to zero?
Don't forget E = L * di/dt and since we want di/dt to approach infinity to have RE effects, L must approach zero.
Even nanoHenries of inductance can cause sufficient and undesired voltage bounces.
Why do you think PCB designs use maybe a dozen SMD ceramic caps around each microprocessor? The multilayer PCB has one layer dedicated only to being a ground plane.
These caps and CPU ground are connected to through-holes directly to the ground plane.
Post some photos in your thread of your physical layout, and I can make some more comments.
Would like to see a scope photo full screen of just the area in the attached photo.
Regards, Earl
Well, I finally got a 10:1 probe, and also got direct data capture over the GPIB bus working!
So, here is an updated scope shot of the driver waveforms on channel 1 (bottom) with a 10:1 probe, and the switched primary coil end waveform on the top (channel 2) with a 1:1 probe:
http://www.falconir.com/pics/withten_one_probe_on_channel_1.jpg
As you can see, the gate drive is not really as bad as you might have thought based on the first scope shot...
Eldarion
Well, here is a quick update:
The old hand-wired board with the MAX627s on it was severely distorting and altering the pulse sequence--I disconnected it and scoped across the FPGA's TTL outputs, and a perfect waveform was observed. I guess there was just too much interference being picked up by the horribly long wire lengths.
So, I will stop this unproductive line of testing and concentrate fully on building the new driver board. I got several of the UCC27322 SMD chips installed, along with their bypass capacitors, and some other through-hole support components. Unfortunately I am still waiting on the main 40-pin right-angle header (in the UPS system somewhere), so I will not be able to install any more parts until I get that installed. This is so that I can test my work as I go, rather than potentially having to troubleshoot an entire "finished" board at once. I do believe we should see some positive results once this board is finished; having the 4-channel digital 'scope is really aiding design and troubleshooting here! ;D
Eldarion
At the docs office on broadband WiFi so I can finally get caught up on all of the threads with pictures.
@eldarion
Glad to hear you figured out where the problem was. Breadboarding is good for really low frequency testing of parts but as you seen it can get a bit "hairy" as speed is increased.
I have my own problems I'm dealing with. I have the need... the need for speed ;-) My 20 MIPS is topping out pretty low when running all 8 channels, and still too low when cut back to 4 channels. I'm not finished with trying coding tricks.
Bob
Hi Bob,
If you like I can send you one of my unstuffed driver circuit boards and the control software--once I have checked everything out on mine, of course! ;) The FPGA can easily go to 500KHz+ on the primary frequency, and while driving 6+ channels. Whether the driver board will keep up without distortion or timing glitches is the only question...
Just let me know the pulsing sequence I should create for 6 or 8 channels, and I will update the control software!
Eldarion
Quote from: eldarion on October 30, 2007, 10:22:15 AM
Hi Bob,
If you like I can send you one of my unstuffed driver circuit boards and the control software--once I have checked everything out on mine, of course! ;) The FPGA can easily go to 500KHz+ on the primary frequency, and while driving 6+ channels. Whether the driver board will keep up without distortion or timing glitches is the only question...
Just let me know the pulsing sequence I should create for 6 or 8 channels, and I will update the control software!
Eldarion
I would love to give the FPGA approach a try for sure. The pulse sequence is like that which was pictured earlier, will attach it here. In order to maintain exacting timing relationships, I am producing pulses for all 3 channels, but directing the missing pulses to unused port pins. I am driving the B channel set and A channel set with a slight adjustable phase offset from one another. If I were to time them exactly together, I could reduce complexity a bit and increase speed. I may end up doing that, but for now I like having the option of phase shift between sets. The other 2 channels are for PMW controlling the control windings, so they are driven different. I will probably change the channel arrangement around and use a dedicated PWM channel pair for this function. As it is now, I am generating my 2 PWM channel signals totally via code instead of using the internal PWM channels where I can set and forget until it needs to be changed. I am reading my ADCs twice per code loop currently to get high priority to changes, that can also be reduced.
Bob
Forgot to mention. My pulse sequence is 12 pulses long, then repeats, makes for some pretty simple code at this point.
Just a quick update on my efforts to get a very fast switch working... I now have clean rise and fall times of 11-13 ns. I am using TC4421/22 drivers and IRF 820A mosfet. I chose the 820 because of the low gate capactiance (only 360pf) and high voltage rating 600V. The only tradeoff is 8A maximum current pulse (2.5A average) .. but this is not a problem if we want to drive fast pulses into anything with even the slightest inductance.
Earl is right -- to get clean fast switch times (both rising and falling) took quite some effort... 1uF ceramic decoupling right on the power supply pins of the driver (SMT cap) with larger cap right on it's back, schotkey diode on the high side of the mosfet with reverse ultra fast recovery diode for handling the back emf. I have everything mounted on a small heat sink as one monothlithic block with absolute minimum lead lengths. Another tip is to run off battery to completly avoid any ringing or interaction back up into the power supply or through unintentional ground loops. Yet another tip -- you can have long runs of wire from your signal generator and to your devices and preserve the wave form if you use good quality twisted pair connections -- get some twisted pair network cable and canabalize the twisted pairs from that ;)
The net result is an extreemly fast switch that runs cold at anything below 1Mhz. I have used the switch block up to 3Mhz and tested it to 11Mhz. To improve things further I am getting in some ultra fst pin drivers from Intersil (http://www.intersil.com/cda/deviceinfo/0,0,EL7158,0.html) these are rated at 40MHz continuous opperation and 8ns rise/fall times into nf's load -- so at least at the switch output I will be able to push transition times down to 4ns or so.
In tests driving primary coils on my toroidal ferrite core the fast switching makes a big difference! ... but her's one strange thing - now I can pull small sparks from both the core and other open coils (but only from one end of other open primaries)... but I cant see matching HV spikes on the scope!!!! I will report when I know more about exactly what is happening -- perhaps I should just start testing with light bulb loads :)
cheers
mark.
Way to go Mark. Seeing sparks with nothing visible on the scope, that shows you are finding exactly what you are after.
You can ask Rich In FL about this, lighting 120 VAC 75 watt bulbs (even to the point of burnout - and the arcing beyond burnout) while nothing reads on a scope, voltmeter, ammeter, or RF wattmeter.
I plan to try mounting well bypassed drivers on tiny boards bolted right to the input terminals of some 1200V/400A bricks to see how well that works ;-)
@eldarion
What size LCD screen do you use for that FPGA board? I found some nice goodies here.
http://www.futurlec.com/LCDDisp.shtml
Bob
First, good job Mark! Are you driving those primaries open-ended or closed? Also, are you using Bob's pulse sequence yet?
I have been quite busy populating the new driver board with parts. Here is a pic so far, everything installed checks out except for the UCC27322s, as I have not yet fired them up. ::) Here is a pic: http://www.falconir.com/pics/DSC02301.JPG
I am currently using a standard VGA PC monitor to display all kinds of data. I am not quite sure if I want to add a small character LCD on to the system yet, if anyone would like one in addition to the main VGA screen please speak up! And also comment on what data you would want displayed on that smaller screen vs. the main one. Also keep in mind that I will be allowing networked remote control of the complete system, so you would not have to have a monitor and keyboard hooked up all of the time.
Eldarion
Quote from: eldarion on October 31, 2007, 12:38:39 AM
I am currently using a standard VGA PC monitor to display all kinds of data. I am not quite sure if I want to add a small character LCD on to the system yet, if anyone would like one in addition to the main VGA screen please speak up! And also comment on what data you would want displayed on that smaller screen vs. the main one. Also keep in mind that I will be allowing networked remote control of the complete system, so you would not have to have a monitor and keyboard hooked up all of the time.
Eldarion
If there was enough processing power to spare to drive a character display, it would be nice to have, at least for a benchtop testing system. I have looked over the information on www.digilentinc.com and it is unclear what the Spartan 3 units are supplied with, and what accessories would need to be ordered at the same time. That was my main reason for asking about LCD character displays, as that other site had a wide selection available.
Also, there are several models, at varying prices. It appears that the FPGA are directly surface soldered to the boards, which would mean that being able to upgrade to a higher gate count FPGA chip does not seem to be an easy to implement option. So, for a Spartan 3 based system, it looks like a choice needs to be made between the 200k - 400k - or 1000k gate Spartan 3 starter board, the 500k gate Spartan 3E starter board, or the 1600k gate Spartan 3E-1500 development board. I guess I am just confused by the unfamiliarity with these products. What software is suggested for use in programming these? I would think that others here would like to know this as well.
Bob
Hi Bob,
I am using the Spartan 3 -1000 version, which does not come with a character LCD display by default. I will be happy to put one on the system, maybe display output power and frequencies or something? I do have plenty of processing power left over.
There are several software packages available for programming the Spartan FPGAs. The free one is the ISE Webpack, the more expensive option that I am using is called the Embedded Development Kit (EDK). The EDK allows me to place a small 32-bit microprocessor into the FPGA fabric, whereas the Webpack does not. (The Webpack will allow you to develop any logic you want on the Spartan 3/3E, but only has a small 16-bit "Picoblaze" processor available, and the Picoblaze is an absolute pain to work with).
Where I am going with all of this is to have a 4Kw+ rack-mount module that has some basic control functionality, but most of the configuration and set-up would take place with a keyboard and VGA monitor, then be saved to nonvoltatile memory. At that point, once the system is installed, the network connection could be used to get power output, system status, etc. I chose the standard network connection because it uses Manchester encoding, therefore small time shifts caused by the cores at full power should not cause a loss of communication with the controller. In addition, I am thinking of having several of the modules converse over the network in order to balance loading, etc. But that is far off in the future! ;)
Hope this helps some,
Eldarion
All,
I have completed parts stuffing for now. The board is complete enough for me to resume testing with--I did find a major error with regards to the ADCs, so before the system is put into production I will need to run another batch of boards. I have reenabled two of the ADCs with some hand work, but the other six will probably have to stay disabled. Other than that the boards are in good shape.
The pulse trains to the IGBT gates are now pretty much perfect, and exactly according to the diagram Bob provided.
I have tried firing the coil, but am still having no success (COP << 0.5 and no DC on the longitudinal windings). I will double-check everything and try again soon.
Eldarion
Quote from: eldarion on October 31, 2007, 09:38:34 PM
All,
I have completed parts stuffing for now. The board is complete enough for me to resume testing with--I did find a major error with regards to the ADCs, so before the system is put into production I will need to run another batch of boards. I have reenabled two of the ADCs with some hand work, but the other six will probably have to stay disabled. Other than that the boards are in good shape.
The pulse trains to the IGBT gates are now pretty much perfect, and exactly according to the diagram Bob provided.
I have tried firing the coil, but am still having no success (COP << 0.5 and no DC on the longitudinal windings). I will double-check everything and try again soon.
Eldarion
Yea I hate when that happens. I've had my share of board mistakes. This is one major reason why I want to thoroughly test a board layout prior to releasing it.
So I should consider buying one of those Spartan 3 starter kits with the -1000 gate option? When I was reading some material about the 32 bit microprocessor, it looked like it took 900k gates to emulate the processor. This was real old material, does it still tie up that much resources for the newer code? Also, would I have to get the EDK to be able to load already written code? This shows how little I know about complex FPGA systems. I hate to have to spend more for development tools than the hardware costs.Cost of development tools was one of the primary reasons I chose the Atmel AVR for the hydroxy gas stuff over some of the better choices available from other mfrs. Also, which character LCD would be a good match for that Spartan 3?. I was thinking about ordering one of these 20 X 4 with backlight shown here
http://www.futurlec.com/LED/LCD20X4BL.shtml
I'm wondering if the Spartan 3E-1600 might not be a better choice for this and future experimentation, despite the much higher cost. Being so unfamiliar with the FPGA has me undecided in which direction to consider going. I like learning, but I also hate making a first purchase only to regret not having spent a bit more for a much more usable device. Oh, is the Spartan 3 board small enough to fit within the hole in the center of the core if stood on end?
Bob
Quote from: Bob Boyce on November 01, 2007, 02:51:34 PM
Quote from: eldarion on October 31, 2007, 09:38:34 PM
All,
I have completed parts stuffing for now. The board is complete enough for me to resume testing with--I did find a major error with regards to the ADCs, so before the system is put into production I will need to run another batch of boards. I have reenabled two of the ADCs with some hand work, but the other six will probably have to stay disabled. Other than that the boards are in good shape.
The pulse trains to the IGBT gates are now pretty much perfect, and exactly according to the diagram Bob provided.
I have tried firing the coil, but am still having no success (COP << 0.5 and no DC on the longitudinal windings). I will double-check everything and try again soon.
Eldarion
Yea I hate when that happens. I've had my share of board mistakes. This is one major reason why I want to thoroughly test a board layout prior to releasing it.
So I should consider buying one of those Spartan 3 starter kits with the -1000 gate option? When I was reading some material about the 32 bit microprocessor, it looked like it took 900k gates to emulate the processor. This was real old material, does it still tie up that much resources for the newer code? Also, would I have to get the EDK to be able to load already written code? This shows how little I know about complex FPGA systems. I hate to have to spend more for development tools than the hardware costs.Cost of development tools was one of the primary reasons I chose the Atmel AVR for the hydroxy gas stuff over some of the better choices available from other mfrs. Also, which character LCD would be a good match for that Spartan 3?. I was thinking about ordering one of these 20 X 4 with backlight shown here
http://www.futurlec.com/LED/LCD20X4BL.shtml
I'm wondering if the Spartan 3E-1600 might not be a better choice for this and future experimentation, despite the much higher cost. Being so unfamiliar with the FPGA has me undecided in which direction to consider going. I like learning, but I also hate making a first purchase only to regret not having spent a bit more for a much more usable device. Oh, is the Spartan 3 board small enough to fit within the hole in the center of the core if stood on end?
Bob
Yeah, go ahead and get one of the Spartan -1000 starter kits. That FPGA actually has 1 million logic gates on it! (Fully loaded with a MicroBlaze soft processor, nine ADC interfaces and 4 three-phase precision pulse generators, as well as the VGA display system and PS2 core, only 70% of the chip is used.)
Yes, you can load a pregenerated EDK file onto your board. All I would do is send you a bitstream file, and you could load that straight into Impact (the downloader/programmer software tool) and program the board.
Any LCD will work; I just have to write the hardware interface. The one you found looks good.
Before I put too much more work into the "extraneous" stuff, I would like to see this thing generate some excess power! ;) I hooked up some power resistors instead of the primary coils and scoped across the IGBT drains. These things are switching almost perfectly, and the pattern is exact. Yet I am still getting tiny amounts of power out?!? ??? Not sure what the problem is.
Oh, and I put the varistors across the primaries like you did on the PWM3E--they get very hot, so I still seem to have a terrible ringing problem. I still do not understand why your PWM3E seems to switch the primaries with almost no ringing and why I am having such a terrible time with a very similar output drive system. Any ideas? ;D I do not have any snubber diodes connected as you do in your PWM3E, maybe I should add them to the circuit? I don't see where that would help, however, as my ringing waveform usually does not go negative.
I even tried the HV potential, cranking it up to 160VDC, and still nothing! I am not sure where to proceed from here.
Oh, all tests were performed with a 7W 120VAC light bulb connected across the output terminals, for some tests (without HVDC potential) I just had the bulb connected across the secondary output leads, and for other tests (with the HVC potential) I had it connected through the HF choke and DC blocking capacitor in their appropriate locations in the output circuit. It lit very dimly in all cases, and as the frequencies were swept, it would go out completely sometimes.
Any ideas? ;D
Thanks,
Eldarion
Quote from: eldarion on October 31, 2007, 12:38:39 AM
First, good job Mark! Are you driving those primaries open-ended or closed? Also, are you using Bob's pulse sequence yet?
I am driving closed loop primaries on my ferrite core. This is still just prelimenary testing -- it has taken a while (and 6 blown drivers!) to get the switch working - both fast and reliable. I can now push transition times to just below 10ns measured on the coil :)
I am working on a design for a monostable switch that will deliver a minimum width pulse. The idea is to get the pulse width so short that it fits within the coil -- traveling along as a pulse down a transmission line. After witching I will then close the loop so the pulse can circulate around the coil IN ONE DIRECTION. This requires a very novel mosfet switch be closed across the coil once the pulse is injected -- the switch is a new configuration of oposing mosfets that form on AC switch with almost no voltage drop at all!
I am also plan to move up to some faster drivers:
Intersil drivers http://www.intersil.com/products/pt/parametric_table_503.asp *** fastest driver you can get and a resonable price - $4.60 each.
EL7518 8ns rise and fal time into 1000pf, 12A peak current, 40MHz continuous opperation.
http://www.intersil.com/cda/deviceinfo/0,0,EL7158,0.html
I am posting parts, data and supplier information in a public forum here http://www.overunity.com/index.php?topic=3544.msg56820#msg56820
If money is no object then the fastest mosfet gate driver soon to be availble has under 2.5ns rise and fall times with 15A current peak. It can create 8ns wide pulses. It is the DEIC515 gate driver from IXYS RF (was Directed Energy) for $21. http://www.ixysrf.com/pdf/driver_ics/deic515.pdf
Currently shipping is the DEIC420 with a 3ns rise time and 20A peak current http://www.ixysrf.com/products/mosfet_driver_ics.html
IXYS have some great application notes on their web site. Here is a link to the PDF for their low side gate driver evaluation board -- it's worth examining for general tips on how to drive mosfets fast http://ixysrf.com/pdf/switch_mode/appnotes/evic420.pdf
mark.
PS. The HV happens when I dont shunt the back EMF back into the supply rail. So the HV is expected - but there are still a few mysteries one of which is... Why I cant see it on the scope-- I am wondering if I am getting a pulse forming/sharpening action as the pulse travels around the core -- the result being that the pulse sharpens up and is simply too fast to see on the scope. This would be happening due to the limitation of the ferrite to react fast enough so the pulse piles up... but why can I draw the same sparks from the core? capacitive coupling from the windings to the core??
All,
Just to let you know, I have found yet another large instability in my pulse generator, so the above null results are probably incorrect. The instability is quite the difficult bug to smash! ::)
Eldarion
Hi Mark,
To get a single pulse to fit into a cable means that you need so have a pulse shorter than the VoP (velocity of propogation) of the coil.
According to Wikipedia: http://en.wikipedia.org/wiki/Velocity_of_propagation
For example a cable that has a VoP of 70% of speed of light means that the signal will travel 1 meter in 4.7ns
The only thing is does coiling up the cable change the VoP of the wire dramatically?
Certainly a wire of about 3-4 meters long should be enough to allow a pulse of about 10ns to fit inside the coil.
As you know IXYS have a range of mosfets and drivers that can handle these narrow pulse widths:
http://www.ixysrf.com/app/high_voltage.html but at a price though.
Could get expensive if things go bang during testing. ;)
In my short experience I think the mosfets are OK to overload with current, its just the drivers that you have to be careful with. I suppose you could use some very low value resistors to limit current. Perhaps a carbon brush could be used.
With the idea of allowing the pulse to travel around the coil continuously you could use a schottky diode in series with the start and end of the coil wires. Can a schottky diode handle the switching speeds though?
Regards
Rob
@Meger Man
I dont wan to divert the thread here so just a quick note. Yes -- a coil as a tranmission line can have dramatically lower transmission speeds. If we have interactin with a coer -- aluminum even then the transmission velocity caould be dramatically lower. This is expecially releveant if there is sone phonon-electron spin interaction in the core that is retarding the wave.
Quote from: eldarion on November 02, 2007, 03:22:33 PM
All,
Just to let you know, I have found yet another large instability in my pulse generator, so the above null results are probably incorrect. The instability is quite the difficult bug to smash! ::)
Eldarion
What sort of instability are you seeing eldarion?
Hey, I just won high bid on a pair of Tektronix 2445A scopes on ebay for cheap
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=230185659755
and
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=230185659791
I figure if it comes down to it, it's better to fry a cheap scope than the one I paid a lot for new. Besides, can't hurt to have spares that are the same as my main scope ;-)
Bob
Hi Bob,
I was seeing an odd, repetitive pulse ordering error, whereby the correct pulse sequence would be generated for a little while (maybe a few milliseconds) and then the position of the F1 pulse would shift to a completely wrong location in the waveform. I had not noticed it until recently because I had never expanded the timebase of the scope beyond a couple of cycles!
I have since completely rewritten the pulse generation core from scratch, and that bug seems to be gone. The new core checks out reliably over quite a number of cycles (20+), and the frequency jitter I was also seeing before is now completely gone! ;D
Ahh, the joys of digital logic design. ::)
Eldarion
EDIT: Here is a scope shot of the MOSFET driver outputs with the new core. Notice the lack of jitter!
http://www.falconir.com/pics/mosfet_driver_chip_outputs_fixed.jpg
Quote from: eldarion on November 03, 2007, 02:49:31 AM
Hi Bob,
I was seeing an odd, repetitive pulse ordering error, whereby the correct pulse sequence would be generated for a little while (maybe a few milliseconds) and then the position of the F1 pulse would shift to a completely wrong location in the waveform. I had not noticed it until recently because I had never expanded the timebase of the scope beyond a couple of cycles!
I have since completely rewritten the pulse generation core from scratch, and that bug seems to be gone. The new core checks out reliably over quite a number of cycles (20+), and the frequency jitter I was also seeing before is now completely gone! ;D
Ahh, the joys of digital logic design. ::)
Eldarion
EDIT: Here is a scope shot of the MOSFET driver outputs with the new core. Notice the lack of jitter!
http://www.falconir.com/pics/mosfet_driver_chip_outputs_fixed.jpg
Eldarion, that pulse pattern does look to be right, if the channel orders are correct that is. Hard to tell without channel references ;-)
Now when you go to test, please try to keep in mind the directions of energy flows, both within, and through the circuits. Remember the two drawings that showed the windings? Where in the first one, the capacitor shorts out a very large portion of the captured LEM energy to ground, while the second one does not lose as much. Keep in mind the reference points and potential differences.
http://www.overunity.com/index.php?action=dlattach;topic=2865.0;attach=11978;image
http://www.overunity.com/index.php?action=dlattach;topic=2865.0;attach=11980;image
Rich in FL has also had success with another load configuration, where the secondary is referenced to HV on one side and the bulb load is directly across the secondary with the DC blocking cap in series with the bulb. He said it takes a wider pulse width to get things going, but if you increase the primary pulse voltages, you can reduce pulse width. It is with this configuration that he was lighting, and recently burned out, a 75 watt 120 volt bulb, and the bulb continued to arc internally. Remember, he can only run in pulsed mode, as he has no controller capable of phased rotational drive.
Bob
Hi Bob,
Sorry I took so long to get back to you; one of my primary servers crashed Friday night and I spent all of Saturday repairing the hardware and reinstalling the OS. :(
The channels are:
Channel 1 (top trace): F3 (lowest frequency)
Channel 2 (middle trace): F2
Channel 3 (bottom trace): F1 (highest frequency)
So for Rich's setup, is this the schematic?:
http://www.falconir.com/pics/secondary_connection_style2.jpg
What size DC blocking capacitor should I be using? Right now I am using a 0.68uF 160V polyester cap, which at an F3 of 10KHz works out to 23.4 ohms of reactance, and at the corresponding F1 of 40KHz it works out to 5.9 ohms of reactance.
In case I haven't already mentioned these specs, here they are:
IGBT source voltage: 13.8VDC
HVDC potential voltage: 160VPDC, pulsed at 60Hz
No magnetic bias is connected
Secondary of transformer in series with 160V HVDC supply for HF filtering
7W 120V light bulb used for testing purposes
I tried firing up the system with all of the above specs, and preliminary results are still null; COP < 1 though higher than before (lights bulb to full brightness on 11.04W of input power, HVDC still has no effect). Any thoughts?
Thanks!
Eldarion
Quote from: eldarion on November 04, 2007, 04:29:08 PM
In case I haven't already mentioned these specs, here they are:
IGBT source voltage: 13.8VDC
HVDC potential voltage: 160VPDC, pulsed at 60Hz
No magnetic bias is connected
Secondary of transformer in series with 160V HVDC supply for HF filtering
7W 120V light bulb used for testing purposes
I tried firing up the system with all of the above specs, and preliminary results are still null; COP < 1 though higher than before (lights bulb to full brightness on 11.04W of input power, HVDC still has no effect). Any thoughts?
Thanks!
Eldarion
Hi Eldarion,
A quick question my friend. Why "no magnetic bias"? ??? I do not think OU is possible without that also hooked up, from what Bob has said. Bob, please correct me if incorrect.
Quote about the magnetic bias from the .PDF: (For everyones learning)
"The one that has a separate DC bias winding can introduce a much greater angular field, but it can be quite the beast to control. This one is like a hurricane that can turn into a tornado at a moments notice. This is one that can initiate intense lightning strikes and other nasty stuff if not kept on a VERY tight leash. Obviously, this one is not shared with the hydroxy gas crowd.
The riskier and higher performance toroidal power system uses
both the DC potential bias and a magnetic bias."
AND
"Vortex mode, which is almost identical to rotational mode. The addition of a longitudinal bias winding and a high enough DC bias can result in the analogy of a "severe hurricane", or a "tornado". This is the highest energy mode, which is very unstable, and can be extremely difficult to maintain control of.
Applied bias can play a major role in the behavior and severity of the "storm".
Oh, by the way, that longitudinal low voltage bias winding is for adding the magnetic bias, not a low voltage DC potential bias,
so it must be a fully closed winding fed with straight DC, not pulses. That longitudinal winding is to be wound in the direction of rotation, from top to bottom. The higher voltage DC potential bias needs to be coupled onto the secondary, and decoupled from the load for most applications. If longitudinal DC is required to run a load, then you can add another longitudinal winding or two to the top and/or bottom of the core before wrapping the secondary. Be sure to follow the direction of rotation or it will not work properly."
Warm regards,
Bruce
Quote from: btentzer on November 04, 2007, 04:57:57 PM
Quote from: eldarion on November 04, 2007, 04:29:08 PM
In case I haven't already mentioned these specs, here they are:
IGBT source voltage: 13.8VDC
HVDC potential voltage: 160VPDC, pulsed at 60Hz
No magnetic bias is connected
Secondary of transformer in series with 160V HVDC supply for HF filtering
7W 120V light bulb used for testing purposes
I tried firing up the system with all of the above specs, and preliminary results are still null; COP < 1 though higher than before (lights bulb to full brightness on 11.04W of input power, HVDC still has no effect). Any thoughts?
Thanks!
Eldarion
Hi Eldarion,
A quick question my friend. Why "no magnetic bias"? ???
Hi Bruce,
I was under the impression that the magnetic bias windings were there to shut down an unstable vortex if it were to form, and that therefore having a constant magnetic bias would lessen the generated power. Now going back and re-reading the PDF; you may be correct and I may be incorrect, but I would like to hear Bob's take on it before I test the converter again. ;)
Hmmm...Bob, can you verify that the required power into the magnetic bias coil is only a few hundred milliamps?
EDIT: I have been thinking about the sharp, narrow pulse requirement. Are we trying to approximate a Dirac delta function here in an effort to inject an infinite number of frequencies into that primary coil? If so, how does this work with the rotational mode?
EDIT2: I have also been looking over your PWM3E circuit, and I was wondering why you included R4 and R7 on the gate of Q1? It almost looks like there will be an asymmetry in turn off / turn on times, with the device turning on slowly and turning off quickly, although not nearly as fast as if you had just coupled Q1's gate directly to the PCP116 optocoupler. Maybe this slowing down of the switching is eliminating any possible ringing problems, like the ones that I am having in my controller?
Looks like this is fixed in the PWM3F system--sorry!
EDIT3: I promise, this is the last edit! ;D I decided to run some simulations to test an idea I had, and the results seem to strongly back it up: the load impedance is
critical! Too high, and you will get terrible primary ringing. Here are two simulations, one with the rough equivalent impedance of a 7W 120V light bulb (in reality the impedance is even higher), and one with a 100-ohm load (about 150W @ 120V). Notice the absence of ringing with a low output resistance!
High output resistance waveforms and test circuit: http://www.falconir.com/pics/high_load_resistance.jpg
Low output resistance: http://www.falconir.com/pics/low_load_resistance.jpg
Eldarion
After running a ton of simulations and reading up on the operation of flyback converters, etc. I have figured out that the ringing on the primaries (or lack thereof!) is highly dependent on the winding inductances, coupling between primary and secondary, and the load impedance. Mess up just one of those values, and terrible ringing will result, along with almost zero power transfer to the load (exactly what I am experiencing on the test bench). Coupling must be as close to 1 as possible. Interestingly, the ringing is not dependent on frequency or pulse width--this is also a very good thing as far as control devices are concerned.
What I am going to do is obtain an inductance meter so that I can not only measure primary/secondary winding inductances and choose a proper load impedance to reduce ringing, but so that I can also measure the transformer's K. My simulations show that the primary ringing can be completely eliminated if the load impedance is chosen properly and the K is high enough, > 0.95.
This is probably why the switching power supply Bob had talked about as the origin of this effect only went into overunity operation under "certain load conditions". With a specific loading impedance, the effect-destroying ringing would disappear, and if the controller then delivered the correct pulse sequence for an instant... Also I would surmise that the K would be quite high in a mil-spec high-efficiency switcher.
It may be quite a challenge to create a controller that will maintain a specific load impedance into the coil under highly variable output loading. Something to think about, anyway! ;)
Pushing forward...
Eldarion
EDIT: I have now run some tests on the coil, and can confirm that when driving only three of the six primaries the leakage inductance on each pimary is quite high--high enough that I cannot even come close to eliminating the ringing in that configuration. Also, as a result, the K is probably quite low, which is probably why the best efficiencies I have seen are only in the 40-50% range. A "normally" wound (not sector wound, as this one is) toroid driven that same way should have an efficiency of 95% or better!
I wonder if the primary coils' leakage is supposed to be that high or not? We deviated from the original hydroxy design when we wound 6 primaries instead of three...
Hi Eldarion,
That is a great bit of electronic detective work, and should help out the guys with their build.
I wonder if the "magnetic bias" will effect the ringing problem.
The magnetic bias is a quarter of the four main components of Bob's device. And without that in the mix, no other solutions can be established. For if you do, they will each have to be re examined when you add the non pulsed dc magnetic bias to that longitudinal (solenoid) winding.
The four main components, of which, if any are missing, is an incomplete, non working device:
1. Properly wound coil
2. Magnetic bias of non pulsed 13.5 vdc minimum for over unity to occur.
3. Secondary HV dc potential of 160 vdc minimum
4. Primaries pulsed with the proper frequencies, and timing pattern with clear waveforms.
I realize this is simplified, but not to "harp", 25% of the device is "missing" without the magnetic bias hooked up. That will effect everything in ways the simulator can not see. ;D
Warm regards,
Bruce
Hi Bruce,
I have actually been hooking the magnetic bias up now and then to see if it would have any effect. So far, every time I connect it, nothing at all happens to either the input or output waveforms. Same with the HV DC potential. I wonder if the potential fields will only have an effect after all the other criteria are met and the energy conversion process starts?
The most pressing concern right now is that high leakage inductance--unless we can get that fixed, the ringing will not disappear and the effect will never manifest. :(
I'll see if I can hook primaries 1 and 2 in series, also 3-4 and 5-6, that way three primaries will be emulated and maybe the coupling will increase.
Eldarion
EDIT: No-go on the three-primary emulation. The leakage is just too high.
Not sure where to go with this, short of re-winding the primaries...even that may not help, as I am not quite sure why there is such high leakage and low coupling.
How many layers of tape are separating the primary and secondary windings? The closer those windings are, the better the coupling. (This is why many SMPS cores are wound multifilar). Right now, we are wasting most of the input power as stray magnetic fields that are not coupling to the secondary or even the core! :o
Unless this is what is supposed to happen, in order to couple into the longitudinal energy flow. Bob, would you mind jumping in here and preventing me from going somewhere I shouldn't? ;D
Thanks!
Eldarion
Hi Eldarion,
I have one layer of HV Electrical tape, followed by one layer of Motor tape. This is over the Secondary, under the primaries. This was done after Bob B. had mentioned this in a phone conference. It was to make for a "flatter" winding of the primaries, if I recall the comment.
hummm........
That might be a part of the problem. I do not know. I think Bob has run 6 primaries before, thus the "Hex" controller, so I would not think that was the problem.
So, yes, some help from Bob here, is needed. ;)
Warm regards,
Bruce
Maybe there's hope yet:
I was running some tests with two adjacent primaries in series and the secondary hooked up to a potentiometer. This configuration damped out the oscillations very effectively when the potentiometer was set to 600 ohms; no additional power draw was noticed when the RCD snubber was inserted on the primary side.
Schematic:
http://www.falconir.com/pics/single_primary_no_oscillation_test.jpg
Scope shot:
http://www.falconir.com/pics/single_primary_no_oscillation_scope.jpg
I was able to confirm the dependence on load impedance--vary that output resistor just a little from 600 ohms, and the oscillations come back full force. Also confirmed is that pulse width and frequency can be varied without reintroducing the oscillations for a given load impedance.
Now we will see if this will scale well to the rest of the primaries or not...
Also, I wonder why this type of circuitry is not seen on Bob's PWM3F? I can understand that the cell stack would present a nice relatively constant impedance and therefore avert the ringing, but there would still be a large inductive spike without the RCD snubber like what I have across my primary coil here.
Onwards...
Eldarion
Quote from: btentzer on November 06, 2007, 04:02:50 PM
Hi Eldarion,
The four main components, of which, if any are missing, is an incomplete, non working device:
1. Properly wound coil
2. Magnetic bias of non pulsed 13.5 vdc minimum for over unity to occur.
...Warm regards,
Bruce
Bruce, Eldarion,
Series resistance is needed with the Magnetic Bias Coil(s) to limit Amps to 150mA. I tested two 10 Ohm 1/4 W in series with the 1.2 Ohm outer winding for 21.2 Ohms TL and found 3.25V yielded the max 150mA and no heat anywhere with just this level of DC. Bob will correct me if I am mistaken. However, what about the inside winding? Shoudn't we be driving this with the same drive as the outer winding?.. or is this already being tried? Both polarities tried? I apologize for piping up instead of being up by now, but I have asked about this spec before and did not hear back.
Ward
Hi Eldarion,
It's great to have someone actually testing stuff :) .. but a bummer that things are so frustrating for you.
This is just my 2 cents worth... I would be looking to run at resonance -- the core resonance, not lumped LC resonance. The ringing is not wasting any energy -- on the contrary it's because the energy is staying in the core that you get ringing.
Let me give you a parallel example -- in good Tesla coil design you aim for a coipling coeficent of 0.05 or so. This lets the primary and secondary circuits ring up with quire high Q's which is essential to opperation. You still get the same energy transfer in the end -- it's just that a lot of energy is buit up in the resonnant circuits as well.
Megnetic biasing I agree with -- but I have no idea how much... to be honest I would drive the magnetic bias resonantly to save energy. I know it's not a DC bias then -- it's an AC bias, but there are arguments for that working well also... can't say more.
Also -- corect me if I missed it, but I didn;t see that you were biasing up the primaries much ?? ... this was one of the predictions I made that Bob confirmed. He said that a *minimum* of 50V DC bias is required before the device reaches break even. After that the more bias applied the more power generated.
I'll hook up my core now and see if I can find it's fundamental resoance quickly.
cheers
mark.
OK. I cant find any magnetoacoustic resonances in this core. -- not like I can in the ferrite core.
I can hear a sharp resonances at 9.2 KHz and 11.93 Khz. Mechanical damping confirms these are comming from the core. Spacial beats also indicate these are multipole resonances. I cant detect *any* acompanying electrical signature of these resonances. I probed poloidally and toroidally -- nothing.
The resonances require some magnetic bias -- but once started they persist when the bias has been removed.
That's all I can tell you about this core. Even although I could only hear them it is clear the Q is quite high. This suprised me as the Iorn Powder is quite soft and I had feared that it would damp out all resonances. As the resonances are mechanical and with quite high Q you will be able to hit harmonics of these -- a lot of harmonics.
Hope that helps a bit.
Mark.
Hi Mark,
Thanks for all the info--I was wondering why I only had the audible resonance, glad to know it isn't just my setup! ;)
Regarding why I wasn't resonating the primary coils, here is a post from Bob. I may be interpreting it incorrectly, does he mean that the resulting waveform in the primary should look like a unipolar negative spike, or should I be injecting the spike and then allowing the coil to ring?
QuoteHello Gary.
You are correct, if you drive any single winding on that toroid with an AC waveform, you will not see anything but single phase AC output on all the other windings. Where this phase factor comes into play is, instead of driving with AC, you drive each of the 120 degree coils in phase with very sharp and narrow unipolar negative DC pulses. The energies that are in phase will be additive. Now apply a nominal 160 VDC positive bias (through a high value resistor or LPF ) at one end of the single winding that is wound 360 degrees. At the other end of that winding, use a DC blocking cap, and apply that signal to a load (120V light bulb works well). Now as you are applying pulses and bias, start to shift phase very so slightly of phase B and phase C and observe the effects. You will not need to adjust but a fraction of a degree. If you do not see the effects, then your pulses may not be of sharp enough rise/fall times, or there may be too much of an impedance mismatch between the drive and the transformer. Oh, all applied potentials are in reference to earth ground. You may want to compare the results with and without the 160 VDC bias, reversing bias and applied drive potentials, ect.. and note the observed differences.
The other mode is rotational with 120 degrees of phase differential between A to B and B to C, then vary B and C a fraction from that. But I highly suggest you avoid that until you have a control loop set up that can allow you to shut it down very quickly. It can go out of control suddenly and avalanche if conditions happen to be just right.
Bob
Any thoughts? This pretty much sums up what I have been doing; incidentally this post seems to succinctly and clearly explain how to at least get
some energy conversion effects--maybe not at full power to start with, but the working device can always be tweaked later on.
Eldarion
EDIT: Here's another quote. This one seems to confirm that resonance is not desired in the primaries, and that any current whatsoever does not help the effect. I do wonder if the primary drive should be pure electric fields, or if some magnetic component is required?
QuoteThere are many ways to tap into this LEM pie. Almost all share a common thread, ultra fast pulses of pure electric potential.
EDIT2: Here is another interesting tidbit. I do not understand the source of my difficulties; this seems easy?
QuoteRotating fields in and of themselves are not dangerous, as electric motors do it all the time. It is rotating fields inside of electrostatic potential fields that can become dangerous. This does not mean that it will always occur, just that there is a great risk of this occuring if things are just right. If it does, there is no advance warning. Without controls in place, it will quickly escalate into a runaway condition that will culminate in an energy avalanche. It is this energy avalanche that will trigger a local lightning strike.
Didn't Grumpy experiment with something like this with the MAGVID whereby he inexplicably burned out one of the stepper motor controllers that he was driving the orthogonal coils with? But I think he pulsed the magnetic bias in some odd way when the system burned out...
I am so fed up with this not working; if anyone has any ideas I would be happy to try them! (No offense to Bob, it is not his fault--I am probably just missing something that others are taking for granted. ;D) My next step is going to be to make sure that I am not swamping the output, as Bob has mentioned before, but after that I'm out of ideas.
I have noticed that the pulse fall time is excellent, but the rise time is rather long because the switched primary end is not being actively driven to +13.8V at that point. Just another thing to consider...
Eldarion -- I just looked back over the thread - no screen shots of signals. Are you able to post screen shots of waveforms -- measured at the coils? It would help to see whats (not) happening.
mark.
Sure! Here you go. I had not posted them because I thought they were too chaotic to be of much help.
Here is the primary coil F3 waveform at 10.0KHz:
http://www.falconir.com/pics/F3_highest.jpg
Here is the primary coil F2 waveform at 20.0KHz:
http://www.falconir.com/pics/F2_middle.jpg
And here is the primary coil F1 waveform at 40.0KHz:
http://www.falconir.com/pics/F1_lowest.jpg
Here are all the signals on one scope screen:
http://www.falconir.com/pics/all_together.jpg
And here are the gate drive waveforms, scoped at the IGBT gates:
http://www.falconir.com/pics/gate_drive.jpg
All of these measurements were taken with the identical settings; 500nS pulse width, 40.0KHz primary F1 frequency, 1-2-4 frequency division, 120.0 P2 phase, 240.0 P3 phase, capacitor and 7W light bulb for load.
Thanks!
Eldarion
EDIT: Fixed some labels ;)
Eldarion,
You scope shots look like positive pulses rather than the negative pulses that Bob specified. My understanding is that you drive the primaries with negative pulses. I can't see why you would want the primaries to ring, so they should be damped.
Ar your primaries open-ended? Sorry, I know that is somewhere in this thread and I am being lazy and not looking.
Hi Grumpy,
No, I am driving the primaries closed-ended, as Bob does. The IGBTs are in a pull-down configuration, so they are supposed to generate negative pulses, but the ringing is just so bad that the primary waveforms look nothing like they should.
Damping the ringing just plain doesn't work. As soon as one of the other primary coils pulses, it sets up ringing of a different frequency in the non-driven primaries. Seems that the only way to get rid of the ringing completely would be to tie the switched end of the coil to +13.8V, which is the equivalent of shorting out the coil.
If I physically disconnect two of the primaries, and drive only one, I can get the waveform to look pretty much perfect. As soon as I connect even one of the other primaries, however, everything breaks loose and I get the waveforms that I posted above.
I also do not like how long the pulse actually is at the primary coils. When the switched primary end is released, it is taking an awfully long time to return to +13.8V...
Eldarion
Sounds like the primaries are in radiating tranversely - conventionally - and you need then radiating longitudinally. I think Bob mentioned this very early on. Mark Snoswell explored this to a great extent. Very interesting results.
See, you have both types of wave going on and you can favor one or the other. They can even be separated, the longitudinal and transverse components that is.
Short fix: narrower pulses. Passive differentiation with C and R. Hell, double differentiate it. Pulse will get shorter and amplitude may go down to some degree - at least on an oscope - the spectrum will likely get very active as all the harmonics start showing up.
Small R and Small C - like a few ohms and a few pF
http://www.st-andrews.ac.uk/~www_pa/Scots_Guide/experiment/diff/diff.html
EDIT: also might try choking the inputs to the primaries. Potential required, not current.
Hi Grumpy,
Thanks for clearing that up! I had, early on, been confused by all the discussion about longitudinal resonance--longitudinal radiation makes a whole lot more sense. Primary resonance does not seem to be something that we want! :D
I will try double-differentiating the output of the MOSFET drivers themselves, with no IGBTs in between, and see if something interesting happens. And yes, the HV potential will be connected. ;)
The only thing I could see as a problem is that there will be two pulses for every input pulse--one positive pulse on the low-to-high transition, and one negative pulse on the high-to-low transition. Wouldn't this be a bad thing? I wonder if a sawtooth input to the differentiator would work better.
I can't test right now, as I am away from my bench, but for sure I will post some results, positive or negative, later on tonight.
Eldarion
Get rid of it with a diode, which is not perfect, after the differentiator. You will have to have enough current to drive the diode at the diode.
You do not want undershoot (overshoot in your case since pulse in negative).
http://www.ortec-online.com/electronics/amp/03_4.htm
adds a pot across the cap - can also use a fixed resistor
OK, thanks! Everything seems OK to me now, we'll see how it does on the bench! :)
Eldarion
I spent *way* too much time last night (most of the night) doing spice simulations -- trying to find the best way to drive the primaries closed loop. The conclusion is that closed loop you really want a switch on each end of the line. Nothing else really works. The sequence target is:
1. Close switches at both ends
2. As soon as current rise starts to slow shut of both switches -- at about 2/3 applied potential.
3. When voltage reaches limit of switches open negative end.
You can think of this as opening both ends to let electrons rush in and build up momentum -- as soon as they do you shut off the ends and let the pressure (voltage) rise to the maximum -- then you open the negative end. The result of this is that you get the negative flyback wavefront propagating in a reverse direction back down the line -- this is as close as you can get to the original effect Tesla observed when they threw the switch on a long DC transmission line.
Driving an open coil is far simpler -- but you can only drive with the applied potential if your rise time is lower than the time it takes the longitudinal shockwave to travel down the coil. If you can?t switch this fast then all you will get is a tame rise of voltage to the applied potential. If your Transition time is *very* fast you should be able to get a longitudinal shockwave travelling down the line - but only at the magnitude of the applied voltage. High current IGBT's will switch the current about as fast as you can -- but they won?t switch the voltage fast enough. In this open ended mode you could use the Hyperfast lower current IGBT I listed on the Materials Parts and Data page.
After last night?s simulation work I am now leaning towards the more complicated two switch sequence and deliberately utilizing the electrodynamic momentum of the B field to drive a large reverse direction longitudinal wave front. I am also wondering if this is what was done in the targe TPU's and if the toroids in the middle were for floating the control for the switched on the ends of the lines.
No guarantees. None of this is a sure fire way of making it work ? My opinion is that until the switching is right there is little chance of seeing anything interesting. But even with the switching scheme I propose here I don?t know what will happen.
I am still testing switches, drivers and simulations. I am getting some great switch times and voltage spikes ? and blowing a lot of drivers at the same time ïÂÅ' This is mostly on the ferrite toroidal core I have. As I mentioned in a previous post I can?t trigger any of the longitudinal magnetoacoustic models in the Iorn Powder core. .. I can hear a ringing from the core but I have no idea what that is and if it?s any use.
I have been talking to Graham Gunderson about his experiences with drivers. He went through the same grief I am having until he came up with a robust and very fast modular driver board that he now uses for everything. He is sending me circuit diagram and made up boards and I will roll that information into a reference design that will be available for everyone to use.
Cheers
Mark.
yup...
Quote from: Grumpy on November 08, 2007, 05:58:30 PM
Keep in mind that the flow of current is retarted in iron or other magnetic wires.
Another intersting approach is an old Tesla coil diagram that used chokes on each side - see attached.
(Someone was serious about blocking current and again you will see the retardation of the current flowing.)
Side note: there is probably wa way to replace the primary coils with something more like a plate since we are not using the coil as an inductive coil.
LOL -- excellent timing :) ... I tried all sorts of chokes on either side in the simulation tests -- it's the first thing I thought of. After fiddling around for ages I decided that with modern fast switches it's easier to just time switches at both ends. It's a moderatly complicated switching scheme -- but once set up you can drive all sorts of different inductance lines and coils. With the chokes at each end everything has to be tuned just right.
On plates rather then primary coils -- I have a set of HF ferrite toroids commming in to do just that - set up a transmition line that should only pass longitudinal waves and is designed as a resonator for them. It will be a set of spiral coils that also act as capictor plates interleaved with toroid ferrite. The idea is the enhance the radial amgneti field while only permitting longitudinal electric field. THis is just a research tool -- not a practical device -- well that's the intent.
mark.
Quote from: MarkSnoswell on November 08, 2007, 05:23:41 PM
1. Close switches at both ends
2. As soon as current rise starts to slow shut of both switches -- at about 2/3 applied potential.
3. When voltage reaches limit of switches open negative end.
You can think of this as opening both ends to let electrons rush in and build up momentum -- as soon as they do you shut off the ends and let the pressure (voltage) rise to the maximum -- then you open the negative end. The result of this is that you get the negative flyback wavefront propagating in a reverse direction back down the line -- this is as close as you can get to the original effect Tesla observed when they threw the switch on a long DC transmission line.
Hi Mark,
That is a vey interesting set of observations--I, too, had come to similar conclusions, but had not had the inclination to build such a complex switching arrangement. I think I just might now, though! ;)
I assume you meant to:
1. Close both switches
2. Open both switches
3.
Close the negative-side switch?
Thanks!
Eldarion
Quote from: eldarion on November 08, 2007, 07:41:23 PM
I assume you meant to:
1. Close both switches
2. Open both switches
3. Close the negative-side switch?
er -- yea. I think. When I get it all working in the simulator I'll send you the spice model if you like. ... and then I'll be doing some testing to confirm that it works as designed :) ... and hopefully I'll stop blowing up drivers too.
mark.
Quote from: MarkSnoswell on November 08, 2007, 08:18:49 PM
Quote from: eldarion on November 08, 2007, 07:41:23 PM
I assume you meant to:
1. Close both switches
2. Open both switches
3. Close the negative-side switch?
er -- yea. I think. When I get it all working in the simulator I'll send you the spice model if you like. ... and then I'll be doing some testing to confirm that it works as designed :) ... and hopefully I'll stop blowing up drivers too.
mark.
Hi Mark,
I know what you mean about the drivers. In the course of trying to drive the primaries with only the MOSFET driver outputs, snap crackle pop...all 12 SMT UCC27322s on my board are dead. :o >:( :-X
I really like the above idea, though, and will start working on a new high and low side driver board for the MOSFET totem pole you are describing.
Onwards...
Eldarion
Quote from: eldarion on November 08, 2007, 08:44:48 PM
I know what you mean about the drivers. In the course of trying to drive the primaries with only the MOSFET driver outputs, snap crackle pop...all 12 SMT UCC27322s on my board are dead. :o >:( :-X
ah -- gasp... the drivers have great speed and current buy lousy voltage ratings. Any back emf and they die.
Right now I am leaning towards these parts:
Drivers:Intersil drivers http://www.intersil.com/products/pt/parametric_table_503.asp *** fastest driver you can get and a reasonable price - $6.30 each. avail in 1x
EL7518 8ns rise and fall time into 1000pf, 12A peak current, 40MHz continuous operation.8 ld SOIC package only. Vcc 4-12 ,18V max.
http://www.intersil.com/cda/deviceinfo/0,0,EL7158,0.html
or IXYS ICDD414 - wich arent quite as fast but have high voltage rating and are reportedly very robust.
Output:STGP6NC60HD $1.17 ea
TO-220; Voltage, Vceo:600V; Voltage, Vce sat max:2.5V; Current, Ic continuous a max:6A; Current, Icm pulsed:21A; Power, Pd:20W; Rise time 5ns
http://www.st.com/stonline/products/literature/ds/13765.pdf
or
IRF820
Isolators:NVE Il711 150 mbps GMR isolator (2500V) --- *the* fastest you can get. Very new spintronic device. from NVE http://www.nve.com/index.php
You *need* separate power supplies for low side and high side drivers and the output stage. The low side driver and output still share a common line which needs filters on it. The high side driver is on a floating supply which gives it some inherent filtering.
Hi Mark,
Ordered samples of the MOSFET drivers, 10 total! ;D I will try to obtain some IRF840s; I used to have some lying around, but alas, they were incorporated into other projects.
One thing is still bothering me: why does Bob's simpler driver board work at all if this kind of fancy switching circuit is required??? No matter what I do, I still cannot seem to get that type of driver circuit to work. Maybe Bob can clear it up when I talk to him Saturday. ;)
Eldarion
Hi Eldarion,
Just thought I would check in and see how things are going with your testing. I just did a quick skim over the last couple of pages of the post and I saw those waveforms that you were getting on the primaries. To put it bluntly, the look pretty crazy coming out of those IGBTs! From the looks of those output waveforms, it seems like your IGBTs are not switching fast enough to turn on and give you a sharp edge. Another problem could be that your power supply is not reacting fast enough to supply the current to the IGBTs. Are you using any filter caps on your V+ supply? If not, putting a few caps (47uF, 0.1uF, and 0.01uF for instance) in front of each of your IGBTs will help to clean up the waveforms very much. That may explain why the waveform looks fine when you only have one IGBT running.
Also, putting a large cap in front of your battery/power supply will give it more kick to help supply the current.
My other thought is about your magnetic bias. How smooth is the output waveform across the magnetic bias winding? I think you were using pulsed input to regulate the output voltage level? Also, how much current do you have flowing through the bias winding?
What were you doing when you blew up your UCC drivers? When I did some testing with them, I quickly noticed that it didn't take much to blow them up. But if you drive the primaries open-ended with the UCC drivers, you *shouldn't* have any problems.
Here's some more tips, the first thing you need to do is confirm that you have a rotating field. I would try placing a cup of salt water with pepper or something in it to see if the field has any effect on it. The rotating field that you need to have should be mainly electric in nature. From what I understand, the magnetic component is just a side effect of the changing electric potentials. Thats why I have been such a proponent of the open-ended drive. If you do it right, you will be able to generate an output on the longitudinal windings. I was able to do this with the green TPU that I posted videos about on YouTube (http://www.youtube.com/watch?v=X6NlO-fVr_U), though I was using much higher frequencies to get the effect.
Thats all I can think of for now, I want to get up to speed on what you are doing and maybe I can help you out more. Meanwhile, I'm still plugging away at my DDS controller, which thankfully is finally in the last phase of the design stage.
God Bless,
Jason O
Hi Jason,
Thanks for the info. I was driving the primary coils closed-ended with a very small pulse width when they blew. Interestingly, on further examination, only the F1 (highest frequency) drivers blew--this is a good thing, as I can route around the dead chips for now and resume testing.
I had a large bank of capacitors at the power supply, but I will try putting individual capacitors on the IGBTs to see if that helps. I had assumed that I was getting some weird waveform mixing in the core and failed to look for more obvious problems. ;)
Eldarion
Quote from: eldarion on November 09, 2007, 12:31:31 AM
I had a large bank of capacitors at the power supply, but I will try putting individual capacitors on the IGBTs to see if that helps. I had assumed that I was getting some weird waveform mixing in the core and failed to look for more obvious problems. ;)
not good enough. You have to have sets of capacitors 0.01u,0.1u, 4.7u ... ideally low ESR and low inductances. The reason to use several sizes is that the caps impedance will drop untill they get to a certain frequency and then their impedance will rise! ... the lower value caps have their minimum ESR at higher frequency -- so by combining several different values you get the lowest ESR up to a good high frequency -- which you need to keep the pulse edges sharp. You need to repeat this supply bypassing for each driver and the output stages. The caps need to be right on (as close as practically possible) the leads to the devices. I tested -- a combo of 0.01, 0.1 and 4.7u SMT ceramics works best -- back it up with 4.7Uf tantalum... and then isolate any common lines between drivers with ferrite beads.
IXYS have some great application notes on their web site. Here is a link to the PDF for their low side gate driver evaluation board -- it's worth examining for general tips on how to drive mosfets fast http://ixysrf.com/pdf/switch_mode/appnotes/evic420.pdf -- study that and take heed!
cheers
mark.
Configuration:
NEV IL711 isolator.
IXDD414 driver. (22V)
2 IRF820 in parallel. (450V)
Load = 5nH 30 ohms. ( air cored primary made from Ni205 resistance wire).
The battery and driver circuit are floated down at - 450V.
22V for Driver from 2 x 11.1V lithium batteries.
Results:
Minimum reliable pulse input of 4ns. (+ 1ns rise and fall times)
Rise time........ 600ps -- yes, no mistake, that's six hundred pico seconds = 0.6 ns
Pulse width..... 12ns
Fall time......... 1.5ns
Notes:
* No ringing, no bounce.
* Peak current in IRF820 = 7.2A each. 30% below rated max. Safely within SOA.
* IRF820 is readily available, cheap and very fast (faster then IRF840) with just 360pf gate capacitance.
* Highest operating frequency of output 41Mhz. - limited by drive pulse width.
* Highest operating frequency of ICDD414 driver is limited by power dissipation (833Mw in 8 pin DIP and 12.5W in TO220):
........in 8 pin DIP package: 700KHz.
........in TO220 package: ~ 10Mhz.
* 22V is slightly overdriving the 820 gates which have a rated maximum SG voltage of 20V.
* Il711 offers 2,500V isolation and level shifting = no problems driving the floating driver circuit at -450V -- still leaves 2,100V protection.
Driver Frequency limits:
The maximum operating frequencies can be increased by operating at a lower voltage of 11.1V = one lithium battery. The trade of is an increase of the rise time from 0.6ns - 1.7ns. The fall time is unaffected. At 11.1V driver supply the maximum operating frequencies can be doubled to:
8 pin DIP... 1.4 MHz
TO220....... 20 MHz
Load Power dissipation limits to frequency:
At 250KHz the load coil will dissipate 20W. The IRF840 output drivers are only dissipating 0.8% of this.
It is clear that the heat generated in the load will be the limiting factor to going to higher frequency. As it is not practical to generate shorter pulse widths the only way to reduce power dissipation is to reduce the output stage voltage or increase the load resistance.
Increasing the load resistance will slightly reduce the rise time but the fall time will degrade linearly with increasing resistance. If a high side load resistor is used to limit power then a mosfet switch could be placed in parallel with it and this would reduce fall times to the same duration as the rise times. Experimentation will be required to determine the most efficient and effective drive mechanism.
Resistance wire vs. high side load resistance.
There are two distinct types of load resistance:
a) Resistance wire coil.
b) Low resistance coil + high side load resistor
In (a) the pulse voltage will drop across the coil giving a distributed field in the device.
In (b) the entire coil will pulse up to the peak voltage.
Experimentation will have to be done to see which mode is beneficial/required for the device to function and produce excess power.
Spice analysis - software:
Analysis done with Linears free LTSpice/SwitcherCAD III software available from http://www.linear.com/designtools/software/
Analysis done from GND to +400V as the spice solver hates to have things reference below GND. IN the real world you just attach the single high side to GND reference and you have negative spikes... I have referred to rise and fall times in the negative going sense: rise = leading edge, fall=training edge.
Cheers
Mark.
PS. Just had a bit of excitment as a big Koala walked around the house :) ... They are prety timid and you can get to within about 2 meters.
Hi Mark,
Good work! The only problem I see with that circuit is the 5nH primary coil--I have a hard time believing that my primary coils are anywhere near that low of an inductance. ;)
When I was mentioning a capacitor for each IGBT, I was just writing sloppily. I was thinking many bypass caps of similar values to the ones you mentioned with near-zero lead length directly across each IGBT and primary coil.
Also, my Earth ground is currently connected to the negative rail, so I should definitely connect it to the positive rail and try again. :) Right now I am generating big, fat positive pulses, not narrow negative pulses...
Eldarion
Ok -- it took a long time but I got it to work in the end.
Here is a circuit and results, for driving the 0.58mH primaries of the BB unit.
The red trace shows the voltage across the coil.
The green trace shows the current through the coil.
All voltages and currents are within the SOA of the parts. Overall power consumption is much lower than driving a lower inductance coil.
Leading edge transition is still 0.6ns. Falling edge transition is 20ns. This is realy the best you can do with these standard parts and a 0.58mH coil.
I will explain the circuit more tomorrow -- exhausted now.
cheers
mark.
Hey guys, Bob Boyce wanted me to pop in here and post a diagram of how my setup is configured, sec, bias cap blocking and load. Please excuse the crudeness of it. Did it freehand in ms paint, pain in the butt. You don't have to use both caps, just one is needed, that's just me.
Rich
Hi Rich,
Thank you for the diagram--it will help a lot to have a confirmed secondary connection schematic like that! :)
Would it be possible for you to post a scope shot of one (or more ;D) of the primary coils while the entire system is hooked up and running? I am having a terrible time getting my system to work, and if I could compare the desired primary waveform(s) (what you get) vs. my waveform(s), I could zero in on a solution much faster.
If you don't have the time or a scope, that's OK--just thought I'd ask. ;)
Thanks,
Eldarion
EDIT: I just realized yet another discrepancy between my build and the working build(s): Rich has a stack of batteries providing the HVDC potential. Those batteries probably have a very low AC impedance compared to my bridge rectifier...I'll see if lowering the impedance of the HV source helps anything.
I was waiting for Rich to post the diagram before responding here. It took him a few days to get the time to draw it up and post. He has been very busy, and on top of that dealing with a terminal illness and recent loss in the family. Thank you Rich for doing this. I guess I was having a hard time trying to put into words what such a simple diagram shows.
The source of the dipole should not matter much, whether it is batteries or a power supply, as long as it is decoupled from earth ground well enough that it does not "short out" the LEM energy flow. The current requirement is so low that you can even plug together a bunch of 9V "transistor radio" batteries into a little "brick" and use those.
My amish friend uses a brick of those 9V batteries to reach over 700 volts as the dipole source in one of his prototypes. Once operating, he can parallel a HV capacitor bank with the batteries, and then remove the battery "brick" from circuit. The capacitor bank is just able to keep it going while powering the load, as capacitors do not convert LEM to TEM as efficiently as batteries. It is interesting to note that when the batteries are in circuit, they increase in voltage while running. The incoming LEM charges them while they are connected. His self-running device uses a HV pulse motor to rotate a commutator that fires the pulses to the inductors. So it is mechanically regulated to prevent runaway and avalanche. The pulse motor and load operates directly off of the dipole source, so no other power source is required. With batteries removed and just running on the capacitors which are being charged by LEM, it continues to run for as long as he wants it to, using HV impulses to flash-light some flourescent light tubes. A totally free light source, as long as the mechanical components don't break down or wear out ;-)
Rich has also noted that the old batteries he uses do not run down, and in fact now hold charge better than they used to. It is as if the exposure to LEM flow has improved the operation of those old batteries.
Bob
Tryed to post,, quess didnt like me.. will try to post the pics anyway, hope that works, will explain after the fact ;D
Rich
Is that thing emitting radiation? :o
Also, is that the bulb intensity when it is in self-running mode?
Thanks!
Eldarion
Quote from: eldarion on November 11, 2007, 04:28:08 PM
Is that thing emitting radiation? :o
Is that the bulb intensity when it is in self-running mode?
LOL Yeah, light radiation ;D.
Looks great Rich! Hey can you take some scope shots of the waveforms at your IGBTs? I think that it would be very helpful to compare them with what Eldarion is seeing in his setup.
God Bless,
Jason O
eldarion, and all, well first time, didn't take, so hope it all works this time. First Picture is the three phase pwm, 1200 volt 400 amp IGBT bricks. Scope shot of the 10.7 khz primary, voltage peak is at 1kv taken with a 100x probe, 100 volts per div.. Primary drive voltage and current are 50 volts dc at 500 ma, apox. 25 watts. Lamp is 75 watt 120 volt. Picture of the germanium LEM detector running. Any more questions, feel free to ask, as I am sure did not get everything. Oh ya that meter is not registering what you are thinking, its what i have available, needed the 50 micro amp movement for the LEM detector.
Rich
WOW! :o
That second picture was of the primary waveform ??? I would never have anticipated that a waveform that noisy would produce good results! Now THAT is encouraging! That thing is ringing to high heaven! No wonder that filtering the ringing made it stop working before!
In that scope shot, were the other two primaries running along with the one shown? If so, could you take a scope shot of just the one channel running with the other two off?
God Bless,
Jason O
Quote from: Jdo300 on November 11, 2007, 04:31:47 PM
LOL Yeah, light radiation ;D.
LOL! Just wanted to make sure I wasn't giving myself a fatal dose of X-rays. :D
From the scope shot Rich posted (which is pretty close to what happens when I scope across my secondary) I would think that the primary ringing is actually desired, or at least does not detract from the effect badly enough to kill it.
I got to thinking--a stack of mostly charged 9-volts will have an impedance of about 600 ohms. If I can decouple my HVDC supply from ground, I think the only remaining step would be to lower the output impedance of my supply, which right now is nearly infinite. At that point I should start to see some interesting stuff--I hope! ;) (I think the infinite impedance, along with the short to Earth ground on one end of the secondary, is probably routing any longitudinal energy away from the load and also blocking it from reaching the load.)
Here's a picture of my setup so far:
http://www.falconir.com/pics/DSC02326.JPG
I have been working very hard to make everything compact, including the gate drive leads. The secondary wires travel over the top of the coil off the right side of the picture. There are also 1.5uF caps directly at the IGBTs to allow for short bursts of current at turn-on.
Eldarion
Jason that is one primary channel only,, and you do want all the harmonics its generating, its hard to see but there is 1kv peak out there too, which is the pwm on and off time. Forgot to mention my sec. bias voltage is a apox 74 vdc, and the battery's seem to be charging ,, after the run I did , battery's measured at 76 vdc after.
Rich
Hi Rich,
Thank you for all the pictures and information!
I just have one more question: is that 75W bulb just barely lit, or is it lit fully and the camera is making it look dim?
Just tying to figure out the COP I should be expecting from this type of setup... ;)
Thanks!
Eldarion
Eldarion, that bulb is about 1/4 lit. I don't want to take it to much further as I am getting close to exceeding the 1200 volt rating of the IGBT's. There is a peak on the scope, its kinda hard to see, near the right side that's at 1000 volts peak. Have already wiped out two of those big bricks. ;D
Rich
Hi Rich,
So your setup isn't quite overunity in the TEM sense yet, correct?
All,
I have cranked the primary drive voltage up to 160V and connected the 60W light bulb directly across the secondary (with no HV potential yet). This step has significantly cleaned up the primary drive waveforms.
Also, with the entire system consuming about 50W of power, there was a very dim glow coming out of the 60W light bulb. This glow took a second or so to appear while the filament heated up. Until the bulb was lit, the primary drive waveform was atrocious due to the high filament resistance and therefore low loading of the secondary.
This scope shot of the F1 (highest frequency) primary was taken with 5uS pulse width, 40KHz highest frequency, 20Khz next lowest, etc. in rotational mode with the bulb dimly lit: http://www.falconir.com/pics/primary_one_highest_frequency_high_voltage.jpg
I have not yet placed an HV potential on the secondary side, as I only had one 160V isolated power supply built. I will need to construct another supply before I can perform that test. ;)
Eldarion
Hey Rich,
Could you possably save your bricks by shunting the HV spikes over X volts back into the positive battery supply rather than burning them up on your bricks? Maybe using zeners or something like that may help save you? Just a thought.
Also, I'm very suprised that your inputs are ringing that much at that low a frequency. Do you see the same ringing if you hook the IGBTs across something like a resistor? I'm curious to know if the ringing is just coming from the brick itself or the coil.
God Bless,
Jason O
Hey guy's, hope to have a new scope shot up soon, made some changes to the primary side that's made quite a difference in current draw and LEM output. Looks like it cleaned up the waveforms nicely too. ;D
Rich
Sounds good! Looking forward to hearing more... ;)
Eldarion
Hey, I thought I would check in and give a brief update.
I now have the second 160V isolated HV supply built. It does seem to be making a difference (finally!) With the HV potential in line with the secondary, I am able to use shorter pulse widths to acheive certain brightness levels on the 60W bulb.
Right now, the system is drawing about 4 amps at some voltage less than 11 volts (have not yet had a chance to measure it, but the two inverters go into low-voltage shutdown mode if I increase the pulse width any further--I am still powering the while thing off of a 3A surge 13.8V power supply). With that input, the bulb is maybe 2/3 of the way lit. So even if not overunity yet, at least the COP is approaching 1!
To answer Bruce's upcoming question, no, I have not yet engaged the magnetic bias. ;)
Eldarion
;D
Sounds great Eldarion!
Does varying the HV bias influence the brightness of the bulb?
God Bless,
Jason O
Quote from: Jdo300 on November 15, 2007, 07:35:53 PM
Sounds great Eldarion!
Does varying the HV bias influence the brightness of the bulb?
God Bless,
Jason O
Hi Jason,
For some reason shutting down the HV power supply (bringing the bias voltage to zero) has absolutely no effect on the bulb brightness. I may have sounded a false alarm here! ;D
I did try connecting the magnetic bias, and besides exerting a powerful effect on my old CRT monitor, it did nothing.
I am curious as to what Rich did to improve his setup? My current COP seems to be very close to what he was originally getting (almost, but a little less than, 1). Speaking of COP, I verified my input power as 41 watts for the approximately 2/3 lit 60W bulb.
I am sooo close, but something is still not right!
Eldarion
EDIT: One other thing: the running converter is generating some pretty powerful magnetic fields--my old CRT gets interesting patterns generated in it when this thing is fired up...
Hi Eldarion,
Just a couple of quick questions, my friend.
What frequency was your F4 during this experiment? And did you attempt to increase the frequency to garner more power? If so, was there an effect?
Did you scope across the "other" longitudinal winding of the core, from the top, bottom, and inside winding on the core, and see ANY signal? Could you post a scope shot of that for us, please?
Thanks, and great job! When the magnetic bais affects things AND also the lowering of the HV potential on the secondary affects things, and when those other windings show current, you will have a "fully" working device as described by Bob. Sooo.... If your cop is at about 1 now, imagine once it is working....! LOL ;)
Are you driving the six primaries as two sets of three, or just driving three primaries? Can you also please post a scope shot of your thin negative pulses off of the primary, since it's fix. Thanks!!
Side note to all:*Please remember that Rich is using a "pulsed" system, not rotational. Also, and Rich please correct me if I am wrong, he has no magnetic bias or "other" longitudinal windings on his core. It is wound simply with a secondary and three primaries. So, in the rotational mode we should be seeing much more power. Time to check everything, again.
Warm regards,
Bruce
Quote from: btentzer on November 16, 2007, 03:40:34 AM
Hi Eldarion,
Just a couple of quick questions, my friend.
What frequency was your F4 during this experiment? And did you attempt to increase the frequency to garner more power? If so, was there an effect?
Did you scope across the "other" longitudinal winding of the core, from the top, bottom, and inside winding on the core, and see ANY signal? Could you post a scope shot of that for us, please?
Thanks, and great job! When the magnetic bais affects things AND also the lowering of the HV potential on the secondary affects things, and when those other windings show current, you will have a "fully" working device as described by Bob. Sooo.... If your cop is at about 1 now, imagine once it is working....! LOL ;)
Are you driving the six primaries as two sets of three, or just driving three primaries? Can you also please post a scope shot of your thin negative pulses off of the primary, since it's fix. Thanks!!
Side note to all:*Please remember that Rich is using a "pulsed" system, not rotational. Also, and Rich please correct me if I am wrong, he has no magnetic bias or "other" longitudinal windings on his core. It is wound simply with a secondary and three primaries. So, in the rotational mode we should be seeing much more power. Time to check everything, again.
Warm regards,
Bruce
Hi Bruce,
My F1 (maximum frequency) was anywhere from 30KHz to 60KHz--no real difference was noted.
Yes, I did scope across the longitudinal--there was a small amount of clipping occuring on the sine wave, so
maybe there was a small DC offset, but it couldn't have been more than a few hundred millivolts.
I'm driving only three of the primaries (see the pic I posted earlier).
I am currently at work, so I cannot post 'scope shots. I do have some other ideas that I would like to try, so I will probably drop off for a couple days and come back with new info and results (I hope! ;D)
Eldarion
Quote from: eldarion on November 15, 2007, 10:37:36 PM
Hi Jason,
For some reason shutting down the HV power supply (bringing the bias voltage to zero) has absolutely no effect on the bulb brightness. I may have sounded a false alarm here! ;D
I did try connecting the magnetic bias, and besides exerting a powerful effect on my old CRT monitor, it did nothing.
I am curious as to what Rich did to improve his setup? My current COP seems to be very close to what he was originally getting (almost, but a little less than, 1). Speaking of COP, I verified my input power as 41 watts for the approximately 2/3 lit 60W bulb.
I am sooo close, but something is still not right!
Eldarion
EDIT: One other thing: the running converter is generating some pretty powerful magnetic fields--my old CRT gets interesting patterns generated in it when this thing is fired up...
Hi Eldarion,
Ahem Ahem..... you said that adding the magnetic bias caused the coil to exert a powerful magnetic force on your CRT? Let me ask you something here.
1. Did the Coil affect the TV at all without the magnetic bias?
2. If you just turn on the magnetic bias without pulsing the coils, does it show a strong effect on the TV?
3. How much current were you putting through the mag bias windings to see the effect?
If you only saw that happen when you applied the bias AND pulsed the primaries, then there is a GOOD chance that you might be starting to form the vortex field. I'm thinking that the toroid should (under normal conditions) completely suck in any magnetic field that is produced. So I find this very intriguing. Hey, try sticking a compass or some water near it to see if there is any effect.
God Bless,
Jason O
Hi Jason,
No, the effect on the TV is just a normal result of a large DC current flowing through a coil. Likewise, the primaries exert their effect on the TV regardless of the state of the magnetic bias.
Sorry! I wish I was seeing more interesting effects...
I am currently rebuilding the entire tri-voltage isolated power supply after finding leakages to ground that could have been interfering witth the correct operation of the device.
Eldarion
Hi Eldarion,
You know, it seems that you are going the very complex route to make your high voltage supply. Why not do something very simple like a 555 timer circuit running on a 9V batter pulsing a transformer to make the HV? Then you just put a large electrolytic cap on the secondary side (through a bridge rectifier of course), and simply tune the frequency to generate the bias that you like. That way you will be completely isolated from the earth ground and not have to worry about shorting the LEM through the supply. You also have the added benefit of the transformer itself acting like a choke to protect the timer circuit on the other side.
I have built several of these circuits and depending on the rise time of the input square wave and the choice of transformer, I've been able to generate voltages anywhere from 50V to over 2kV! Very simple to make!
God Bless,
Jason O
Hi Jason,
I am rebuilding it in a manner similar to what you suggested, but not exactly the same. Remember that I am using 160VDC to power the IGBTs, so I need something high power and preferably regulated, as well as isolated.
Here is what I am doing:
I am using a small 120V inverter (95W max) to power a 12.6V transformer. That is then rectified and sent to power the control electronics. I have also hooked up two more 12.6V transformers in reverse to get two separate, isolated 120VAC outputs. Both are rectified; one is used as the HV potential, and the other is for the 160VDC power to the IGBTs.
Because the outputs are isolated from each other, I can tie the grounds of the 16VDC output and the 160VDC output together, and then attach the positive lead of the 160VDC supply to Earth ground, thereby floating the IGBTs to -160V and the control electronics to -152V (This is causing me some grief, as the CRT is tying the FPGA's ground to Earth ground--sparks! Without the CRT connected, everything is OK, so long as you don't touch the FPGA and Earth ground at the same time! ;))
I then take the HV potential supply, connect its negative lead to Earth ground, and use the positive lead as the HV potential. This gives me 320V of potential difference between the pulses and the HV potential, so I hope this will help to kick-start the effects.
So far, my 75W inverter is doing strange things with the inductive load of the transformer, so I will sub in my 95W inverter and see what happens. Obviously I am very nervous about hooking this directly up to the mains! :o
If you see any problems in all of this, please point them out! ;D
I still wonder what Rich did to make his setup work better?
Eldarion
EDIT: Well, my way just isn't working--it's sucking down 30W in idle. I had been concerned about this possibility. :( I will go ahead and try to build a couple of the circuits Jason mentioned, and see if they will work.
I have been pondering a few things, and I wonder if all applied potentials must be relative to the core itself. That thing should be sitting pretty close to Earth ground, so hence the reason everything we apply to the drive coils has to be referenced in some way to Earth ground? Generating those negative pulses with respect to Earth ground is a pain, as it means I have to float the entire drive system to -160V, and if I want to interface Earth-grounded peripherals such as my monitor, I have to try to optoisolate them, or else clip off the ground pin. Grr...
Rich,
Would you mind posting the value of the chokes you added to your setup? I think it was 30mH, but I am not sure.
I tried adding some small toroidal inductors, and they didn't help very much (or get warm), so they were probably just the wrong value. I also tried the secondary of a 120V-12V transformer, and while this made an excellent HVAC generator on the primary side, it didn't clean up the primary waveforms at all.
With a 0.025F 200V capacitor bank on the output, along with a Schottky bridge rectifier, and 700ns pulse width at 40KHz F1, 20Khz F2, and 10KHz F3, my setup is drawing about 150mA at 160V. The capacitor bank seems to have a power input to it of about 20 watts, so once again very close to unity. Engaging the HV potential is still making no difference, and I know it should be making a large one! The only thing I can figure is that my primary waveforms are still just so noisy that I am getting almost no energy collection.
I can't even try to loop the system yet because my inverter and controller are sucking down about 8W at idle. To be able to loop the system, I need to see the capacitor bank charging at a rate that is about double the TEM input power.
I looked for the snap-together ferrite cores, but I am not sure which variety you were suggesting--my guess is the ones where you loop the wire several times through the core.
Also, could you post a scope shot of your cleaned-up primary waveforms? I would like to know about how much improvement I should be expecting.
Thanks!
Eldarion
EDIT: What you are suggesting sounds an awful lot like an AC line filter(http://my.execpc.com/~endlr/a_B2.gif). Could you try putting one in your setup instead of the chokes and see if it has the same effect? These things are very common, and if it works just as well as the discrete choke solution it will take a lot of guesswork out of that part of the design process. (Obviously you would just leave the Earth ground connections in that diagram disconnected). They suggest these values, which sound pretty close to what you had mentioned.
C1=0.1 uF (the X capacitor)
C2=4700 pF (the Y capacitors)
L1=22 mH common-mode choke
R1=1M ohm bleeder resistor
yup...
Hi Grumpy,
I will definitely take that into consideration. Right now, my top priority is replicating Bob's setup, which uses the "long" pulse widths in the hundreds of nanoseconds. Once I get something working, then I can start varying the parameters (trying shorter pulse widths, etc) and see if the change is enhancing or killing the energy generation.
All,
I was able to locate two large inductors in an AC line filter. An educated guess would put them at somewhere between 50mH and 70mH. Placing them in the locations designated caused one of my IGBTs to blow--I will assume this is a good sign, as I may have cohered some energy that overloaded the IGBT. (There was no reason for it to pop on its own--besides changing the inductors, I did not touch any part of the setup, and this setup has been running off and on for a couple of weeks now with no problems Also of interest, the middle frequency IGBT blew, not the highest or lowest frequency IGBTs).
Now I know why Rich was warning me about using 160V! ;D I will see what I can do to prevent this in the future (probably scale back to 80V or so for now.)
Just to reinforce this for everyone: the choice of the inductor's value is CRITICAL. Too low, and no filtering will take place. Too high, and the core will slowly build up a magnetic field, and once again no or very little filtering will occur.
Hope this helps some experimenters out there! ;)
Eldarion
Hey guy's, got some more pictures to post. Showing the chokes, split cores, scope shot, votage and current. Some pictures of the setup and the 450 farad cap bank I am charging.Ã, Please forgive the mess.Ã, ;DÃ, Eldarion, I don't have an inductance meter, so I can't tell you what the inductance values are, sorry.
Rich
Thanks Rich! :)
From looking at your chokes, they may be of a similar value to the ones that I scrounged up--the only difference is that mine are not toroidal. My waveforms (without chokes and with the wrong value chokes) did not look nearly as clean as those, so that is probably the issue.
I like your capacitor bank. ;)
Eldarion
Hey Rich,
Quick question: are you seeing more than 1.1547V rise per hour on your capacitor bank if you start with it discharged? If so, then your setup is overunity. I am going by the readings on you meters, which say you are putting in 10W. Over the course of an hour, that is 600J, so with that 450F bank on the output you should be getting 1.15V or more starting from 0V if in overunity mode.
The chokes from the AC line filter really cleaned up my primary waveforms, but I still have no TEM overunity. I do feel like I am blind here, not having an LEM detector to see where the problem is! :P
With 8W in, my 0.025F capacitor bank charges to 10V in 1 second. That is only 2.5W out; maybe electrolytic capacitors are just really bad at storing/converting LEM? ::) The controller is sucking down 5W just to power itself, so the TEM COP isn't as bad as you might think, just not overunity yet.
Bob (or anyone who knows ;)),
When I run the system at 200KHz to check for DC on the longitudinal windings, should that be 200KHz on all the primaries or 200KHz P1, 100KHz P2, and 50KHz P3?
Thanks!
Eldarion
All,
I was browsing around this forum and I came across this:
http://www.overunity.com/index.php/topic,3457.msg60495.html#msg60495
So what do you think? Could he be tapping into the dominant energy in a similar fashion to what we are trying to do? He's certainly got enough info to replicate the system on his website. I found the part where his load (the LEDs) inexplicably burns out eerily familiar...sounds a bit like a micro-avalanche (if such a thing exists).
A snippet from his website (emphasis mine):
QuoteFrom looking at the above power peaks we can assume that there are three distinct frequencies where amplification takes place and they are; '12,0xx,xxx', 15,1xx,xxx', '18,6xx,xxx'.
Or he could be full of hot air. :D EDIT: Or maybe not so much...looks like he has been dancing around whatever phenomena Bob is using for some time, but not quite putting everything together to make a system like Bob's:
http://www.drstiffler.com/ecat2004.asp
If nothing else, his circuits may shed some light on the basic phenomenon at work in all of these devices.
EDIT2: Wow, he even found the main 42.6KHz hydroxy frequency!
http://www.drstiffler.com/bfh2gen.asp
Here's a question: did he hear of Bob's work and rip him off, or is this one of those bizarre cases where two people come up with almost identical ideas on their own? ;)
Eldarion
Hi Eldarion,
Don't get too excited, LOL! ;) Stiffler says 42.6 Mhz, not 42.5 Khz Bob uses. I think that Stiffler's work with the cold electricity is more related to SM's unit, just my own opinion of course. And Bob B. did say that there are "hundreds of ways" to tap that dominant power.
Back to the work at hand! ;D
High regards,
Bruce
Happy Thanksgiving Holiday Eldarion and to all the replicators!
Hi Bruce,
Yes, it is important to stay on-task here--Bob's technology has far more potential (pun intended) that anything Stiffler has shown to date.
Having said that, his ECAT unit had some similarities to Bob's unit, just on a much smaller and less controlled scale. It would even seem that he may be creating some LEM, based on the difficulty he had in getting a pure TEM O/U output.
Why is this important? From looking at his results, it looks like he might have unknowingly given us another way to convert LEM to TEM and actually get a polyphase system like Bob's to go overunity in terms of TEM output--assuming his load-powering difficulties were due to LEM being created in the first place.
It can't hurt to try something new on the output side! :D
Happy Thanksgiving to all!
Eldarion
Eldarion, the voltage and current readings from the most recent pictures where taken while still running a light bulb as a load, to show the difference inserting the chokes made. The capacitor bank is made up of 6 2700 farad 2.5 volt capacitors in series. When I was charging the 450 farad capacitor bank, the primary input voltage was apox. the same, current fell to 90 ma., apox. 4.5 watts primary input power. The observed charging rate was apox. .4 to.5 volts per hour. I charged the bank over a 4 day period at 8 hours per day. That would be 36 watt hours per 8 hour charge period or 144 watt hours total. The capacitor voltage at the end of the charging period was 12.5 volts. Each capacitor can store 8400 joules of energy each. So if you charged one capacitor to 2.5 volts by its self and put a dead short load across it, you would have 3360 amps of current until you reached 1.25 volts or half voltage. Just so you know, most of the readings are apox. They would definitely have to be done under more controlled conditions. Oh and by the way happy thanksgiving all. ;D
Rich
Hi Rich,
If you are only putting in 8640J to charge that bank to 12.5V, then you are definitely overunity--by 4x!!! :o
I gotta get some larger capacitors. ;D
Eldarion
I have to verify that you are meaning Farad and not uFarad here. I can find no place to get such caps. 2700 Farad X 6 is one lot of cap. If so can you point me to your seller as I want to get a few to play with. I have two 1F 5v caps and they are neat. Just like small batts.
thaelin
Quote from: Rich SAS on November 22, 2007, 08:43:34 AM
Eldarion, the voltage and current readings from the most recent pictures where taken while still running a light bulb as a load, to show the difference inserting the chokes made. The capacitor bank is made up of 6 2700 farad 2.5 volt capacitors in series. When I was charging the 450 farad capacitor bank, the primary input voltage was apox. the same, current fell to 90 ma., apox. 4.5 watts primary input power. The observed charging rate was apox. .4 to.5 volts per hour. I charged the bank over a 4 day period at 8 hours per day. That would be 36 watt hours per 8 hour charge period or 144 watt hours total. The capacitor voltage at the end of the charging period was 12.5 volts. Each capacitor can store 8400 joules of energy each. So if you charged one capacitor to 2.5 volts by its self and put a dead short load across it, you would have 3360 amps of current until you reached 1.25 volts or half voltage. Just so you know, most of the readings are apox. They would definitely have to be done under more controlled conditions. Oh and by the way happy thanksgiving all. ;D
Rich
Hi Thaelin,
Yes, Rich is talking Farads here! ;) Those are ultracaps (sometimes called supercaps, take your pick).
DigiKey does not stock them, but they do have a catalog entry:
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=495-2064-ND
And an assembly of 6 like what Rich has:
http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=495-2068-ND
Here's a picture:
http://rocky.digikey.com/WebLib/Epcos/Web%20Photos/ULTRACAP%20150.0L,75D.jpg
I have yet to find a place to buy them for a reasonable price, however. :(
Eldarion
Hey guy's, I got those ultracaps direct from maxwell technologies.Ã, http://www.maxwell.com/ultracapacitors/products/large-cell/bcap3000.aspÃ, Link to the data sheet on a 3000 farad unit.Ã, Also working on a more sensitive LEM detector setup, using 1n191 germanium diodes. So far looking pritty good. Oh by the way those pc2500 ultracaps where on closeout that's why I was able to get them for 35 dollars each ;D
Rich
Hi Rich,
Could you please post the two germanium diode LEM detector circuit?
Will it work with Schottky diodes instead? ;D
Thanks,
Eldarion
Eldarion, to meter + -------|<-----|<------- to meter minus, the pickup or sample wire go's to the anode, cathode junction. As far as diodes or transistors go, I have tried the 2n404's , 1n4148 and 1n191 so far and found the 191 to be the most sensitive. But others can be tried. As far as meter movements go you can use 0 to 10 ua, 0 to 25 ua or in my case, I am using a 0 to 50 ua. The 0 to 10 ua meter and 1n191 diodes would make the most sensitive arrangement, might be too sensitive. Hope this helps. ;D Oh ya, Jason is trying to figure out how you came up with the 4x over unity for my little experiment.
Rich
QuoteWhen I was charging the 450 farad capacitor bank, the primary input voltage was apox. the same, current fell to 90 ma., apox. 4.5 watts primary input power. The observed charging rate was apox. .4 to.5 volts per hour. I charged the bank over a 4 day period at 8 hours per day. That would be 36 watt hours per 8 hour charge period or 144 watt hours total. The capacitor voltage at the end of the charging period was 12.5 volts.
Input power:
8hours * 4days = 32 hours = 115200 seconds
115200 seconds * 4.5W = 518400J
Output power = 0.5 * C * V^2 = 70312.5J
So the COP is actually 0.135!! Sorry, not sure what I was thinking earlier! :-[
(Probably forgot one of the multiply by 60s)
I'll try the LEM detector next and try a Tesla type system to convert LEM to TEM.
Eldarion
Bob, Rich,
Do you have any ideas on how I could try to see some anomalous effects (other than energy gain)? I tried setting the frequencies to F1 210KHz, F2 105KHz, and F3 52.5KHz in rotational mode, and adding 0.5A of magnetic bias along with 160V of HV potential, and let it run for 10 minutes or so, but nothing seemed to be happening. I also tried the same thing with F1, F2, and F3 all at 210KHz, but nothing interesting there, either.
I can't help but think that I should know when this thing kicks into anomalous mode, regardless of LEM output, based on energy out and anomalous effects starting, especially with a system this powerful. Am I right?
Bob, if you like, I would be willing to loan my controller to you so that you could mate it up with a known-working coil and see if it is the controller or the coil that is the problem...
The only thing I have noted as of yet is that I can vary the primary drive frequency all I want, and to some extent the pulse width, and no matter what frequency I use the HV back-EMF pulses on the primary always occur at 10uS intervals--even when there are drive pulses in between the observed back-EMF pulses! Also, they are a bit odd in shape, perfectly symmetrical about a peak and there is no indication of ringing. I haven't really seen that in any other inductive system.
So, any thoughts? ;D I personally can't help but wonder if the three disconnected primary windings are somehow interfering with the rotational field, possibly by creating gaps in the field patterns...
Eldarion
EDIT: Further thinking and musing about the field patterns has yielded an interesting idea. I will elaborate futher over the weekend when I have more time (and test it as well! ;))--suffice it to say for now that it involves driving all six primaries in rotational mode just as Bob said, but with an added criterion. If my thinking is correct, this will reduce my IGBT-frying back-EMF considerably, while allowing energy to build up in the core over time, as well as possibly explain why the avalanche mode happens so infrequently and unpredictably.
Quote from: eldarion on November 27, 2007, 01:07:18 AM
The only thing I have noted as of yet is that I can vary the primary drive frequency all I want, and to some extent the pulse width, and no matter what frequency I use the HV back-EMF pulses on the primary always occur at 10uS intervals--even when there are drive pulses in between the observed back-EMF pulses! Also, they are a bit odd in shape, perfectly symmetrical about a peak and there is no indication of ringing. I haven't really seen that in any other inductive system.
So, any thoughts?
That sounds like an indication of a parasitic oscillation with the driver mosfet. I fyou drive the mosfet with a low voltage then you can put it into an oscillation mode wher in the off state it till switch on very breifly at very regular intervals. I ahve done this myself -- the worlds simplest pulse generator - even simpler than blocking oscillator.
-or- it could be something else.
Mark.
Hey guy's, here is a picture of the maxwell 2700 farad ultracap. Put a six inch ruler next to it, to show size. The next picture is the new black sand or hematite cores, one torus and two rods. I included the micrometals T650-52 core for size comparison.
Rich
Quote from: eldarion on November 27, 2007, 01:07:18 AM
The only thing I have noted as of yet is that I can vary the primary drive frequency all I want, and to some extent the pulse width, and no matter what frequency I use the HV back-EMF pulses on the primary always occur at 10uS intervals--even when there are drive pulses in between the observed back-EMF pulses! Also, they are a bit odd in shape, perfectly symmetrical about a peak and there is no indication of ringing. I haven't really seen that in any other inductive system.
Eldarion
@Eldarion,
It sounds kind of like the Solitron Scaler Wave transmitter based on Cadeous / Tensor long, standing coil. So strange that this pulse is not attached to the stimulus pulse at all. I bet it has to do with what is posted in Grumpy's Pulsers thread showing the animations of TEM and LMD reactions to stimulus being distinctly different from one another.
I tried to find the link to the Solitron, but found this by accident at Youtube: http://www.youtube.com/watch?v=NQHTrZbyUE0&feature=related
Wow, Obelisks were scaler wave controllers that draw water up like a pump. The video shows an installation doing this and oh yes barbed wire and armed guards. In USA, the public has tolerated individuals being arrested and held without charges for "Photographing critical infrastructure". If you get caught doing so, they do not even tell your family they are holding you, even when asked and shown missing persons Police reports.
Beware of what we have become... much like what the Soviets were once despised for. Waterboarding is drowning, resuscitation, repeat, until death by 'natural causes' (lack of endurance to hypothermia, inhaling water, etc). I do apologize for talking politics when it has no place here. I mean only to warn others in USA NOT to photograph... that you have no rights anymore, and maybe never really did... just an illusion really, restrained by occasionally 'uncomfortable' public visibility .
I will post the Solitron link when I can find it among my failed bookmark imports from WinXP.
Ward
Hey,
Just thought I'd drop in to say that yes, I am still alive and I am still working on getting this converter working! ;D
Attached is a scope shot of one of the 500ns pulses across the primary coil. As you can see, things have improved considerably since my last postings in terms of pulse shape and ringing. I am still attempting to remove the last bit of ringing that is still present; hopefully once that is gone we will start to see some effects.
Just a note to all builders: Bob and Earl are absolutely correct when they say to treat this thing as if it is an RF circuit. This holds true for all interconnects, not only the MOSFET gate drive! There are frequencies flowing through the coils that are in excess of 50MHz, and they help to create the nice, clean square waves. Using high resistance, high inductance interconnects and unsoldered connections will really foul things up.
I have switched back to using IRF510 MOSFETs, purely for the switching speed. I mounted three of them on a nice big heatsink (with appropriate insulating hardware), and have not yet blown any of them (they don't even get warm with 500ns pulses).
Low-inductance speaker wire seems to work wonders, as well as the liberal use of distributed power supply filter capacitors. Don't just stick all your power supply filter caps on one piece of perfboard 1.5 meters from the switching devices and expect the signals to be clean (yes, this is the voice of experience talking here--pretty stupid!) I have a small value filter capacitor sitting right on the coil's positive leads where they all connect, and it reduced the ringing duration by almost half. DON'T USE A HIGH VALUE ELECTROLYTIC--it will not work at RF, and the voltage will sag horribly! (Not the voice of experience here, simple math will explain this one)
Well, more coming soon, I hope...
Eldarion
Eldarion,
I don't understand the waveform. It seems to me that the first part of the waveform is when the drain shorts to ground. This should look like the straight red corner. Don't understand the lazy, slopping descent to ground. Second, there are spikes going negative that I don't understand. Third, after the ringing settles down, there remains a constant amplitude sine wave, which I don't understand; would expect to see a flat, horizontal line as shown in green. The first cycle of this sine wave is very noisy. Fourth, don't see any HV spike(s) going positive to 100, 200, 400, 1000 Volts. A lot of puzzling things, which indicate to me that something(s) are not quite right in your circuit, measurement, or both.
It would be important to know if this constant amplitude sine wave is an artifact (hum??bad grounding??) or an [interesting] anomaly, e.g. what is the source of this sine wave? I t appears that the period is roughly 500ns, which means about 2 MHz. I have seen an anomaly of a constant amplitude sine wave after perturbing a magnetic field, but in my case the frequency was about 100 MHz.
Earl
Quote from: Earl on December 15, 2007, 04:35:08 PM
would be important to know if this constant amplitude sine wave is an artifact (hum??bad grounding??) or an [interesting] anomaly, e.g. what is the source of this sine wave? I t appears that the period is roughly 500ns, which means about 2 MHz.
Nothing here looks out of the ordinary to me -- all things I have seen in practice. The initla poor pulse shape looks like the result of insufficent bypass caps and/or inductive wiring between the mosfet switch and the load.
The 2MHz ringing is not constant amplitude -- I have seen ringing like that in both simulations and in paractice -- ringing that can presist right into the next pulse. If you look at the envelope over a longer time frame you see that it is decaying as expected.
It takes a *lot* of effort to get clean pulses at the coil. However I dont know that this is a goal that is productive.
Mark.
Hello all,
It was my understanding that the pulse shape (incuding that the pulse is unidirectional, i.e. no ringing!) is extremely important to get the effects to show up. Bob has mentioned several times things like "If you do not see the effects, then your pulses may not be of sharp enough rise/fall times"; "drive each of the 120 degree coils in phase with very sharp and narrow unipolar negative DC pulses"; etc.
If I am wrong here please correct me, as it is very difficult to get the pulses cleaned up and get rid of the ringing! ;)
It looks like the inductance of my interconnects is still way too high, so I will try some Litz wire and see if that might help. These are not long connections (max. six inches, simply due to the way the coils are wound), but they are wreaking havoc on the pulses and the pulse generator. As of right now, the entire assembly wants to go into a very powerful 2MHz resonance, which sucks down about 2A @13.8V, raises the voltage at the other end of the power wires by almost double, and strongly affects my CRT monitor. Not good! :o
I am using three 10-ohm power resistors to dampen the ringing, but the inductance of the interconnects has completely drowned out any effect they might be having. My circuit looks good in simulations, I just have to get all the parasitics reduced as far as possible.
I wonder if short runs of coaxial cable to the coils might be of any help here? Right now the coil leads themselves are about 6" long, but I am hesitant to cut them off right near the actual coils, as there is no way to undo that operation. ::)
Maybe Bob can offer some guidance here? ;D
Eldarion
A follow-up to my previous post:
I have now cut a good 6" of lead length off of all the primary coils, and soldered everything up. This had the added bonus of giving me several 6" pieces of that special silver wire to use for power interconnects. ;)
Attached is a picture of the setup, as well as a pulse out of it. This particular pulse was 150ns long. There is no trace of the prior 2MHz resonance, which is a very good thing! ;D
I do not know how to remove that last bit of ringing. I still think it will mess up the effects, but maybe Bob can confirm or deny this?
Also, I wonder if there is any way to make the falling edge steeper. I'll scope around and see if the MOSFET just isn't turning on fast enough for some reason (very likely).
Eldarion
Ok -- now I can see your setup a few comments:
Put bypass caps on the Mosfets -- right up against the body of them. Use 4.7uf and 0.1uf ceramics on each device. You should also put ferrite beads on the V= and GND lines just before the bypass caps. I would also put them back at the other end as well to supress any HF pucked up by the leads.
You wont speed up the falling edge -- not unless you use Mosfet driven by complimentary signal which will give you a low impedance to pull up the signal fast.
... You can also just drive the coils directly with IXYS 414 or simillar drivers. These have very low output resistance and can deliver 10A at 30V. If you do this you must put TVS's (transient voltage supressors) on the outputs to protect the drivers from BEMF spikes.
You can also clean up the output by putting in a fast diode to supress the BEMF -- this will completly kill the ringing (it does for me).
Oh -- you will get faster times with IRF820 mosfets. Dont put these in untill you have everything working cleanly. To get the fastest switching time drive the gate with the maximum permissable -- 30V. If you do everything right your leading edge transition time can get doen to a few ns or so... thats what I read with a 150MHz scope which is the limiting factor with these fast times.
Mark.
Hi Mark,
Thanks for the tips! What I meant when I said falling edge [of the output waveform] was the rising edge of the gate drive, therefore the driven portion of the waveform. ;)
So the MOSFETs' bypass capacitors would be soldered directly across the drain and source of each MOSFET?
Thanks!
Eldarion
EDIT: Just as I was fearing, there is a problem with the MOSFET drive signal. Top trace is the gate drive, bottom is the switched end of the primary coil. I have to think of a good way to get rid of this induction...
I solder the gate driver back to back with the Mosfet -- there is perhaps only 3 mm lead length between them. If you do this then there is no ringing on the gate drive... if you dont then you are faced with having to put a resistor in the gate drive path to kill the ringing. Unfortunatly this will also slow the switching time.
Eldarion,
you are moving in the right direction.
The MOSFET driver, along with its multiple ceramic SMD capacitors between driver VDD and ground must be physically located at the MOSFET with roughly 3mm or less lead length. There is simply NO other way to correctly do things.
In the drain lead going to the coil, I would insert a series ferrite RFI bead. If you don't have a bead, use a 1/4W or 1/2W carbon resistor with a 4-turn coil wrapped around it's body and soldered to the resistor on each end. Values of around 47 Ohms work well. The parasitic low-Q choke or ferrite bead reduce the Q of any eventual HF/VHF/UHF resonant circuits to a value that stops HF oscillations as the transistor goes through a high-gain linear region when switching.
Earl
Quote from: eldarion on December 15, 2007, 09:39:56 PM
Hi Mark,
Thanks for the tips! What I meant when I said falling edge [of the output waveform] was the rising edge of the gate drive, therefore the driven portion of the waveform. ;)
So the MOSFETs' bypass capacitors would be soldered directly across the drain and source of each MOSFET?
There are no capacitors between drain and source of the power MOSFET. The by-pass capacitors are between driver IC Vdd and ground.
Thanks!
Eldarion
EDIT: Just as I was fearing, there is a problem with the MOSFET drive signal. Top trace is the gate drive, bottom is the switched end of the primary coil. I have to think of a good way to get rid of this induction...
Hi Earl,
That's what I thought! It wasn't making too much sense to be putting a capacitor on the MOSFET itself. ::)
I have a solution to the gate ringing problem, hopefully. I will be posting more info and a better quality picture once I have everything soldered up and online.
Eldarion
OK, here it is, gate ringing try 2.
The top picure is the promised better shot of the setup.
The middle picture is the gate drive to the MOSFET.
The bottom picture is the signal to the MOSFET driver from the FPGA.
I really need to get a shielded cable run between the FPGA and the MOSFET drivers to cut down on all the RFI induced in the signal lines when the coil is operational! :o
At least this seems to be a step in the right direction...
Eldarion
P.S. The MOSFET driver chip (the UCC27322) is underneath the ceramic capacitor marked 104.
So much better than before :) -- actually looks like a half decent square wave pulse now.
I am about to set up a new mosfet driver/switch conbination with 320 - 640 V to the Mosfet switch -- I'll let you know how I go with these higher voltage pulses.
cheers
mark.
Hi Mark,
That sounds interesting! What MOSFETs are you using for those tests?
I just realized I had forgotten to post a scope shot of the pulse measured at the primary coil itself. So, here it is. MUCH improved in my opinion! ;D
Power tests are coming soon...
Eldarion
EDIT: Along with this comes a question for those who may know the answer:
Is the whole point of the fast switching times to generate large, fast, narrow, unidirectional back-EMF spikes that then modulate the HV potential field? Basically, I need to know whether I should suppress the back-EMF spikes or not! If the spikes are at the heart of the effect, then I've just removed all chances of my setup working by suppressing them, and would need to get rid of the suppression network.
BEMF or not -- I don't know. If you think about it the BEMF is a symptom of current flowing with nowhere to go. On the other hand you can apply and EMF with little current flow. So the question is do we want current or voltage? ... and induced voltage the same as applied voltage? WIth all the information I have seen to date I don't know -- we should be prepared to test everything.
mark.
Hi Mark,
I would think that letting the coils generate a back-EMF would be quite similar to just applying a fast, narrow HV pulse across the primary coil as you suggest. It is possible that the primaries are just there to have enough inductance to create these HV spikes; who knows (except Bob!) ;D
I am starting to think that it might be a good idea to let the BEMF spikes occur, as the end result would be a rotating electric field. There isn't too much in the literature on such fields (which makes them an interesting avenue for research), but I did run across this abstract:
Quote[PP1.109] Stability of a Penning trap with a quadrupole rotating electric field
T. Hasegawa (University of Hyogo, Hyogo 678-1297, Japan), M.J. Jensen, J.J. Bollinger (NIST, Boulder, CO 80305)
We will present theoretical and experimental studies of the center-of-mass (COM) stability of ions in a Penning trap with a quadrupole rotating electric field. The rotation frequency of an ion plasma in a Penning trap determines the plasma density and shape, and it can be precisely controlled by a rotating electric field. A quadrupole rotating-field scheme can control pure single species plasmas in contrast to a dipole field, which is effective only for plasmas composed of two or more species of ions(X.-P. Huang, et al.), Phys. Plasmas 5, 1656 (1998).. However, the quadrupole field can modify the trap stability because of the spatial dependence of the electric field. In this study we theoretically and experimentally determine the COM stability condition for ions in a Penning trap with a rotating quadrupole field. The experimental results agree with the theoretical prediction.
Luckily I can de-suppress the BEMF spikes simply by removing the resistor network. Now just to rig it so that I don't blow the MOSFETs and drivers... ;)
Eldarion
The simplest way to control the BEMF is via the pulse width... you just control how much energy goes in and that is proportinal to the BEMF peak. In practice it worked almost exactly as I simulated.
A very big defference with BEMF is that although the current runs the same direction the generated emf is opposite to the applied emf. If you need unidirectional emp pulses this could be bad --
I will be speaking with Bob tomorrow and I'll bring it up.
cheers
mark.
@All
As I see it, if the back EMF HV-pulse has the correct polarity, the lower voltage pulse has the wrong polarity and therefore perhaps an implied wrong rotational direction. It is also possible that the relationship concerning the shock to the environment is according to the square of the voltage. If this is true, than assuming 12V versus 1200V means the wrong polarity has 10 thousand times less effect than the correct polarity.
It is easy to generate back EMF with a switch and a coil; we will see how things work out with, say, avalanche generators. I am building a custom mini METGLASS core to do some experiments with HV generation. Will be using 5mm wide METGLASS strips cut myself from a larger 1 inch band, and 5-minute epoxy glue to give rigid structure and eddy-current reduction. It will be a mess to do, but for 1 or 2 research pieces, why not?
Earl
I was just talking to Bob... based on his past experiences he didn't think there was any significant difference between driving primaries closed or open loop -- and therfore no significant difference between apprlied EMF and BEMF.
so -- who knows? I think Earl is right - done corectly the EMF is insignificant compared to BEMF. Aureliano's group (the Mixicans) have also seen that the output is exponentially proportional to voltage.
As for clean signals -- Bob said that you want to apply clean pulses to the coils but that thay can ring however they want -- the ringing is not a problem.
cheers
mark.
OK, thanks for clearing that up! ;D
I will remove the suppression network and see what happens!
Eldarion
@All
I remind everyone that in my experience, a series damping network consisting of a 1 nF ceramic in series with a ferrite bead soldered directly between drain and source knocks out the overshoot on the back EMF pulse, without effecting the main BEMF pulse itself very much. Such a lossy-L / C network is fundamentally different from a R/C network, often used. I achieved fantastic results by choosing the correct bead from a R&D variety designer's kit..
Every scope has a rise time and every scope probe also. Both have to be fast in order to see narrow spikes that may contain enough energy to destroy a transistor.
Earl
Hi Earl,
Thanks for the tip! Right now, I am just going to let the coils ring as they wish, like Bob said to.
All,
I have removed the suppression network and tried a couple of tests with no load on the output. With 13.8V in, I can get an 80V p-p sine wave on the output with one end of the secondary tied to ground. Applying 12VDC "HV" potential to one end of the secondary (the one that was tied to ground earlier) merely shifted the waveform up by the 12VDC, and unfortunately did not increase the p-p magnitude of the signal at all.
So something is still wrong; any ideas? ;D (My first thought is that the "HV" potential is too low, and I will see if anything changes with 160VDC HV potential).
The waveforms on the primaries are rather interesting, I will post a scope shot later on.
Eldarion
EDIT: I tried it with 160VDC as the HV potential, and unfortunately there was no change.
I am runing all of these tests without the magnetic bias. It would be really nice to know if the magnetic bias is required to see any effects or not! ;) (From what I understand, rotating electric fields in a vertical magnetic field cause odd and interesting things?)
Quote from: eldarion on December 20, 2007, 03:06:48 PM
EDIT: I tried it with 160VDC as the HV potential, and unfortunately there was no change.
I am runing all of these tests without the magnetic bias. It would be really nice to know if the magnetic bias is required to see any effects or not! ;) (From what I understand, rotating electric fields in a vertical magnetic field cause odd and interesting things?)
Hi Eldarion,
I have enjoyed watching your progress on the waveform.
As for what you say above, I say, "fire it up and let 'er rip!" LOL Then you will know if there are any effects or not. ;) Just make sure that the sky is clear above....
Hi Eldarion!
It has been great watching your progress with your build and I commend you for your tenacity to see this through. I have been moving back to my home town of Toledo for the last few days so I don't yet have any of my lab things unpacked. But I did a few interesting experiments over the break that I will pass along to you later.
As for your question about the magnetic bias, from what I understand, the rotating field is completely useless without it. We need the orthogonal vector to turn the 2D planar field into a 3D vortex field. Thats what the MAGVID people were saying.
God bless,
Jason O
Quote from: Jdo300 on January 07, 2008, 02:47:34 PM
Hi Eldarion!
[snip]
We need the orthogonal vector to turn the 2D planar field into a 3D vortex field. Thats what the MAGVID people were saying.
[snip]
It is that simple. Rotate electrical charges in a circle, then then punch it orthogonally to create a vortex. The only things between now and success are a "few small details" to be trimmed up.
Earl
Hi all,
Thanks for the encouragement. ;) This whole setup has been on the back burner for a couple of weeks while I worked on other things, but I think it may be time to pull it back out and try some things again!
Regarding the details to be worked out:
If the goal is simply to rotate charges in a circle, there are probably far simpler ways to do this, but I won't go there. ;)
I did fire it up last night and noted that there is a small (~60mV) DC bias on the output from the secondary. My guess is that we need to have high voltage spikes on the control coils (possibly from back-EMF spikes on control coil disengage) in order to drag the charges around the secondary, just like Tao's posting so long ago.
The only issue with that is that when the pulse fires off, the generated electric field will both buck and boost the DC current flow in the collector. (Think of placing a high voltage charge near a wire--it will repel electrons both directions down the wire, not just one) I have yet to figure out a way around this issue.
Jason,
I would be very interested to hear about your interesting experiments. You never know what might suddenly cause everything to "click" and start working. ;D
Onwards as always...
Eldarion
Quote from: eldarion on January 07, 2008, 04:44:21 PM
Hi all,
Thanks for the encouragement. ;) This whole setup has been on the back burner for a couple of weeks while I worked on other things, but I think it may be time to pull it back out and try some things again!
Regarding the details to be worked out:
If the goal is simply to rotate charges in a circle, there are probably far simpler ways to do this, but I won't go there. ;)
I did fire it up last night and noted that there is a small (~60mV) DC bias on the output from the secondary. My guess is that we need to have high voltage spikes on the control coils (possibly from back-EMF spikes on control coil disengage) in order to drag the charges around the secondary, just like Tao's posting so long ago.
The only issue with that is that when the pulse fires off, the generated electric field will both buck and boost the DC current flow in the collector. (Think of placing a high voltage charge near a wire--it will repel electrons both directions down the wire, not just one) I have yet to figure out a way around this issue.
Jason,
I would be very interested to hear about your interesting experiments. You never know what might suddenly cause everything to "click" and start working. ;D
Onwards as always...
Eldarion
Eldarion,
You are correct. As you pulse a primary, there is a shock wave that travels in both directions, but you are looking at it from an overall perspective, and not accounting for the tiny details that are so important. Consider this in relation to time and space (direction}. As fast as it occurs, to really understand this, I want you to slice that pulse event up into tiny fragments of time, and picture in your mind what occurs as that pulse travels down that wire in the winding. As you pulse the winding, the potential impulse travels from the driven end of the coil, down that coil, imparting a bi-directional spin (2 axis rotation) to the shock wave that is traveling towards the undriven end. The energy of that impulse is additive from turn to turn in the driven winding, imparting additional spin energy and amplitude to the shock wave as it passes along. This is what I meant when I likened this to a particle accellerator. The counter shock wave on the other hand, does not get this reinforcing energy, as the counter of each has opposing 2 axis rotational energies. The result is the shock waves travelling CCW are well formed, in spin phase, and are very uniform. While the CW travelling opposing shock waves are poorly formed and out of spin phase. This means that while rotation in one direction is correct, the other rotational component of spin is in the wrong direction. These counter shock waves do not even cancel out the desired shock waves due to the phasing and velocity difference. They do not add up, therefore they are just as noise and are to be ignored.
This is the main reason that the direction of winding is so important in this design. Direction of rotation can be related to the orientation of the earths magnetic field, that only determines whether the unit bucks or boosts in relation to the location on earth. While that impacts efficiency, it is not nearly as important as the direction of spin. The wrong direction of spin determines which subatomic particles we are coupling to.
I hope this helps in visualizing.
Bob
Hi Bob,
Thank you so much for the information about the rotation of the shock wave. I have been agonizing over understanding how the shock wave can only go one direction when the field is ejecting waves going both directions when you fire the coil. Is there any specific name for the effect of the waves adding going one way and canceling when going the other way? I would like to study this effect more to understand it better. I know you talked about how in one direction the wave couples to the molecules and in the other direction it doesn't. Is that magnetic resonance or something else responsible for that?
God Bless,
Jason O
Hi Bob,
Thank you for that explanation--it helps quite a bit.
I can now also take a stab as to why my setup isn't working--I am using IRF510s, and they are taking almost 25ns to switch on. Not good! With that slow turn-on time and the "short" primary coils that we are using (comapred to the speed of light), there is no shock wave forming in my setup.
I am ordering some IRF540s, IRFPS3810s, and PCP116s (as you have recommended for the hydroxy system) and I will see if I can get faster switching times with those.
Eldarion
OK, time for another update!
Swapped out the IRF540s for the IRF510s, still have an atrocious waveform even across a resistor, so I need to do some work on improving that. It is quite possible that I am undervolting the UCC37322 drivers, and therefore enough current may not be going into the gates, causing slow switching. I will definitely look further into this, as I am now quite convinced that clean, fast switching is an absolute must. (I know Bob has been saying this all along; I just didn't quite get how fast--a 25ns pulse rise time with some hefty ringing on the pulse is NOT going to work here ;))
I found reading about the Dave Lawton system quite interesting, including the bit about placing a light bulb on the output, tuning for the best light output, and watching the input power actually drop. Sounds quite familiar, actually. :D He did offer one more bit of information--when he looked really close with his scope, he found extremely short, extremely high voltage "kicks" when the MOSFET transistioned. Furthermore, they were required to get the system to work properly--it was underunity without them. Also, he has managed to close-loop a system this way (sound familiar again? I think these are exactly the same technologies here, but Bob has a much better grasp on what is actually going on and how to extract lots of power!)
You can read more about Dave's system here just for grins:
http://www.panaceauniversity.org/Chapter5.pdf
Back to the bench... ;D
Eldarion
Quote from: eldarion on January 12, 2008, 08:43:27 PM
OK, time for another update!
Swapped out the IRF540s for the IRF510s, still have an atrocious waveform even across a resistor, so I need to do some work on improving that. It is quite possible that I am undervolting the UCC37322 drivers, and therefore enough current may not be going into the gates, causing slow switching. I will definitely look further into this, as I am now quite convinced that clean, fast switching is an absolute must. (I know Bob has been saying this all along; I just didn't quite get how fast--a 25ns pulse rise time with some hefty ringing on the pulse is NOT going to work here ;))
I found reading about the Dave Lawton system quite interesting, including the bit about placing a light bulb on the output, tuning for the best light output, and watching the input power actually drop. Sounds quite familiar, actually. :D He did offer one more bit of information--when he looked really close with his scope, he found extremely short, extremely high voltage "kicks" when the MOSFET transistioned. Furthermore, they were required to get the system to work properly--it was underunity without them. Also, he has managed to close-loop a system this way (sound familiar again? I think these are exactly the same technologies here, but Bob has a much better grasp on what is actually going on and how to extract lots of power!)
You can read more about Dave's system here just for grins:
http://www.panaceauniversity.org/Chapter5.pdf
Back to the bench... ;D
Eldarion
Hi Eldarion,
Thank you for the update! Sometimes stepping back for a moment and then digging back in, can give one, fresh perspective.
I like the route you are on, and believe and hope that it will result in success! ;D
@ All
Perseverance and persistence is what is required. Especially for this project. Bob Boyce has shared with us what he knows. He explained that it had been over a decade since he had messed with the direct energy version of his Toroidal Power System.
I am convinced of a couple of things in these months on this project, and Bob's impeccable integrity is just one of them. I have no reason to doubt a single word he has told us, either by writing or telephone. I have said in private and all along, that if the system does not appear to first work, the fault is with either the controller (waveform, etc.) or the coil is wound improperly.
That said, I would urge, in the strongest way possible, those reading this, to continue to see it through to the end. Some have nearly finished their controllers, while others have set it aside for a brief moment. Even Bob is tied up with that new hydroxy project he spoke about, not to mention battling continued illness on the family front.
So, press on in this new year, to success and be encouraged! ;) :)
Warm regards to all replicators,
Bruce
Still working on getting it working; just thought I'd stop in with this thought: ;)
On a normal 3-primary Boyce coil, each primary is about 40 feet long. The required potential shockwave should form if a pulse is applied to the primary whose rise and/or fall time is less than 40ns in length (a nonretarded light pulse will travel about 40ft in 40ns). These rise/fall times are well within the limits of a conventional MOSFET driver IC with standard MOSFETs such as the IRF540Z.
With our six-primary coil, each primary winding is probably only 15 to 20 feet in length (I don't have the exact numbers here, just guesstimating from the dimensions of the coil). Now we will need some serious circuitry to keep the rise and fall time under 15ns. (I think I now have the means to do so, I will test it and post back here with the circuit if it works as it should--2ns rise time, 8ns fall time advertised; sounds good to me!)
This may or may not be a correct analysis, of course, but it seems to make sense to me. Fits right in with stopping a pulse before it reaches the end of a wire, like Tesla mentioned.
A main point here is don't do what I did and try to overlay the hydroxy controller, even a small part of it, onto the six-primary rotational technology--it won't (and indeed can't) work that way! ::)
More coming soon as always...
Eldarion
Hey Eldarion,
Great to hear that you are still plugging away. While I'm still getting settled back in at home (and reconnecting my lab), I have been distracted by the OC MPMM magnet motor that has been causing a lot of buzz lately. But this little motor has also rekindled my interest in the little magnet experiment that Marco did a while back. Remember Marco's dancing magnets video?
http://video.google.com/url?docid=5540717206741162529&esrc=sr1&ev=v&len=209&q=marcos%2Bdancing%2Bmagnets&srcurl=http%3A%2F%2Fvideo.google.com%2Fvideoplay%3Fdocid%3D5540717206741162529&vidurl=%2Fvideoplay%3Fdocid%3D5540717206741162529%26q%3Dmarcos%2Bdancing%2Bmagnets%26total%3D2%26start%3D0%26num%3D10%26so%3D0%26type%3Dsearch%26plindex%3D0&usg=AL29H21jDmJstvePyrlQS0-SdxBAYvHsuQ
I was just talking to Earl a couple of days ago about how I was interested in replicating it and he mentioned to me that Marco actually used some kind of bias on the little toroid that he was pulsing the magnets with! Later, I went through the forum reading the thread about his dancing magnets device and came across this from him:
"it seems to be a DC wave in nature and the coil did get a continous DC voltage and the pulses were added on top.
next i will try to cancel out one half of the sequence to see if i can end up with the phenomenon of induction."
Which also reminds me of what Sauron mentioned once before:
"if we have a powerful magnetic wave present @7.8 Hz and we make the coil cancel one half of the wave , the other half will "hit" the coil which can result in a vibrating "shock" effect and the output will be the other half wave , so basically a alternating half cycle which could look like a pulsed DC is some fashion."
Now this really has me thinking :).
God Bless,
Jason O
Quote from: Jdo300 on January 15, 2008, 03:03:07 AM
Which also reminds me of what Sauron mentioned once before:
"if we have a powerful magnetic wave present @7.8 Hz and we make the coil cancel one half of the wave , the other half will "hit" the coil which can result in a vibrating "shock" effect and the output will be the other half wave , so basically a alternating half cycle which could look like a pulsed DC is some fashion."
hmmm...must have missed this one...
Quote from: Grumpy on January 15, 2008, 10:54:47 AM
Quote from: Jdo300 on January 15, 2008, 03:03:07 AM
Which also reminds me of what Sauron mentioned once before:
"if we have a powerful magnetic wave present @7.8 Hz and we make the coil cancel one half of the wave , the other half will "hit" the coil which can result in a vibrating "shock" effect and the output will be the other half wave , so basically a alternating half cycle which could look like a pulsed DC is some fashion."
hmmm...must have missed this one...
This was in a PM that he sent me :).
Thanks for sharing.
Hello all,
As some of you already know, I recently acquired a Tektronix 500MHz digital scope--this has allowed me to see an additional pulse generation problem with my FPGA-based controller.
For an unknown reason, it generates a perfect pulse sequence on each of the channels, but only for a certain number of cycles, with dead time in-between!!! It looks like I have never actually fed the correct sequence to the coil, ever. (This type of problem is nearly impossible to see without a digital storage oscilloscope with enough memory; hence the reason I never saw it before. On an analog scope everything looks perfectly fine, even though it is not.) Just to add to my befuddlement, the Verilog pulse gen code does not have any errors in it, so it must be Xilinx's synthesis tools acting up again. I am not impressed with Xilinx's synthesis program in many areas; you can guess why! ;)
So it looks like another rewrite of the pulse generator is coming. :( At least I know what the problem is, finally. ::)
Eldarion
Hey Eldarion,
Great to know that you figured out what's up now. I'm very excited to see how things will go for you once you input the correct frequencies!
Keep up the great work; we're all rooting for you!
God Bless,
Jason O
Quote from: eldarion on January 16, 2008, 08:27:30 PM
Hello all,
As some of you already know, I recently acquired a Tektronix 500MHz digital scope--this has allowed me to see an additional pulse generation problem with my FPGA-based controller.
For an unknown reason, it generates a perfect pulse sequence on each of the channels, but only for a certain number of cycles, with dead time in-between!!! It looks like I have never actually fed the correct sequence to the coil, ever. (This type of problem is nearly impossible to see without a digital storage oscilloscope with enough memory; hence the reason I never saw it before. On an analog scope everything looks perfectly fine, even though it is not.) Just to add to my befuddlement, the Verilog pulse gen code does not have any errors in it, so it must be Xilinx's synthesis tools acting up again. I am not impressed with Xilinx's synthesis program in many areas; you can guess why! ;)
So it looks like another rewrite of the pulse generator is coming. :( At least I know what the problem is, finally. ::)
Eldarion
Hi Eldarion,
I hate to say it, but this is one case where bad news is good news! :D
Now that you have diagnosed the problem, perhaps the smell of sweet success is ahead...
No way to get those shock waves Bob described, without the correct pattern as everyone knows. So this is very encouraging indeed. ;D
Warm regards,
Bruce
This Bedini modded schematic should provide you with constant load matching WITHOUT oversaturation and be very effecient. The youtube is just to show how well it does it's job.
http://www.youtube.com/watch?v=4l1PfmTheFM&feature=autoplay&list=ULHx7aWaPbBo4&lf=mfu_in_order&playnext=1