Ward's Build of Bob Boyce's TPU

[HumblePie]

hi Ward @all

I just posted some infor on Earls thread http://www.overunity.com/index.php/topic,2996.new.html#new

After I confirm simulation with experiments (today I hope) I will post a heap of information. It looks like fairly simple circuits can be used for driving -- but you do need to select the right parts and component values. I'ts quite amazing just how well you can do with a simple design -- but you do have to pay attention to detail.

cheers

mark.

Dr. Snoswell,

* Are rise and fall times to open-ended drive as pulse width is to closed-ended drive?  {Yes that is also pun on Schoolastic Aptitude style questions}. 

EDIT 01-Dec-07 ** Is rise/fall time the tweak to be tweaked for open-ended drive builders?  Ooops duplicated the first question by accident. 

Do-Over:
** What is the difference between driving a folded wire from both ends with seperate FET's Vs soldering the two driven ends together and driving them with one FET?
In the link above you talk about closed-ended drive pulse desired not being shortest possible, but that width that yields peak BEMF. 

*** Isn't this another way of saying 'peak resonance'? 

**** If so, does this say 'resonate the Prime'?  (as mentioned in a group discussion where we could use a Bifilar Pancake pickup coil to detect such with and without shielding)  "One Prime at a time 'with other Primes inactive' ". 

     I only ask '****' because of capacitive coupling dominating at 90 degrees as in the one wire experiment Jason showed us recently... and the TPU...
...and how it is said that 'the TPU is really a Tesla Magnifying Transmitter-Receiver rolled into one',

     But In the Meyl video and papers he says 'the resonant Rcvr 'fulfils the conditions required for resonance' and hence is of opposing phase with respect to transmitter.  So my mind wants to think; make the Secondary resonate in response to the the Primary, but they do not ever 'co-resonate' together this way nor do they have any linked phase relationship. They are not common, but the circumference is. 
   
So then I think of the experiment you shared and Jason duplicated with the solenoid coil capped at one end with the folded Bifilar pancake.  (I re-cap this so it makes sense for others.) 

     That test showed:  Drive the solenoid, pancake picks-up in-phase at both ends, pancake upside down no difference (because it is 'flux blind'), BUT drive pancake and solenoid pickup shows 90 degrees lagging one end and 90 degrees leading at other end. 

So now I see the solenoid should be bent into circle and its ends would be 180 degrees out of phase, fulfilling the conditions required for resonance.. in the toroid itself, not the Primary or the Secondary after all.  So the optimum Primes should be folded 'Bifilar' to make them fluxless (not real Tesla Bifilar like Wiki' shows).  I bet this is how they transmit two directions at once because they are flux-free... like Giant's avitar once illustrated.  Bob actually says this.  'drive a coil from both ends to really see results' to paraphrase.  So Dr. Snoswell, can your simulation for driving from both ends fit my proposed folded Primaries?  You said it was the very best way to drive them.

The opposing phase certainly explains how the sum of ALL winds on the Toroid is/are the longitudinal open-Capacitor path of the ~LMD Transmission Line and why all should fully touch... and how it could be made resonant at a totally different frequency than either the Primary or the Secondary, being forced into nodes by circumference.   The opposing phase explains why my Primes near the middle resonate oddly higher than the others (whether 1X or 10X probed, relatively speaking... BTW - one is supposed to use a pick-up coil probe, not connect right to coil else you are adding L & C to them with the probe.

~See the LMD Trans' Line at http://jnaudin.free.fr/html/lmdtem.htm   

I look forward to the simulations, and I wish I had time to learn Spice or Jason's favored alternative simulator because the reports are very encouraging.  Still you blow drivers though, but fewer I bet.  I still have 4 DEIC420 drivers I'll try after I get my balance.

*******************
@All, 

My GAL card 'seems' 100% clean now with DDS 20 as the CLK source, but it won't fit in middle unless I really hack some posts down, and wires going around it would be a last resort. 

My Tektronics 465 O'Scope Time Base Shaft just stuck in B'Dly mode and it is the perfect excuse to spend money on a new scope. 

Please advise with your favorite Scope picks if you were to buy one today.  I am on a budget, but they last so long I'll splurge!  I know I will be glad I checked your opinions before I buy tomorrow.  Thanks in advance.

Ward

[MarkSnoswell]

Hi Ward,
   I don't have time to answer all the questions now --

I did some testing today. I am not blowing any drivers now... but I am blowing mosfets at an alarming rate... and at *way* below their rated voltage and current.

I have to do more tests -- but something is wrong. Before it dies the output waveforms do match the simulations very well although their amplitude is low -- and the mosfets die with alarming regularity. I am concerned that there may be some energy spikes that I am not seeing on the scope -- and that it's this "invisible" energy that's killing the mosfets.

Mark.

[Earl]

Hi All,

here is why I want to "blow up" my transistors or diodes.

http://www.overunity.com/index.php/topic,2996.msg61991.html#msg61991

Regards, Earl

[HumblePie]

@All,

My camera's display just died so no photos for awhile.  See below.  I removed the Op-Amp oscillator from my GAL board and cut the board narrower so it will fit in the ring beside my DDS 20.  Can't make the Opto-isolators listed for the PWM3F work anywhere above 60KHz and I am using the current levels this OCP-PCP116 opto's data sheet calls for... so I will try the IL710-2 in PDIP Pkg rated at 300V.

I could have waited to obtain 1% Capacitors (code 'F') to retest the jitter and thinking that is what delayed action.  However, even if jitter is solved, this oscillator is not near fast enough.  My GAL uses 12X CLK.  I can not use a 6X CLK for divide by 1.5 as the TP900 Pulser does.   A CLK multiplier is really the only way to get stable and adjustable higher freq's without my DDS 20. 

My RC Dly's will suffer phase stability jitter too from the 20% tolerance Cap's.  I thought +/- 20% only affected the actual measured value Vs labled value... it seems the cycle to cycle repeatability is the major issue.

Do List:  Fix and test GAL Card wires being redone.  Change Power Supply and Kill Switch wires to coax, and reinstall 22M Ohm Resistors on the HVPS Capacitor.  I am very close, but am having momentum problems for various reasons.  I will succeed.

Edit 3-Dec-07 - PC just died and the motherboard is extint!  O'Scope, Camera, PC... 3 sigma says it should be smooth sailing after I get a new PC.  Using Vista box I'm borrowing Vs WinXP is going to be like running in mud and I all those book marks I hope I can recover.  Maybe the time for Linux has arrived.

I don't know what I lost yet.  Be safe -> REMEMBER TO ARCHIVE ALL OF THIS DATA...(all text, all files, all bookmarks)... ESPECIALLY THE OLDEST STUFF LIKE RECENT POSTS OF TAO, EM, MARCO, JASON, MDR, GIANT, etc (many others... I just read these first... order is random Thank you all even those not mentioned).     

THIS DATA WILL NOT STAY STAY HERE FOREVER.  CAN YOU ACE IT ALL BEFORE IT EXPIRES?  Trust me on this.Ward[/img][/img]

[HumblePie]

@Hey All,

New horrible Vista OS PC has worn me down after a day wasted 'trying' to import my WinXP bookmarks into Vista.   An expert assured me it had to be done as was shown to set ownerships for importing only to find I could not per URL format Vs html. 

Ha.  I felt like the guy salting before tasting I did.  I could have tested a few files before clicking about 10,000 clicks (really), but it gave me time to think about things. 

While I burned that time I kept replaying Dr's words about 'content' of that previous post I made, I assume about the DC to DC power supply and/or tuning characteristics I saw. "Interesting" I thought.  It wasn't even a toroid.  I was so focused on wanting that inductor to be very tiny I only tried one that was toroidal... a bead really about 5mm diameter.  It ate mAmps like candy. 

Its circumference was so small and this apparently is what made its impedence so low when driven at such low frequencies as a 555 delivers.  It did not matter how short its pulse width was.  The low frequency was inappropriate for the circumference.


Then there was the Ramp Wave pattern seen in the DC2DC HVPS output that I kind of breezed over describing.  Remember to/(I) start(ed) with lowest freq' and shortest pulse... which always uses the least mA of any pulse width.  As I increased frequency with shortest pulse width, the voltage out followed in a linear fashion until it SNAPPED up dramatically with PLL characteristics... meaning I had to reverse and decreased frequency more than expected to get back to the low voltage out 'low-peak'.     The wire length was trying to hold the wavelength of the frequency into integer mutiples of its length it seems, but why the asymetrical behavior?  Why is the ramp wave pointing with ramps going up and towrads the low frequency side of the graph ( I made a mental graph, else I'd show you).  Try it if you have not, else be disadvantaged I bet.

Sounds just like SM telling us about his " 15" Stereo Speaker Amp that he had to leave at 35KHz and change because it fought his tuning attempts to reach 35K... so he left it there at this frequency where he could not hear that > audio band anyway... SM was telling us what the circumference does to resonance. The exact change over 35K is related to circumference. 

BUT my little inductor yielding 500VDC out is just a 1:1 chip Transformer.  NOT a toroid at all.  The frequency was fitting the length of wire and that wire length forces nodes.  Forget about if it is integer # of circumferences that fit wave or wave length of the FREQUENCY fitting the circumference (no keep pondering that too!).

  The "Phase velocity" of the pulse or transition we drive is << than SOL in Copper and has something to do with the "Acoustic wave"... but somehow it is much slower than the pulse or transition we drive into this thing.  I read start of EM's thread 'Exploring Addition of Signals in SM's TPU' and got to thinking about what in the heck EM means by "Phase Velocity".   

http://www.overunity.com/index.php/topic,3138.0.html
My take on it is:  ->   In here he talks about how the wavelength traveling at SOL IF matching the circumference would have its peak fall at the very same place around the toroid each and every cycle... by nature of the wavelength fitting the circumference exactly.  So actually when using a slower frequency we seek the next best thing... an integer # of circumferences fitting exactly into the wave length.  Yes the signal will overlap onto itself and add to itself, but the same each and every cycle, as long as exact integer multiples apply.     If SOL, we need 3.2GHz I told EM.  He said 'the acoustic wave is slower' I think he said.   He describes rotation emenating from slightly less than or more than integer # of multiples (in my interpolation... he really says literally wave length = circ', but forget that for now).   

So then I replay thoughts about
MAGNETOSTRICTION... the audible buzzing when inductor pulsates and stretches conductors and affects magnetic impedence and vica-verce (if there is such a thing).     Audible/acoustic implies low frequency... what?   Well if that slightly "Off-tune", it would tend to rotate the peak aroud the toroid because its end would land at a slightly shorter landing zone than each previous cycle did if it was anywhere less than 1/2 a circumference shorter than a perfect integer # of circumferences in wavelength.  When it is a little shorter still, say 3/4 of one circumference, it is just like being 1/4 > than a perfect integer # of circumferences long, and the pulse peak rotates the other way.  NOT, same way, opps. BUT it is the OPPOSITE direction that the pulse is being driven!  This landing zone of the 'start' of pulse (or end of pulse if it makes you think easier) seems to be what is being called:

PHASE-VELOCITY...

I think... and this can be made to travel very very slowly around the ring either CW or CCW... all with one frequency.   INTERESTING stuff EM.  Now why not drive folded Primaries so they are driven at both ends (same end both because folded back on itself) to cancel flux so there will be bidirectional transmisisons of pure longitudinal pulses capacitivly coupled around the toroid?  That way you get bidirectional counter-rotaing longitudinal "Clashes" kind of like MagClashes.pdf (a masterpiece of pdf'ing BTW... nice clash image!)   
     AND why not just solder the side by side leads together of this folded bifilar Prim so can be driven perfectly simultaneously with a single FET?  Oh yes I want to try it fast.  I am the slowest hands on guy I you know, but I'm trying hard to understand how this toroidal one-frequency MT works and this slows me down just as my lab partners tell me it paralyzes them.  'Keep thinking every day' I urge them/him.  It is a little similar to Jason's green mini-TPU here:

http://www.youtube.com/watch?v=X6NlO-fVr_U,

but the drive of the Primary is not same as I describe.


Haste makes waste.  I must try the salvaged PC power supply sized toroid to see what it does. Toroidal DC to DC power supply I read was recalled from Min'Man'Miss'l 'deployment' for being 'ou'.  I see this vending machine is said to be also... notice the big toroid:

Bruce's sticky to  http://altenergy-pro.com (no FORGET the pictures really, just MEMORIZE the text there called 'Magnetic Resonance Specrum")... it says to me take EM's rotation in 1 ring with 1 freq and put a hundred identical passive rings above and below that will rotate for frre-ride on the back on the active ring WITHOUT DIMINISHING THE SOURCE... oh ya... INTO INFINITY without diminishing the source.  Like gallery photo #2 shows with the stove pipe satalites resoanting for free in presence of active transmitter in center.  An MT I bet... Lindsay would not answer when I asked about "Pipes, tubes, and plastic" back in late June in SuccTPUreplicat' thread.  It is not like an alternator that loads your engine down as you turn on the lights... he he!  It says that anyway.  I will try.

About doing everything (meaning some things) backwards from what we've been taught (taught where/when ha ha).  There is some contratdiction by specification so be prepared to test all hypothesis.  If they prove false, test the opposite for grins.  I would have bet lots that min pulse width to the 9V DC to 500 VDC PS would have been more efficient than max pulse width, but opposite was true to about almost always on pulse width where mAmps climb finally.


AND realize I am a fool if I do not place the bridge accross a toroid instead of 'any inductor'.  Like the Japanese self-powering vending machine appears to have a Boyce sized toroid in it keeping its batteries charged always!   This bridge saves what the diode blocks and is normally wasted... sending it back to the battery, with a Cap accross it to smooth and sooth the battery input.

To see the Capacitor equivlent circuit that keeps batteries charged, see gnosys.com links "MUST READ" and you'll find:
http://theverylastpageoftheinternet.com  and a 'slightly' wrong schematic of the Capacitive circiuit from an Engineer who heard Mr Bedini speak about it and just built it.  Both modulate the equilibrium.

OK so my brain is cleared of this thinking.  Now how is the transmitter and receiver of this toroidal MT to work without a tank circuit.  Giants new Hetridyning tread jut posted links back to this open-capacitor Tesla MT again and it has tank Ckts.  Grumpy points out these coils are being used like Capacitors, not inductors and could be replaced with some kind of 'capacitor' plates ('implied')... hmmm.  Interesting.  I will eventually connect all this, but not without hands on!  I bet the EM style rotation direction change with slight freq'/wave length change is somehow related to the ramp wave I see in HVPS V-output Vs frequency.

Ward