user TURBO?s replication of Steven Mark?s TPU ?

[giantkiller]

Congrats GK
I agree  the best way is to share the info and replicate the heck  out of it so there is no "single point"... I am rereading it all  and will get wound within a day or two...some  points are a little tough but most of the info is there. Thanks
BTW I found a nice 555 at Centerpointe Elect. FG-500K $38  will go 1-100 khz
Ay more details would be great !!!!!

the 555 here is a timer chip, right? the lousy ones can be to 1.5 mhz. and not $38bucks? u mean $0.38? But they drift bad and that is good starting out. Or get a real good stable clock and add the thermal shutoff attached to the coil and a foot operated dead man switch.

Remember: drift is you power up friend.

--giantkiller.

Hi GK and All,

I think 'boxofsparks' means the type of a function gen which costs $38 but call it a 555.

You probably have heard of the CMOS version of the 555 timer IC, that is LMC555 or TLC555  (but not including here the MC1555) and they work up to 3MHz in astable mode from 5V supply voltage. See data sheet: http://www.national.com/pf/LM/LMC555.html
And I found a very 'handsome' Design Idea at electronic design magazin site: a new stable RC pulse generator circuit with the 555 chip, see a printer friendly link here:
http://www.elecdesign.com/Articles/Print.cfm?ArticleID=6235 

So I think three of this pulse generator could serve as a stable frequency source for driving directly the power MOSFETs. Independent duty cycle and frequency adjustment!

My question for you on your TPU is: how much the drain currents are when you switch on the MOSFETs?

regards
Gyula

Hi GK
Sorry I miswrote on the 555...Its a complete signal generator 1hz t- 100khz stable. Heres the link.
http://www.cpcares.com/ElencoElectronicsInc/EFG500K.html
Good work on your end thanks for info.
Jon

hi,
I use bipolars - tip41a

I have run with 1,2,3 freqs. Greater then 2 and they have to be synced together! I saw the 3 clocks I was running with drift and when they randomly crossed I got kicks. This has to be controled because that is where the heat came from. SM explicitly warned against this! So please keep that in mind.
The runaway effect is very high speed. That is what rocked my world today. Yes it is a cool result. But i got real lucky. Nothing can stop anybody from trying anything. But be aware. I also got warned about the radiation effect today. I am sitting right next to it testing. Some days we are dummer, no?
So now you have a very real part of the experiment with SMs warning right out of the videos and prior posts.

[mrl]

GK - Pull yourself together and stop freakin' out!

Point taken. I do know this. Tonight I will see more results just as easily as before. Today will mark the 1st day of 2 sets of results, morning and evening. It is just inevitable that when big things change quickly in a positive manner,  elation takes hold with suspicion. Things are cool. I am trying to take the next step safely. But man, this happened damn fast!

Everyone has everything they need since page 55. Garden wire is standard size and there are 6" leads coming out of the collectors. My 555s are @ 5v the tip41a-s are @ 12v. Bias ohms are 100.
All the content is off site already and I don't know who or where. I also have an HTML copy. Anybody can get their own too. But remember, it does not copy the quotes or the graphics.

I need to implement very stable clocks instead of the 555s. And that is after the foot switch and the thermal shutoff is installed. Thats right. I expect more phenominal results. Anybody want to jump on the clock design? I could use the help and they could use the inclusion.

I know that by using the different 555 cap values, it saved my butt by keeping the 3rd freq off. I approach this next result set with caution and intrepidation.

Again:
coil diam is 6"
3 layers of:
collector coil is garden wire, 22 turns.
control coils are 4 segments of 30 gauge magnet wire, 200 turns on each collector.
the feedback coil is 16 gauge speaker wire (single run) wrapped around all layers in 4 segments of 17 turns.

Then interconnect the whole system by Ottos diagram.
I power the 555s @ 5v and the tip41a-s @ 12v.

Freq range is 200hz for all 3 555s.

My new power supply is 12v @ 8.25a

First law of uncontrolled TPU power up is: It gets hot real fast when you hear the clicking.
You been warned!
If you are doing this then tell me on the side if you don't want to go public here.

--giantkiller. All posted again for the utmost enjoyment of others.

You don't need to go to a crystal clock.  All you do is take a 74HC14 hex Hex Schmitt Trigger,  rig it up to be an RC oscillator and use a multi turn pot.  Run it in the mHz region, then pump the output through a frequency divider.  You can get multi stage counter dividers in one chip or you can use a 4 bit ripple counter.  This will divide your frequency by 16, or more if you use more counters feeding each other.

What the above configuration will do is give you and fairly stable oscillator.  It is stable because the affect of the drift and  jitter of the the high frequency raw oscillator will be divided down by the divide by 16 counter.  The more devisions you have the more stable your output will be.

mrl

[aether22]

[Airpland Joke] Giantkiller, get ahold of yourself. shake shake.  Here, let me take care of this, Giantkiller snap out of it, slap slap. Doctor you're your needed over there, let me take care of this... [/Airplane Joke]


Directions:

Wind 3 coils of iron (or is it steel?) wire just under 6 inches (garden wire) of 22 turns each with 6 inch leads, optionally on a dielectric (SM used wood, GK used no dielectric core AFAIK)

GK, can you plese clarify the thickness of the wire, no idea what 'garden wire' is, is it iron or steel?

Wind a control coil of 200 turns on each iron core with 30# copper wire, wind it in 4 segments (50 turns a segment), covering 90 degrees each and connected as shown in Otto's fig. 4 otto_control_coil wiring1.jpg. The way I understand it each ring recieves a different frequency on it's control coil?
CORRECTION, 200 turn segments x, that's 800 for each ring, that's 2,400 times your going to pass a small spool of wire through the center! (not counting the feedback coil)

GK, It is not clear if the segments are magnetically bucking or not. (is the flux circling inside the ring, and if so cw or ccw?)

Wind a feedback coil with 16 gauge speaker wire (single run) wrapped around all layers in 4 segments of 17 turns. (likely connected the same as the control coils)

GK, in the picture (snap 0261), I can see that you have what appears to be wire with 2 insulated cores, which makes sense as you said speaker wire, but I only count half the number of pairs you mentioned. (17 seems about right if I assume it's single core though)

Connect all coils as per Otto's diagram,

Howeever GK said 'Bias ohms are 100', GK are you using 100 ohm resistors (what wattage?) when though Otto uses 12 ohm 1 watt resistors?


For the driver, 3 555s @ 5v and the tip41a-s @ 12v.
http://www.alldatasheet.com/datasheet-pdf/pdf/5769/MOTOROLA/TIP41A.html

Freq range is 200hz for all 3 555s.

>My new power supply is 12v @ 8.25a

>I went down to around 200hz. The tuning fork example of 'C' puts it at 261.6hz.
>My 3 555 are @ 2.2Uf, 2.2Uf, & 4.7Uf. Yes they don't match!!!!!!

I suppose he means the caps he is using on the 555, along with a single turn 500k pot.


Running: It can take many seconds to really do anything.
    Use magnets on the inside
    Watch it, it will get hot FAST! (when you hear the clicking)
    Viabration and gyroscopic effects are present!
    I (aether22) would guess that it is probably OU based on how fast the metal heated up. (2 secs)


You need a scope, but an old analog (or new analog ;) without storage (ability to freeze) is ok apparently.

Also GK used Dave's 3x 555 circuit I believe, can someone dig it up? (it was so many pages ago)


Questions for GK:
recap of questions for GK so they hopefully all get answered

1: Is it iron wire or steel wire? What is it's thickness? Does it appear to retain much magnetisum (before you used it in the TPU)

2: As the different segments are wound in different directions, are the segments bucking or attracting? In other words are half creating a cw magnetic field with the other half creating a ccw magnetic field? If not is the magnetic field cw or ccw inside the coil?

3: In the picture (snap 0261), I can see that you have what appears to be wire with 2 insulated cores, which makes sense as you said speaker wire, but I only count half the number of pairs you mentioned. (17 seems about right if it were single core though which clearly it isn't)

4: Bias ohms are 100?, GK are you using 100 ohm resistors (what wattage?) when though Otto uses 12 ohm 1 watt resistors?

[starcruiser]

For All,

A small correction aether22. The control coils are 200 turns each for a total of 800 turns. The smaller the wire the more turns you can get. I am thinking the 30 AWG is probably better but you can get results with 26AWG.

The collector core can be either iron/steel/zinc composite wire, the garden wire can be found at the local hobby store or garden center, coppper is preferred by some builders (Otto and myself, I think it will provide better reaction). Personally I am using 26AWG solid core copper for the control coils. The feedback coil has been 18AWG (for Otto's) or 16AWG stranded (GK), I am thinking the feedback should be 16 or 14AWG copper, better for handling current and heating.

Remember, we are still testing these things out so the specs can change. Also remember Otto is using all copper and has achieved output of over 400V with the wiring diagram posted. He found better results using MOSFETs which may provide better isolation from back EMF from the coils.

So I say have at it boys!

[aether22]

Starcruiser, have you had gyroscopic forces display in your TPU?

Have you had extreme anomalous heat?

As far as I'm aware GK is the only one to have anything truly weird. (anything stranger than a step up transformer effect)