Steven Mark`s TPU

[ronotte]

Earl,

as always you are concrete and pertinent (your experience does show...): thanks for your comments.

"why the very slow oscillation in output power ?"
Well in the first I thought them out of a 'non perfect synchronism' between the 3 freqs, but now I'm not so sure, why?......the point is that the 3 freqs for themselves are NOT SYNCHED, they are automatically put, let say, in synchronism by the processes actually running into the 3 rings: it seems LIKE they get synchronized...HOW?: I don't know.
What is clear to me is that them ARE PERFECTLY SYNCHED as the scope synchronism is  triggered with extreme precision. I mean that if I change less than 1 Hz one of the 3 freqs  NO WAY TO SYNCH the scope!  You know what I mean...very, very difficult to force a professional analog TEK scope even if in 'manual trigger' to get in synch with random signals......that's also one of the reasons why it's so difficult for guys without prof equipment to deal with this kind of devices. To come back to your question my feeling is that sayd effect is not due to freq sliding and consequential less waves superposition....so what else?...I'll wait for any contribution better if with practical way of verification on the spot!

"I assume that if you reduce scope intensity and slow the horizontal trace very slow, you will also observe this on the scope"
OK of course I already did it and it does show just what I sayd.

"Can you attach some permanent magnets...."
That's interesting and I'll manage to make that test...consider that I'm taking it slow because I don't want to trigger a run-away.

Last question:
YES as always I'm preparing an engineering report that I'll share with all , so work-is-in-progress.

Roberto

[MeggerMan]

Hi Roberto,
Results look good!

Is the input power to drive the 60W lamp = 1.3A x 30V = 39 watts?
Can you show a photo of the coils.
Is the output/collector coil at 90 degrees to input coils to rule out mutual coupling?

Regards
Rob

[ronotte]

@Rob,

thanks for your encouragement.

As I sayd for the moment NO OU....but I'm near: What is still needed is to enhance the various processes, for example use narrower pulses: actually my electronics & wirings & stray capacitances do not allow the use of narrower pulses ( now 1 - 2 microseconds) so I will have to re-think the various possibilities.....

YES the architecture used is that suggested in the 3StackTPU by dfro: you can find in this thread all the details (90 degrees coils & mobiuses....included of course). The practical implementation has been done by Otto as also posted in this thread.

I strongly suggest you to replicate this project as it worths the needed time...

Roberto

[wattsup]

@Roberto

Again good work indeed. I am thinking that if the output electricity is taking the path of least resistance, it will fill up the capacitor before it fully lights the cold bulb. I am wondering as the bulb lights up, does the resistance of the bulb become lower so then the capacitor starts to discharge faster until it gets to a lower level, then it fills up again. This may be why the bulb is going at a 6-10hz cycle. Do your have any other output capacitor (smaller) to see if this will vary that result.

Again just the best right now, and the most important thing is this result seems to be directly under your control and not a random event, which for this stage is a major success.

[Gustav22]

Hi,
as far as I understand Ronotte is currently trying to improve the behavior of his CCs, so that all 3 of them auto-resonate at the desired harmonic frequencies.

In the meantime I am thinking about the behavior of the collector. As Otto has measured in a friend's lab, the 2 parallel conductors A and B of the collector's lamp wire (what Otto calls the 'mobius') have capacitance, i.e. they behave like 2 plates of a capacitor. This capacitor seems to get charged up by the 90? induction.
I think it is noteworthy, that none these 2 strands of th collector is grounded.
So this strange capacitor can retain it's charge and push it through the load in certain intervals, creating the characteristic frequency of Ronotte's output signal.

If one compares this to the L and N strands of our normal 220 V electricity supply lines (Europe), one realizes that the N is usually grounded. This prevents capacitance in our normal transmission lines, because one of the "capacitor plates" is connected to earth ground and thus can never get charge up. So the charging effect  - which we desire in the TPU's collector - is made impossible in the normal power transmission lines of our electricity grid.

I was thinking, that it might be a good idea to strengthen the capacitance in the collector windings and was looking for a way to do this. I think it can be done, if each strand (A or B) is in close contact with a maximum of strands of the other kind, so that the collector does not just have 2 capacitor plates (strand A & B) but instead resembles a capacitor with lots of alternating plates.
Such as A-B-A-B-A...
(a bit liker in a bifilar coil).

I read some posts in other threads and then tried to wind an example for a collector, which hopefully meets this criterion (lots of alternating plates). In a 3stack one would obviously need 3 of them. Here is a pic:


There is no hollow space or tube inside. It consists solely of lamp wire. (Yes, it should be done more neatly)

Maybe I will get a chance to visit my friend's lab again, then he can help me to test whether this makes a good collector, as I hope.

edit: also again braiding of the 2 lamp wire strands comes to mind as a possibility to augment inter-conductor-capacitance.

more edit (a completely different aspect):
Hi Roberto, I also noted:
The frequencies for a collector of 15" diameter are:
35 kHz, 70 kHz, and 245 kHz (1:2:7)
You use a collector of 5" diameter.
That is 1/3 of 15"
Now if the frequencies for a smaller TPU have to be proportionally faster, they can be calculated as follows:
35 kHz * 3 (because the circumference of your collector is 3 times smaller than a 15" collector) = 105 kHz

This corresponds to the number I read on the upper oscillator in your video
http://www.youtube.com/profile?user=ronotte
title: TPU V10-7
at 1 min:22 sec.


I gather you use:
15 kHz, 30 kHz and 105 kHz (1:2:7)

I think
105 kHz, 210 kHz, 735 kHz (1:2:7)
would be ideal, but I know this presents difficulties, because  at such high frequencies the mosfets are too slow to achieve the short pulses (5% duty-cycle) which you use ....
But maybe you may want try these frequencies with a 50% duty-cycle.

and then the carnival edit(ion):
@SM
Hi sir,
I would like to ask you, whether you ever made a unit with 3 collector strands instead of only 2, to get a 3-phase output?