Steven Mark`s TPU

[AhuraMazda]

In the missing post Otto said he connected his Rodin coil.
I don't know if that is still the case.

[otto]

Hello all,

@Earl

my "output" strands with the bulb connected are in paralel connected.
If I would connect my "input" coils with the pulses in a Mobius way, the cap should have a lower value because when a 2 strand wire is connected in the Mobius way the capacitance of such a wire is dramatically INCREASED!!
Its worth a try.

@EM

thanks, its the right drawing.

The cap in paralel: first I used a 100nF/400V cap but it was not good. This cap was soooo hot that I decided to use a 47nF/1kV cap and it was OK.

This 5 - 10Hz vibrations I see on my scope. 7,8 Hz vibrations?? Yes, I think so.
But only with the 245kHz. The other 2 frequencies that we have to use just increases the sine waves that I see at the 245kHz.

My power supply is a profi made with an 60V/5A output. The current can be regulated but Im using the maximal 5A. Its not OU but something with this setup is hmmm....strange. Why this 245kHz?? Why are sine waves formed? Little sine waves?

@Stefan

I used 12V from my power supply. The current was 1,3A, or so. When I tuned my frequency nearer to the 245kHz I saw that the voltage/current rised more and more. Then, near 245kHz, the voltage rised to 40 - 60V and the current to 2 - 4A. I managed to have my frequency stable and the bulb was almost at full brightness. But after maybe 20 - 30 seconds the effect dissapeares. Like I charged something very fast and then I have a slow discharge. Maybe my cap is discharging? With a bigger cap I had the same effect.

It can happen that this effect doesnt start so I rised the voltage from my power supply to 14V and then the effect started again.

Yes, a picture will be fine.

As I wanted to control this beast I wound a control coil with 57 turns all around the collector and pulsed this control with an unknown frequency. I didnt connect the control to the power supply!! It worked. Not sooo good but..Im learning.

The main problem is when I have to use all my 3 frequencies. I found NO WAY to control the beast. Blowing my MOSFETs is normal. This is the main reason for building tube oscillators because MOSFETs are cheap but when you blow in 1 day 2 of them then you have to decide: to buy every few days a lot of MOSFETs or to build tube oscillators. So Im building tube oscillators.

@AM

Yes, I made the same with Rodin coil collectors. Fine and nice but its not the time to play with such collectors. They are dangerous and Im not able to control them. I hope that with tubes ( they dont blow??? ) I can continue to have fun. With MOSFETs, no way anymore.

Otto

[BEP]

I now doubt coaxial cable has anything to do with a TPU but you can see a pronounced version of this effect by using coax.

The best way I've found is using cheap TV coax that has a copper coated steel center conductor and a continuous foil shield. Use the shield as the primary and the center conductor as the secondary. Or the reverse. Even more interesting is running two opposing signals into the primary.
The coax thing was a learning experience for me and the results were the same as Otto's even with 245kHz.

@Otto,

I understand the losses on MOSFETs. I've also learned that @ronotte has the best fix for that - opto or transformer isolation between the MOSFET gates and the mains and a current limiting resistor on the source lead. No amount of freewheeling diodes and caps will stop the failures at that point.

BEP

[Funkydcat]

@ Otto,

        I recently came across an old manual for an HP 200CD oscillator, and much to my surprise it was all TUBES. It has a range 5Hz to 600KHz sine wave. They sell relatively cheap on eBay. Just thought this might be helpful, as you said you were having difficulties building Tube oscillators from scratch . I'll post the schematic later.

Cheers,

Julian

[ronotte]

@Earl,

About self-sustaining TPU V10-7 operation:

I equipped all the 3 power Mosfets with Gate-to-Source 10 KOhm resistors: NO WAY to stop auto-oscillations (once triggered) that now happens to be at about 100KHz with all 3 mosfets contributing.

The weird thing is that even if the consumpcion is only 1.3 A at 30V (PS)...the 3 heath-sinks gets soon hot. If I try the same PS consumpcion but actually 'driving' the beast with 3 harmonic freqs...well the heath sinks stay cool......

One possible explanation is that when I drive it is with 2 microsec pulses while in auto-sustaining mode it does operate also in linear region  (?)...is it possible? 

Remember that in both cases the mosfets do operate in avalanche mode and do light 60W lamp even if at different level as while in narrow pulse mode to obtain the same light it's necessary to operate at 40-60V.

Roberto