SSTC Feedback controller / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

SSTC Feedback controller

1 2 next

Move Thread

LAN_403

Dr. Dark Current

Thu Jan 01 2009, 08:40PM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Hey all,
in my first SSTC I used secondary base current feedback and blown a handful of FETs (mainly at ground strikes). I then upgraded to antenna, blowing just a few of them. Finally I implemented a start-up osc for the antenna feedback, this worked pretty good (though i didn't do much ground strikes) until I finally exploded my last pair of FETs, which went with a loud and proud Bang.

I'm wondering if anyone has experimented with various types of feedback/start-up oscs and found out which gives the least FET heating/ failure rate?

I'm also particularly interested in using a simple fixed-frequency oscillator, as this can be tuned precisely to get the biggest sparks (no feeedback delay), prevents failures at ground strikes/arcing/even with complete secondary shortage, and mis-tuning cannot blow the FETs as the primary current is highest at resonance (of course "mistuning" to extremely low frequency can blow the FETs from primary reactance but nobody will do that).

Of course then there's the PLL but I (and many others I believe) don't fully understand it and tuning it seems to be a PITA for a lot of people.

What do you think?

LithiumLord

Thu Jan 01 2009, 10:57PM

Registered Member #1739 Joined: Fri Oct 03 2008, 10:05AM
Location: Moscow, Russia
Posts: 261

Well, this is a pretty interesting question. Antenna is a sort of a sledgehammer approach. Unless an arc is sufficient to overload your drive circuit and the field is inducing enough voltage on the antenna it gives a perfect switching, I had only a couple of failures with it, all related with a short ground path (also had an interference issue a few times however, I'm redoing the casing for the drive electronics now as it all was a kind of a temporary version, will see if under any circuimstances the issue could be recreated on the new, perfectly fitted, variant).
The CT needs current, if it's not high enough some various effects may kick in, like delays etc.
As for the open-loop version, try out a PLL - I tried 494 as a PWM and that's a real pain in the rear! Needs retuning all the time whenever you move the coil around, still fails on short ground arcs sometimes, also an untuned bridge will definitely fail - the ZCS will not be applicable for the secondary-induced component any more. Also keep in mind that long sparks that exceed the secondary length cause enough detune for the coil even probably not being able to break out if set on such a frequency in the interrupted mode.

As a side thought - a current shunt! I'll try it out once done with everything else, you can make it using a higher-current zener, a few UFs to act as zeners or a PNP with a given bias voltage applied to the base resistor (whenever the voltage on the secondary tail exceeds the threshold value, the base current is amplified in order to produce the necessary load).

Wolfram

Thu Jan 01 2009, 11:31PM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

I've had best results with antenna feedback and startup oscillator as well. Have never blown FETs due to anything but stupidity and mistakes.

Anders M.

HV Enthusiast

Fri Jan 02 2009, 01:08AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

you want to use primary feedback. secondary feedback, as has been shown, is no good and does not guarantee soft-switching.
antennas are just too hokey IMHO and too finicky.

Frosty90

Fri Jan 02 2009, 01:59AM

Registered Member #1617 Joined: Fri Aug 01 2008, 07:31AM
Location: Adelaide, South Australia
Posts: 139

I've had moderate success with CT base feedback and a startup oscillator, but it needs a bit of fiddling with the coupling resistor values to get it working right. I was using a 555 for the startup oscilator. But like you said, ground strikes arent good for it.

Edit:
And like lithiumlord said, the CT needs current, ive found with my latest coil, that it only works properly beyond a certain power level, other wise the feedback doesent always kick in, and you get a sort of 'spluttering' effect, but above about 800w, it seems to run more smoothly; maybe some experimentation with different CT turns ratio? I've just been using 1:60, because everyone else was.

LithiumLord

Fri Jan 02 2009, 05:05AM

Registered Member #1739 Joined: Fri Oct 03 2008, 10:05AM
Location: Moscow, Russia
Posts: 261

With high CT ratio it may mess things up due to the leakage inductance imho, also a secondary-rigged CT may behave "strange" on startup. If it's all about DRSSTCs, it feels clear how to make it right - but how to deal with generic ISSTCs? So far I see nothing better then a current shunt (or antenna, again :) ).
Ps tried a CT+comparator trick, the idea was making a 0-12v CT range with 5v midpoint, that makes the system very sensitive - even any stray static affecting your secondary triggers the oscillations. The bad side was - I had a strange effect of abnormal delay, giving a highly asymmetric signal on the bridge. Will take a closer look into this some time later, so far I want to play around the shunt concept instead.

Frosty90

Fri Jan 02 2009, 07:02AM

Registered Member #1617 Joined: Fri Aug 01 2008, 07:31AM
Location: Adelaide, South Australia
Posts: 139

What about a CT or shunt, with maybe some sort of shift register running at many mhz (adjustable) to add a variable 'phase' (delay)? its probably a rediculous idea but wouldnt this give some control over the phase/delay with out the need for complex pll circuits? Or are there any other ways to easily adjust the phase with secondary feedback? you could easily compensate for the leakage inductance of a CT this way

Cheers,
Jesse

vasil

Fri Jan 02 2009, 08:50AM

Registered Member #229 Joined: Tue Feb 21 2006, 07:33PM
Location: Romania
Posts: 506

I used antenna a lot and I was very pleased with it, but the results depends by the position of it (you have to experiment until you get the best position). I then choosed the CT feedback, just to experiment with it, but only for DRSSTCs (you need more power indeed). It works OK. I am not allowing ground strikes...well it strike to the targets, but not grounded. I can not use high BPS because I have only little BUP IGBTs here and they are fragile. I can not use use primary feedback because the lack of some of components board and a scope.

But if we use a primary feedback (at high or low or intermediary pole) what would be the difference versus a fixed primary frequency? With an fixed frequency oscillator the primary will stay on the frequency, even for secondary strikes to ground. I remember that it was the Jimmy Hynes approach for the first DRSSTC and he did not report many IGBT failures.

Dr. Dark Current

Fri Jan 02 2009, 11:11AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

After all, it would seem that a SSTC resembles a LLC/LCLR network. Does this seem about right? If so, are there the nasty phase shifts around the resonant frequency as with the induction heater?

HV Enthusiast

Fri Jan 02 2009, 04:28PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

primary feedback is the only practical way to sample your current to ensure proper switching 100% of the time. with secondary base feedback, you may find the current feedback "mode-hopping" or jumping between the two split frequencies of the coupled system or even to other harmonics of these as well.

i use CST-206 type current transformers which are small and cheap for this purpose and they are good for at least 500A peak current on the primary at 100-300kHz.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob