Self-resonating micro SSTC oscillating in wrong mode? / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Self-resonating micro SSTC oscillating in wrong mode?

first previous 2 3 4 5

Move Thread

LAN_403

Dr. Shark

Thu Mar 02 2006, 08:50AM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Last night I had another go at getting the PLL SSTC thingie to work.

All the electronics seem to be fine - the PLL locks in and tracks the frequency if I vary the resonance by bringing by hand near the coil, the GD circuit stays cool and produces a nice square wave on the gates. But: I get no sparks. I tried powering the primary off 24V and 50V, and in both cases I only get a faint spark if I directly touch the breakout point or the secondary. Performance is even worse than with the manually tuned driver, where at least I could get some sparks for an instant when sweeping through the frequency.

Is there anything obvious that I am missing? This is Steves PLL-SSTC Mark I with the only difference that I used the CMOS 4046 which runs off 5V so I can use this for small, high f_res coils. Gate drive is 4421 4422 push-pull with 20-turn trifilar GDT. I tried only 10 primary turns for a 1:2 step-up, but this seems to cause core saturation.
The power electronics is back to a half-bridge, as to keep it simple.

My onw thoughts are:
1) Do I need decoupling caps across the bridge? I think not, as I am running CW and I have the capacitive divider in place anyway.
2) Am I not saturating the IGBTs gates after all, with my 12V drive? Is there any way I could test this? If I do a differential measurement across the primary, I get a semi-decent square-wave, which makes me think that the IGBTs are switching correctly.

vasil

Thu Mar 02 2006, 02:22PM

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

Didn't you forget to enable the PLL, after you found the resonant frequency?
20 turns on GDT seems a lot, I would go with 12-16.
How big is the decoupling cap on the GDT primary?
What schema do you use? This?

http://scopeboy.com/tesla/dw-isstc/dirty_pic.gif

When i made first the PLL setup I used separate resistors for R7 and R6, and it was difficult to get the right value. Then I switched to a dual 100 kohms pot and everything was ok. You have to disable the PLL and find the resonant frequency (maximum brighting in a neon tube), then enable the PLL and use the phase adjust potentiometer for the best output.

It is possible to need more current on gates. Can the TC44xx can be paralleled as the UCCs? If not, use a 100 uF electrolitic on the rails of the driver chips as a charger pump.

Dr. Shark

Thu Mar 02 2006, 02:44PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Thanks for your input Vasil. I am using the circuit from

with the output going directly to the GD chips, no interruptor or anything. My coil resonates at 350kHz (calculated) which means that I should be using less turns on the GDT than most people, but if I do the gate driver chips just get hot (and I mean really hot, they sit on a big heatsink). BTW they are good for 9A peak, so I don't think I need to parallel several of them for my small TO247 IGBTs.

I dont know what you mean, I should first find the f_res and then switch on the PLL. The PLL is always on in my circuit. It seems to lock in fine, so is there anything wrong with this?

vasil

Thu Mar 02 2006, 02:54PM

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

I mean, I built the same schema as you and didnt have succes until I went with a resonant cap in serie with the primary.

After switching to the last PLL schema of Steve (see above), the problems dissapear. With separate R7, R6, I always have problems (no scope though) until i turned for the dual pot.

Yes, in your variant, the PLL is always ON.

I think you are the first trying PLL with IGBTs in CW. Do you have the same results runing in half wave rectifyed?

Dr. Shark

Thu Mar 02 2006, 03:34PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

Um, is there anything wrong doing the CW-PLL thing with IGBTs? I know I could use MOSFETs, but I happen to have just IGBTs.
I'd love to go interrupted, but the GD chips I am using do not have an enable input, so I don't know how I could actually implement that. Also, I don't have any logic chips like flip-flops, and I'd rather not get involved with this digital stuff.

The 4046 data sheet says R7 and R6 set the center frequency and and frequency offset, but never mentions what this offset acutally is. Since this is the only real difference between the two schematics it seems to be critical. I guess I should get some dual-gang pots.

Steve Conner

Thu Mar 02 2006, 03:52PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I think you ought to try grounding the aluminium plate to the mains ground. Otherwise the antenna will tend to pick up equal and opposite signals from the topload and the plate, and it will turn out that the strongest mode is the half-wave one with the sparks coming from the middle of the coil.

Dr. Shark

Thu Mar 02 2006, 05:06PM

Registered Member #75 Joined: Thu Feb 09 2006, 09:30AM
Location: Montana, USA
Posts: 711

I am actually doing that (through an oscilloscope ground clip), and the maximum output is from the top. If you dont spot any obvious flaws (it's your design after all) I'll just try around a bit more.

first previous 2 3 4 5

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