GDT Problem / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

GDT Problem

Move Thread

LAN_403

ScotchTapeLord

Thu Jan 08 2009, 03:34AM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

Hello,
I am new to this stuff, but I've been studying the theory for a while now.
I decided to try to build a DRSSTC modeled after Richie Burnett's SSTC design.
I am using a TLC555 astable multivibrator with a pot that allows me to adjust the frequency from about 150KHz to 1MHz (just for fun, so I can try different secondaries and such). I have a bipolar push-pull circuit driven by the output (and inverse output) of the TLC555 that drives the GDTs that branch off into IRF740's. I decided to test the circuit by connecting the mosfets in series and hooking the first one's emitter to the positive terminal of a 6V battery, then testing the voltage from the second's collector to the negative terminal. When the circuit is off, there is 0V, as the mosfets are completely off. But when I turn it on I actually get a voltage higher than when I test the battery alone! At first I figured I hadn't hooked up the GDT outputs to the mosfets properly so that one would be off when the other was on, so I tried switching the gate drive outputs, but still got the same result. I was only operating it at about 150KHz. I'm guessing the mosfets aren't turning off properly or at all. Yet even that doesn't explain the rise in voltage (to me)! If the mosfets are being turned off/on by inverse waveforms, I should never have any voltage with my test, correct? considering I hooked them up both ways, one should have resulted in 0V.

The only problem I can think of would be the freewheeling diodes I omitted in the push-pull circuit. Am I generating extra oscillation without them that doesn't allow the mosfets gates to turn off? I have UF diodes on the way but I have the feeling it won't solve the problem.

HV Enthusiast

Thu Jan 08 2009, 04:19AM

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

Don't use Richie's circuit for a DRSSTC. Check out simple DRSSTC designs that Steve Ward has developed. Richie's designs, as far as I know are based on standard SSTCs and high frequency (class-e for example) CW designs.

You want a design which autotunes (current feedback).

That design you describe above seem pretty hokey for a DRSSTC plus an IRF740 is no good for a DRSSTC as well.

Go visit Steve Ward's website.

ScotchTapeLord

Thu Jan 08 2009, 10:39PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

Oh... pardon my terminology- I saw the H bridge and a capacitor with the primary and thought it was a DRSSTC (just DC blocking here?). So I guess I'm just making a SSTC.

But since I haven't started the primary it doesn't really matter... I just want an H bridge driver whose frequency I can control, using IRF740.

In any case, the problem is that my MOSFETs aren't turning off and on in sync... which would strike me as impossible if they are run off the same GDT, especially if I tried connecting them both ways. And why is it raising the voltage of the battery I'm using to test it...? I'm guessing the UF freewheel diodes in the push pull circuit will fix this... but what else could be my problem? I will upload a schematic if I'm not being clear enough...

HV Enthusiast

Fri Jan 09 2009, 01:31AM

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

Yes, the series capacitor is just a capacitor to ensure you don't create a flux walking condition (imbalance) on the core, although thats pretty tough anyways with an aircore. But it does keep DC (offset) out of the core which could be a problem for the driver.

ScotchTapeLord

Sun Jan 11 2009, 04:24PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

I just got an oscilloscope and I found one problem... my inverse square wave wasn't cooperating. When I got it to, I blew my TLC555 and my voltage regulator got really hot (good thing I have it on a heatsink). I made a new circuit in which I can control the frequency (now 50K to 500KHz) and pulse width. Also, instead of feeding the push-pull circuit directly from the TLC555 and the inverse, I am feeding it with the inverse and an inverse circuit fed from the original inverse circuit. It makes one waveform a bit curvy on one side but I don't think it should matter so much. I'm just afraid to hook it up again to the push-pull part for fear of damaging my regulator, because it obviously has a problem, though it's modeled after exactly after Richie Burnett's design. Has anyone ever done a SSTC with a TLC555?

Dr. Dark Current

Sun Jan 11 2009, 04:56PM

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

I wouldn't recommend a 555 because it usually cannot generate 50% duty cycle, which results in gate waveform problem (one polarity of higher voltage than the other one).

For a reliable design (fixed-frequency), use a CD4047 or other oscillator with 50% duty cycle. Then use the outputs (Q and Q inverted) to drive two gate drive chips, which you can get as free samples if you must.
This produces by far the cleanest drive waveform, cleaner than any simple discrete design.

Edit: You can also put small RCD combos on the inputs of the GD chips to slow down the turn-on with the purpose of adding some dead time.

ScotchTapeLord

Sun Jan 11 2009, 08:48PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

I have two potentiometers in series with the astable multivibrator's standard 1k resistors, and I am using a fast diode to bypass the second resistor when the capacitor is charging. I have complete control over frequency and pulse width, and I have checked it against my oscilloscope. Will this approach solve the problem, as long as I am very careful when adjusting the potentiometers and keep an eye on duty cycle? I can get it pretty close to 50%. I like the idea of variable frequency, seeing as I have a couple secondaries I'd like to test. Not to mention... I plan to stick with my original designs, more or less, because I already have all the stuff for it. I will vie for cleanliness after being disappointed with the output once it's done. I have a feeling that for a first try it will turn out decent. My equipment should allow for about 1KW of power, but I bet that's pushing it. Is an interrupter recommended for this sort of design?

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