SSTC tips for bridge / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

SSTC tips for bridge

1 2 next

Move Thread

LAN_403

Killa-X

Mon Apr 19 2010, 01:16AM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

Maybe I'm just doing something wrong that I'm not even noticing. To start, I'm upgrading my SSTC (Steves SSTC MINI) from a halfbridge to a fullbridge. Right now, on the gate drive board I have a 10uF capacitor on one side of the GDT primary. The mosfets have a 2.5ohm resistor on them. The GDT consist of 6 primary turns, and 4X 12 turn secondary. I have 15V zeners on the 4 mosfets, which are rated 500V 50A (roughly a 500W max power peak) hiperfet.

So the problem is the GDT. I'm just bearly able to get any output from my SSTC at 30V (maybe 2mm) and I wasnt sure if maybe it was the waveform.

In this picture, You can see green blips on top and bottom. These are just about exactly 15V. The squiggly line, I have no idea why that is there. The more capacitance I added to the Primary of the GDT, the lower it dropped. I already have 10uF, so I figured something is just...wrong.

If you need a larger image, Right click, copy link, and past. For any pictures below!
Wave1

Here I have the coil turned on, You can see it's at least getting feedback from my coil.
Wave2

Here the coil is on again, I turned the Volts / Div up, and this is the ~350khz waveform of the coil
Wave3

Does anyone have any tips for me on solving this problem? Spent maybe a week trying different capacitance..resistance..I changed up the GDT multiple times...Also, Is there a schematic that shows how to properly hook up the CT for feedback. I just have one connection of the CT to ground, The other connection goes to a resistor, then the 74HC14N. Some reason I recall you need a resistor and capacitor...But I wasn't sure, so I don't have one!!
Bridge design
Fbridge3

PCB board layout 60mm X 60mm board
Fb1

Kipmans

Mon Apr 19 2010, 11:38AM

Registered Member #91 Joined: Thu Feb 09 2006, 03:03PM
Location: The Netherlands
Posts: 45

I would first try to get a good looking gate drive waveform before trying anything else or powering up the coil. That is, just apply a square wave of the right frequency to the inputs of your gate drivers and start working from there.

Are you sure you are measuring correctly? Try measuring the gate drive waveform with your scope set to DC and not AC.

What kind of coupling capacitor are you using? 10 uF sounds pretty large to me, I've had good results using a 1 uF MKP X-2 capacitor (the X-2 rating is probably overkill, but I guess at least some kind of MKP pulse capacitor is recommended).

Also, you should be able to drive the MOSFETs using a GDT with a 1:1 ratio, as they only need >10 V to turn on.

It's a matter of debugging, really. Start measuring at the beginning (which would be the inputs of your gate drivers in this case) and if the signal gets messed up modify the offending part until everything looks good from start to finish. When you've got decent gate waveforms, you can start working on the feedback and all that.

HV Enthusiast

Mon Apr 19 2010, 01:45PM

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

I agree. Work on the gate waveform before doing anything else. Keep the MOSFETs connected to the GDT, but don't apply power to them and work on cleaning up your gate.

cjk2

Mon Apr 19 2010, 03:39PM

Registered Member #51 Joined: Thu Feb 09 2006, 04:17AM
Location:
Posts: 263

Perhaps one of your fets is shorted G-S? What frequency is your coil running at? 6 turns sounds a bit low for the goldmine cores I use atleast. I needed about 10 turns to get nice square waveforms at 300khz. Is the first picture your gate drive waveform? I would remove all but one FET from the output of your gate drive transformer, feed your driver a signal near the resonant frequency of your coil and take a look at the gate waveform again.

Killa-X

Mon Apr 19 2010, 07:13PM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

Will do. I know when I scoped the waveform coming out of the driver board TO the primary of the GDT, I had good waveforms. I'll change the 6 turns to 12 turns so its a 1:1 ratio.

As started earlier, The capacitors on the primary of the GDT is a 10uF. I'll stick it down to 1uF. Each gate has a 2.5ohm resistor (and I have a 10ohm mounted so I can switch easily). The capacitors on the bridge is a 400V 0.68uF (0.007ohm ESR) capacitor and 600V 6uF capacitor. I figured since my design is very compact, I wouldn't need 10uF. However, I do have 10uF worth of caps.

I'll post waveform pictures later tonight on what I get. Thanks!

Dr. Dark Current

Mon Apr 19 2010, 08:01PM

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

If the mini SSTC is not getting feedback, one of the gate driver just goes HIGH the other LOW on the interrupter pulse, and the GDT primary then "rings down" with the series capacitance, that's the damped waveform you see and its normal with no feedback to the antenna.

HV Enthusiast

Mon Apr 19 2010, 08:11PM

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

Also, those caps are WAAAYYYY overkill for the given application. With a simple SSTC, you can live with small board mounted electrolytics - no need for monster cans. Might save you some room. In fact, you really don't need the caps at all. Just feed half-rectified AC into the bridge. You won't really see much of a difference.

cjk2

Mon Apr 19 2010, 10:53PM

Registered Member #51 Joined: Thu Feb 09 2006, 04:17AM
Location:
Posts: 263

"Will do. I know when I scoped the waveform coming out of the driver board TO the primary of the GDT, I had good waveforms. I'll change the 6 turns to 12 turns so its a 1:1 ratio."

Yes, add turns to your GDT. Also, scope the output of the GDT with a FET connected. Looking at the input of the GDT won't tell you too much.

Killa-X

Mon Apr 19 2010, 11:24PM

Registered Member #1643 Joined: Mon Aug 18 2008, 06:10PM
Location:
Posts: 1039

C1, 10uF capacitor. I was told by kizmo that this capacitor does NOT have to be low ESR. Only the 1.36uF capacitor. (C2)

For C2 I have a 7mOhm ESR capacitor...Snubbers, worked great for the halfbridge coil.

C1, Since Kizmo told me any will do, I just banked some 2.5uF film caps together to equal 6uF. Just to verify, is this true?

Also I need either a larger core or use 2 cores..I tryed 5 wires twisted (Cat-5 cable) and it filled the core...

Arcstarter

Tue Apr 20 2010, 02:03AM

Registered Member #1225 Joined: Sat Jan 12 2008, 01:24AM
Location: Beaumont, Texas, USA
Posts: 2253

That capacitor needs low ESR (and low ESL is even more crucial) more than any other capacitor in the circuit. It should be very low ESR, and extremely close to the IGBTs. It is there to keep high voltage spikes out, it is very important. The less parasitics in your circuit, the smaller that capacitor can be.

I would suggest you use smaller heatsinks and have the mosfets closer, like Steve Ward's fullbridge. Like EVR said, if you want smoothed DC input, you do not need such big caps. But, if you want real CW, do not expect those mosfets to hold up. A fullbridge on doubled mains, running CW will pull an insane amount of current, like i said on MSN. The mosfets not only need to be higher current, but the wattage dissipated would be huge. Unless you use a bunch of turns on the primary to limit current...

1 2 next

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