My SSTC (and questions) / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

My SSTC (and questions)

Move Thread

LAN_403

Zum Beispiel

Tue Mar 27 2007, 06:57PM

Registered Member #514 Joined: Sun Feb 11 2007, 12:27AM
Location: Somewhere in Pirkanmaa, Finland
Posts: 295

So I finally managed to gather the courage and took a shot at building a SSTC.

Its pretty much the same as Steve Wards Micro SSTC, I just use a UCC37322 as the driver and a crappy to-220 cased 120V Fet instead of an IRFP460, and the secondarys dimensions are different. Also, I didn't have any germanium diodes, so I just used some silicon signal diodes for clamping the antenna. I use 4 primary turns

It works just great! On first light with 12V in I got some 1.5 -2cm streamers (hard to measure) and when I upped it to 20V input, the streamers got a little longer(3cm, or a bit more) and hotter. The only problem is, now I want a bigger one

I'll have to try and get some pics of it.

Now, to the guestions:

In the original schematic Steve has a push-button to inject a high signal into the gatedriver to get it oscillating. However, when I turn the power on, it starts up by itself, and very reliably too (every time). Is this normal? I know that I'm using an UCC instead of a TC4420 to drive the fet, but is this normal with UCCs?

Then about grounding the secondary:

When I made my first test on last saturday, I just breadboarded the thing and sucked power from a small lead-acid battery. I just grounded the coil to a small copper plate under the secondary and it worked fine. Then when I build a more "neat" version of it, I used the same grounding method. The coil makes nice streamers, but I get RF-burns if I touch any metal parts on it. I also think I lost a few mosfets due to this (no big loss, they're only 0.50â‚¬ or something a pop)

When I took some wire and grounded the coil to earth (as in the earth in my garden), there was no voltage build up on the metal parts (as expected). The streamers didn't change in size, but the discharge got more quiet.

Now, I know you guys use the mains ground for multi-kW systems, but I'm a bit paranoid about it. I don't want to accidently fry my (or someone elses

) computer, or something else, so are there any precautions I should make before I ground it to mains?

And one last question, about MOSFET failure modes:

The FETs that got killed had all their pins shorted together (gate-drain-source, all a dead short). Does anyone have any idea on what killed them? I guess it was probably the high voltage build up that killed them, but what do you guys think?

They all died when the coil was unable to break out, so I guess that would cause more voltage to build up on the metal parts (heatsink, power wires), and ultimately destroying the fet. However, I havent lost a single driver IC (yet), maybe it was just luck, or something else killed the poor fets.
Maybe they are just too crappy to work on a coil

But what do I know. Any ideas?

I hope my long post didn't bore you too much, but this is my first coil, so I have a lot of questions. Thanks for reading this.

Marko

Tue Mar 27 2007, 07:20PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi!

Steve's micro SSTC tends to operate simply as a sort of flyback converter and at any higher input voltages mosfet will die from massive avalanching caused by heavy voltage spikes.
You could probably put a TVS or snubber across the mosfet, but it just moves heat from one place to another.

A mosfet with higher voltage rating surely helps, and I never actually used anything below 500V in this app.

You ofcourse also have a lot of switching and ohmic conduction loss.

If you intend to get anything much more than 3-4cm of spark you'l need to switch either to a bridge or class E drive depending on your resonant frequency?

How big is your secondary? Have you calculated/measured the Fres yet?

Zum Beispiel

Wed Mar 28 2007, 11:18AM

Registered Member #514 Joined: Sun Feb 11 2007, 12:27AM
Location: Somewhere in Pirkanmaa, Finland
Posts: 295

The secondary is 13cm high and wound with 28-30AWG (I'm not sure what gauge it is, I took it off a fan from a microwave oven). The former is grey 5cm diameter PVC pipe. I haven't calculated the resonant frequency yet, but I guess it's pretty high. I would have made it on a larger diameter pipe, but I didn't have any. Seems to work well, though.

I haven't had any MOSFET failures after I grounded the coil. It's grounded to a steel water pipe now and works fine. I'm still too much of a coward to ground it mains ground, though.

I know I won't be able to get much better results with single transistor drive. I just build this coil because I was bored and didn't have anything better to do over the weekend (and I wanted to see if I could do it).

When I get some money and spare time, I'm probably going to build a bigger coil with half-bridge drive, that will be a much better performer. This one was just for some quick fun.

Marko

Thu Mar 29 2007, 09:25AM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

That's like 1,3Mhz without a topload. I think you could improve your coil for much if you upgraded it to class E as on steve's page... got oscilloscope?
MAke sure to use a 500V mosfet, IRFP450 is fine.

If it is done properly you should be able to get 15cm or so sparks from that coil.

Regarding the coil spontaneously starting, your driver chip may be turning on spontaneously for some reason, either because you pulled the input up or from noise.
Different resistor arrangement can fix it if it bothers you.

Brett Miller

Sun Apr 01 2007, 05:45PM

Registered Member #593 Joined: Tue Mar 20 2007, 12:32AM
Location:
Posts: 50

Zb,

Hello. I recently built a similar version of Steve's MicroSSTC. It was my first SSTC as well. I went ahead and built and interrupter from the beginning, and also made the 12v power supply for the UCC and the 555 separate from the MOSFET supply in order to have more options for input power. Actually I didn't have the 555 noise generator in the circuit at first, and apparently because of that I had trouble getting oscillation to start, even with a long antenna placed close to the secondary. In any case, as soon as I added the 555, I got oscilation and the expected 1.5 - 2.0" (5cm) sparks without the interrupter.

At that time I was using an IRFP460, 8" winding length 3.5" secondary (20cm by 8cm) PVC, wound with 30awg wire. Primary is 3 turns 16awg on a 4" (10cm) pvc form. Cap is a Panasonic poly .0033uf. The bottom turn of the primary is about even with the bottom turn of the secondary. Most of the time I have powered this coil using a 30v transformer full wave rectified and smoothed with a 6200uf 100v electrolytic, which gives around 48.5v with no load attached.

I few weeks ago I tried an experiment where I flipped the primary former in reverse to raise the primary about an inch and a half up in order to increase coupling. I noticed the pulled arcs (in cw mode) got quite a bit hotter and I could now pull them nearly 3" from the bolt on top of the secondary. However, I should have reversed this mod when I decided to run the system off line through a full wave rectifier and 460uf cap. I ended up frying the 460 (it failed open) and blowing one of the legs off. Remember, if you are going to run this off mains, be sure to fuse the input in case the MOSFET fails short circuit.

Anyway, I had an IRFP250 around from my ZVS stuff, so I sub'd it in and everything worked great. Same sparks with the 30v transformer. I now have an IRFP260N (200v 50A) and get about 3 inch streamers to air with the remote interrupter I built. I need to tweak the cap for Class E operation, but my main most reliable scope is an LCD firmware based thing that I don't think would fare too well around efields. I have a TEK, but it has a horizontal sync glitch that makes it problematic to take simple measurements. I'm pretty satisfied with the sparks anyway, plus this coil is a lot of fun with the interrupter...plus, I too am working on my first half bridge coil...it should be finished in a couple days. Zb, you ought to build the interruter and hold a flourescent tube near the coil while sweeping the PRF and ON time. It's an impressive display.

I guess, one of my main points is that if you keep the voltage under 70 volts or so (i've had no failures of any mosfets up to that voltage with my specs) you can use quite a few different fets for this coil.

Hope this helped somehow...

-Brett

Zum Beispiel

Sun Apr 01 2007, 06:30PM

Registered Member #514 Joined: Sun Feb 11 2007, 12:27AM
Location: Somewhere in Pirkanmaa, Finland
Posts: 295

Thanks for the reply, Brett. BTW, are you on 120V or 230V mains? Cause with 230V input the 500V fet would surely die from the huge voltage spikes.

So anyway, I got some quick pictures with my friends camera. In the pictures I have only two primary turns, which made the arcs hotter, but shorter.

I also tried running it with an interrupter, and got some very nice sparks. Unfortunately the mosfet exploded after a while. Using the interrupter must cause larger spikes than CW operation. I'll have to get some higher voltage, 500V or so, FETs when I order some more parts. Then we'll see what this little thing can do

Firkragg, while Class-E sounds tempting, I'm not sure if I'm ready to go down that route. I do have a scope, but it's about 30 years old uncalibrated piece of junk. I think I'll save class E for the future.

Yeah, anyway, pics.

Brett Miller

Mon Apr 02 2007, 02:11AM

Registered Member #593 Joined: Tue Mar 20 2007, 12:32AM
Location:
Posts: 50

Zb,

I am on 120v mains, still, the voltage spikes can get high enough to blow a 500v MOSFET. During an email discussion with Steve about Class E operation of this coil, he mentioned that you will notice as you try larger Capacitance, the voltage spikes diminish, but you reach a point eventually where output will diminish as a result of too much C. So (for class E) you would want to strike a balance of the best performance you can get while keeping the voltage spikes under 500v (or whatever the Vds of the MOSFET you are using). As seen on the scope, the right capacitance would give one voltage "bump" before the mosfet turns back on.

Of course, I haven't experienced this empirically yet myself. I am in sort of the same situation with a scope from the 1970's. Of course, no worries about getting it too close to the e-field either. I am in the middle of doing the final testing on my halfbridge SSTC this week, so I'm pretty excited about that right now. But I am planning on getting those 15cm Class E sparks eventually!

One thing of note...whether or not you are operating proper class E or not, it would make sense to keep your capacitance larger whenever you are worried about possibly losing a MOSFET...also, I usually run my microsstc with the interrupter at around 30-40 pps or so most of the time, with varying on-times. The FET mostly stays ice cold with the interruptor, even if I am drawing sparks. CW power arcs will heat it up rather fast. I would guess that if you went for a larger cap your problems with the interruptor might disappear.

-Brett

Steve Ward

Mon Apr 02 2007, 02:41AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

One thing of note...whether or not you are operating proper class E or not, it would make sense to keep your capacitance larger whenever you are worried about possibly losing a MOSFET.

Well, not quite. If the cap is too large, it remains charged to some degree. When the MOSFET switches back on, the capacitor is suddenly shorted through the MOSFET. If the capacitance is large, or the voltage high, this relates to a lot of energy dissipated in the fet on every cycle! So while the chances of over-volting the fet are reduced, you could still pop it from simply too much heat dissipation (and it can be a transient thing, where the heatsink was still cool since the internal heat didnt escape fast enough).

When guessing at the capacitor value, start with low input voltages and actually check the Vds waveform for proper pulse shape (or at least something close).

Brett Miller

Tue Apr 03 2007, 02:38AM

Registered Member #593 Joined: Tue Mar 20 2007, 12:32AM
Location:
Posts: 50

Steve,

Thanks for the clarification on that issue. There are some things I need to study up on regarding Class E operation, also some time with my own MicroSSTC and my Tek would be in order.

Zb,

Another thing I forgot to say before...no need to apologize for this post. I found it anything but boring, since building my copies of Steve's Micro and Mini SSTC's are my current projects. I really enjoyed reading about someone else just getting into building solid state coils.

-Brett

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