Author Topic: SSTC build as per Steve Ward's Mini SSTC - Final Design (Read 25399 times)

Thunderstruck · « **Reply #20 on:** January 25, 2019, 01:29:31 PM »

GDT is connected, but there was no feedback signal to the antenna.
Do you think it will make a lot of difference ( feedback signal )
I will have a go at at that tomorrow...

Fumeaux · « **Reply #21 on:** January 25, 2019, 02:02:16 PM »

There is a huge difference between oscillating and not oscillating, as every switch (on to off, off to on) for the UCC's draws a bit of power. You don't need an extra feedback signal if you have a coil running. But, no feedback signal (real or fake) -> no oscillation -> no current draw

Thunderstruck · « **Reply #22 on:** January 25, 2019, 10:18:51 PM »

I am waiting for a proper toroid cores to arrive, Mads noticed that my current core is not suitable for frequencies needed. They should arrive next week.
I do not have any 24 awg wire for the windings, I used slightly thinner wire in my current GDT, do I need to compensate for this ? Maybe I should use slightly thicker wire ?
Im sure there is maths behind this

My secondary is almost finished, last polyurethane layer is drying as we speak.
For a primary coil, is it better to use solid core copper wire or multistrand is ok ?
My SRSGTC uses 6mm copper conductor for primary windings, that seems to work well....

Mads Barnkob · « **Reply #23 on:** January 28, 2019, 09:22:54 AM »

AWG24 is plenty for a GDT, it can continues carry 4A, your average current is lower than that. Thicker wire for higher current capability, but you will sacrifice some coupling for that. So it is best to use the smallest possible both for not using a overly large/expensive core just to fit the wire. I make almost all mine from Ethernet cable, which varies from 24-26AWG.

Skin effect due to high frequency makes the use of solid copper conductors non-ideal due to the most of it not carrying any current, but at least the thermal mass or the copper can help cool down between the pulses. Hollow tubing has the advantage that you can run water cooling through it.

So I almost always use tubing, its easy to bend, its cheap from the scrap yard, you can water cool it if necessary etc. For SSTC I use regular machine tool wire as there is no need for tuning taps and tight coupling is needed, so might as well use a wire with already added insulation on it.

Thunderstruck · « **Reply #24 on:** January 31, 2019, 12:28:02 AM »

Thanks Mads, my new cores just arrived, so soon I’ll be able to test my waveform with proper GDT ( I hope it will be good anyway

)
They are blue - that’s a good sign, and all the way from Czech Republic to Australia

Thunderstruck · « **Reply #25 on:** January 31, 2019, 11:40:32 AM »

Ok, I made a new GDT, using new cores and this time I used magnet wire. I scoped the gates of the Mosfets and I do not know what exactly is going on.
Signal does not look clean at all
I was expecting inverted signals, but they look the same.

I'm starting to think that there is something seriously wrong with my circuit....

Fumeaux · « **Reply #26 on:** January 31, 2019, 11:53:40 AM »

First of all I think the waveform is good enough to drive mosfets.

There might be one of these things wrong:

Maybe you have you scope connections switched
Maybe you have you have your GDT connections switched
Maybe your scope alines the two waveform (probably not this one)

I guess you could try if you have a short from + to -
(with a 12V supply and a lightbulb between supply and +)

If there is, you have your GDT connections switched.

Teravolt · « **Reply #27 on:** January 31, 2019, 06:19:26 PM »

keep in mind that if your probe grounds are to long fake induced voltages can apper on your scope diplay. close your ground to the tip and see if you can pick up noise around your circuit. other problems might arise from ground loops.

Thunderstruck · « **Reply #28 on:** January 31, 2019, 10:07:55 PM »

Thank you for your comments. Fumeaux, I will try the light globe test as you suggested.
The way I understand the GDT windings, first one is connected to UCC’s, second is connected to Mosfet 1 and third is connected to Mosfet 2 but it is swapped / twisted ( maybe I got this bit wrong ).
Thinking about it, I probably got it wrong right there....

Teravolt, I noticed that the waveform changes as I move the ground leads around, so you are probably right about picking up noise.

pentode · « **Reply #29 on:** February 01, 2019, 08:30:26 AM »

If the GDT connections are both in phase you will blow your mosfets
as both will be switched on together !

Fumeaux · « **Reply #30 on:** February 01, 2019, 09:19:21 AM »

Actually not, as the lightbulb is not a short curcuit.

Thunderstruck · « **Reply #31 on:** February 01, 2019, 11:10:49 AM »

Ok, I went back over the circut, and yes I swapped one of the GDT windings - the one that goes onto the gate of Q2. It is meant to be reversed, but I reversed it twice

.
I fixed that, now I got inverted signals on the gates !

Then I decided to connect it to my lab power supply instead of mains, and to my surprise nearby fluorescent tube STARTED LIGHTING UP !!!! YAY !!

I increased the voltage to about 30V and managed to get a small zap from the secondary breakout point.
I noticed that the current draw was about 2 Amps, and it was pulsing.
Forgot all about the run time, had it on for about 5 min continuously, before realising that I might fry something.
Mosfets got pretty hot Q1 more than Q2, fortunately I fitted a large heatsinks so they did not fry.

What would be next step before I connect it to my Variac ?

dexter · « **Reply #32 on:** February 01, 2019, 12:15:44 PM »

Quote from: Thunderstruck on February 01, 2019, 11:10:49 AM

What would be next step before I connect it to my Variac ?

in nearly all my antenna feedback SSTC's i had to reverse the polarity of the primary side of the GDT so that the primary coil is in phase with the secondary coil
If the phasing is wrong the output of the coil will be very small to the point i had to draw an arc in order to see it but as soon i fixed the phasing i got few cm streamers even with around 30V on the bridge.

Fumeaux · « **Reply #33 on:** February 01, 2019, 12:19:20 PM »

Does it run with antenna feedback already or still from a function generator?

In order to decrease the current draw you can:

decrease the ontime from the interrupter
increase the inductance of your primary

experimental: You can just add windings or build a new primary (with taps).
technical: Calculate the primary reactance and adjust it, to achive a wanted current draw. Mads discussed the powercalculation here https://highvoltageforum.net/index.php?topic=455.msg2763#msg2763

Teravolt · « **Reply #34 on:** February 01, 2019, 06:33:27 PM »

When you fire it up make shure tha any of that noise doesn't and couse ant shoot trough in the fets. Is there any dead time betwean each fet turning on?

Thunderstruck · « **Reply #35 on:** February 02, 2019, 08:32:42 AM »

@Dexter,
Seems like GDT is working fine, waveform on Fet gates is correct, so I think i got the GDT right this time.

@Fumeaux,
Function generator was not connected at all, that is probably why Q1 was heating more than Q2.
I connected the generator today, and - NOTHING, absolutely no activity. Then, I decreased the frequency to about 200kHz, and the nearby fluorescent tube lit up along the entire length, current draw went up as well, but no sparks.
Both Fets were equally warm. I’ll have to do some maths to work out the current draw as you suggested.

@Teravolt,
I would like to be able to answer your question, but I do not understand “dead time” I’ll google it and see if I can figure it out. This is my first SSTC, I’m OK with electronics, but nowhere near where I would like to be.
I built SRSGTC - mechanical switching, mechanics I understand very well

Fumeaux · « **Reply #36 on:** February 02, 2019, 08:51:33 AM »

Desdtime is the time between one mosfet turning off and the other on. As both fets are connected to the same gdt they both get the same drivesignal. As far as I am concerned, there is no deadtime, which makes the curcuit a bit less efficient, but thats not a problem. If you wanted to implement deadtimes, the driver would have to become quite complex.

And you can try to reverse your gdt connections. Then you should get an output.

Thunderstruck · « **Reply #37 on:** February 02, 2019, 11:12:46 AM »

Quote from: Fumeaux on February 02, 2019, 08:51:33 AM

Desdtime is the time between one mosfet turning off and the other on. As both fets are connected to the same gdt they both get the same drivesignal. As far as I am concerned, there is no deadtime, which makes the curcuit a bit less efficient, but thats not a problem. If you wanted to implement deadtimes, the driver would have to become quite complex.

And you can try to reverse your gdt connections. Then you should get an output.

I suspected exactly that, but wasn’t sure. In that case it would be easy to measure the “deadtime” on the scope, however in my case it wouldn’t be easy because waveform is so messy.
When you say reverse the GDT connections, do you mean rotate the GDT 180 degrees ?

Fumeaux · « **Reply #38 on:** February 02, 2019, 12:47:00 PM »

Quote from: Thunderstruck on February 02, 2019, 11:12:46 AM

I suspected exactly that, but wasn’t sure. In that case it would be easy to measure the “deadtime” on the scope, however in my case it wouldn’t be easy because waveform is so messy.
When you say reverse the GDT connections, do you mean rotate the GDT 180 degrees ?

Yes you can measure the deadtime that way.

I suspect that your feedback is out of phase. Therefore by switching the connections it could improve your output.

dexter · « **Reply #39 on:** February 02, 2019, 02:57:06 PM »

Quote from: Thunderstruck on February 02, 2019, 08:32:42 AM

@Dexter,
Seems like GDT is working fine, waveform on Fet gates is correct, so I think i got the GDT right this time.

The waveform on Fet's gates will be correct regardless if the phasing is not
Depending on how you connected the primary and secondary sides of the GDT (especially when the leads ate twisted) and the way you connected primary coil to the bridge you can end up in a situation where the signal picked up by the antenna is in anti-phase with what you have on primary coil.
When this happen you get just a small glow at the breakout point. This is easily countered by reversing the connections A and B

High Voltage Forum · « **Reply #39 on:** February 02, 2019, 02:57:06 PM »

Author Topic: SSTC build as per Steve Ward's Mini SSTC - Final Design (Read 25399 times)

High Voltage Forum

Recent Topics and Posts