First SSTC / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

First SSTC

Move Thread

LAN_403

dan

Wed Mar 15 2006, 05:33PM

Registered Member #223 Joined: Mon Feb 20 2006, 06:42PM
Location: Ottawa, Ontario, Canada
Posts: 125

I managed to put together my first SSTC last night. It used an s494p(same as TL494) I pulled from a PC PSU. It has two of its output transistors wired in a push-pull configuration. The output is connected to the gate of a single IRF740 mosfet. The fet pulsed 12v into a 10 turn primary wound on a 1.5'' PVC pipe. The secondary was my 1200 turn 12'' tall 3 1/4'' dia place over the primary. Fiddling with the 50k pot I managed to get 1/4'' tiny streamers. However I had to put my hand 6'' or so away from the output to achieve this. I don't know the exact frequency since I don't have an oscilloscope or frequency counter. Obviously many improvements need to be made but I'm a little reluctant to work directly off of mains yet. I only have 6 IRF740's and am afraid to blow any of them but failures are probably enviable.

However I have a few questions. I have made a GDT out of a toroidal core also pulled from the PC PSU. It uses 10:20:20 wound with 30awg magnet wire. To improve the design above I had the idea of driving the GDT directly from the s494p in a half bridge configuration. At very low frequencies I can get the two LEDs connect to the GDT to blink back and forth very dimly. At higher frequencies the LEDs get brighter and brighter. However I'm not sure if driving the GDT directly off of the PWM is a good idea or not since it can only source 200ma of current. However with my LED test the PWM doesn't even get warm so I guess it's ok. I'm trying to keep part count to a minimum since I am practically on a $0 budget. And not having an O-scope really hurts when it comes to really high frequency stuff.

One thing I notice is that with Ct = 220pF and just near the end of the 50k pot(last 5% or so) the LED's connected to the GDT dim suddenly to almost nothing. Is this and indication of a bad GDT or is it that LED's don't work well above a sertian frequency.

vasil

Wed Mar 15 2006, 09:59PM

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

1.5 inches diam primary and 3 inches diam secondary? why? Put your turns at the base of the secondary (wind them at the secondary bottom).
I didnt see anywhere, the chip output directly in the GDT without push pull transistors. You have to turn completely the FET ON when using higher voltage input (mains). At low voltage can work, but blown FET danger at higher voltages.
Did you turn the pot to the end while it sparking? If the spark is still on top, there are not GDT problems.
What pot? freq? duty cycle?

dan

Wed Mar 15 2006, 10:51PM

Registered Member #223 Joined: Mon Feb 20 2006, 06:42PM
Location: Ottawa, Ontario, Canada
Posts: 125

I have since wound the primary on the outside of the secondary. Itâ€™s just that I had that primary around from another much smaller VTTC. The 494 PWM has internal transistors with both their Collectors and Emitters accessible from pins on the IC. (Pins 8&11 are the collectors and pins 9&10 are the emitters). I thought I was being cleaver about just adding external capacitors to create a half bridge to drive the GDT (see attachment - attempt 1). However it appears that the PWM doesn't like this and it becomes unstable. All the bjt transistors I have crap out at a few 10's kHz (most my components are pulls from old equipment).. In desperation I even tried to make a buffer with lm741 op-amp(see attachment - attempt 2) as voltage follower but this also craps out just above audible frequencies.. I guess Iâ€™m just going to have to wait until the right components come my way.

oh well I guess it's back to my VTTC.. Meh VTTC's are way cooler anyway..

Simplified diagram:

...

Thu Mar 16 2006, 12:10AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

I would look into getting a new gdt. I am almost positive that the one you have is an iron powder one, which isn't good for this high frequency work... Might be something to look into...

HV Enthusiast

Thu Mar 16 2006, 01:36AM

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

... wrote ...

I would look into getting a new gdt. I am almost positive that the one you have is an iron powder one, which isn't good for this high frequency work... Might be something to look into...

Well before you wind a new core, you should probably take a closer look at it. If you have a ferrite core, it will appear silvery-gray, while powered iron cores usually are colored (yellow, red, etc...)

Also, if the core you pulled from the PSU had two windings (i.e. common mode) or four (4) connections, it is likely a ferrite core.

dan

Thu Mar 16 2006, 04:34PM

Registered Member #223 Joined: Mon Feb 20 2006, 06:42PM
Location: Ottawa, Ontario, Canada
Posts: 125

Well I smashed the GDT with a hammer and it turns out it was powdered iron.

I found a ferrite beed around the HV wires on a flyback. I have since wound a new GDT and from my first tests it looks promising. Even at a calculated fixed freqency of 500khz all seems good.(This is the frequency I used from the following tests. I am driving the GDT directly from the PWM. See attempt 1 in the partial schematic in my above post.) I have connected two fets in series accross the 12v and 0v rails to check for signel overlap. If there was an overlap the fets would be shorting out the supply and would probably begin to heat up but they didn't.(so I guess that is good). I also added an LED in series with the two fets to see if there is even the smallest overlap. (the LED didn't light up at all so I guess this is good too.) Now I wanted to see if they were turning fully on.. Using one fet driven by the GDT the fet switched a 5 ohm 5watt resistor. The resistor got blazingly hot in seconds but the fet was only a little warm to the touch(no heatsink). However 12V is a long way from 120V so this slight warming at 12V bothers me a little. Then agian they will be heat sinked when I get around to building the half bridge. Are the resaults from these tests promising? I don't have an Oscope right now but it would probably be a good investment. A new one is out of the question right now and I have been screwed on ebay more than once and vow never to buy anything from there agian.

However i'm not sure about the 500khz frequency.. I am using bread board to prototype this circuit and it probably has enough stray capacitance to effect the calculated frequency.. I am using a 220pf timing capacitor and i'm sure the capacitance between two strips on the bread board is at least a qaurter of that if not more.

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