SSTC EMI problems / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

SSTC EMI problems

Move Thread

LAN_403

Thomas

Mon May 01 2006, 04:23PM

Registered Member #120 Joined: Fri Feb 10 2006, 07:07AM
Location: Westchester New York
Posts: 83

My most recent SSTC is exhibiting the most interesting problem that I just can't answer. I have simple circuit based off of Eastern Volt's SSTC1 (to give you an idea as to what i'm using). One thing that I like is to run my coil driver's at 15Vdc, since I feel that it's less susceptible to induced EMF, well this experiance sure proved me wrong. I never had a design problem like this before, so lets get to the point. Right now I'm still finishing up my coil, well, all I must do is install the pots in a project case and hook 'em up to a ethernet cable then to the driver board. Now my problem is that this coil in particular won't tune. I mean it won't tune no matter what I do. However, I can tune it with my hand by moving it's position and a little tinkering with the freq. res. pot. When I touch the pot's either for duty cycle or for the res. freq. adjustment my coil immedaitly loses it's tune and about 90% of the time it never comes back to it's state before I touched it. If I adjust the position of my hand between the secondary coil, my vairac, ground wire and pots my coil can be semi tuned (for testing the gnd wire was to my left, the pots were right in front of my on the floor, my viarac was to my left and the sec coil was in front of me on the coil's cabinet (above and a little forward from the pot's)). This is all to strange to me, but it seems as though my pot's are being induced with EMF from the secondary coil. Do you guys think this is happening and that once I install my pots in they're project case the problem will deminish? Or is this something much greater? Oh may I also add (off topic) should I ground my H-Bridge heat sink? I grounded mine but I would just like to hear that is it is or isn't a good idea, thanks.
-Tom

HV Enthusiast

Mon May 01 2006, 05:55PM

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

Firstly, the frequency adjustment pots should be close to the TL494. If you have them via ethernet cable somewhere else, then i definitely expect that to produce some sort of problems.

Also, my SSTC1 design is an open loop design and tuning is extremely sensitive, especially to proximity effects.

Your best bet would be to modify your coil and go with a self-resonant design.

Steve Conner

Mon May 01 2006, 07:21PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

I totally agree that your pots are getting zapped with EMI. If they have metal cases, the cases ought to be grounded to the ground rail of the controller circuit. That should help it.

Thomas

Sat May 06 2006, 06:09PM

Registered Member #120 Joined: Fri Feb 10 2006, 07:07AM
Location: Westchester New York
Posts: 83

Well I just made a new circuit. I used a little of Steve and Easternvolt's ideas. I made it very DRSSTCish but without the interrupter, I just used a comparator instead. Now I just have to test some of the values in my board.

Terry Fritz

Sat May 06 2006, 11:10PM

Registered Member #393 Joined: Tue Apr 18 2006, 12:30AM
Location:
Posts: 297

Hi Tom,

I posted notes about my DRSSTC noise problems here:

Maybe something there will help.

If your pots are mounted far away from the timing circuit, the capacitance between the leads in the connecting cable might be relativily "enormous". Perhaps greater than the real timing capacitor. Also, those signal lines are going right to the timer chip's comparators so any glitch or spike will trigger the timer chip. A metal shield calbe probably will not help and Cat5 cable is not sheilded. Yor really need to have the pots close to the timer chip circuits. Like inches...

Grounded metal boxes are great for stopping voltage fields but poor for magnetic fields. A metal box can be put on the top toroid with no problem but not on the primary coil. In the case of Tesla coils, the primary current is very high and it can induce significant voltages on near "digital logic".

So you have to keep the electronics sort of far away and try to put the longer wiring at a 90 degree angle to the primary coil windings.

If you ground the heat sinks, that provides a very high current path if (when) the IGBTs explode. Of course, it is safer if you are holding them at the time

I would float them since there is no way you should touch them when they are on. If they go hot during a failure, there will be no current flow to ground then which should help limit damage. If they float, they will radiate more noise though. There is no "one answer" but just things like that to consider.

Maybe you have it all fixed now, but these are the things I would worry about. "Noise" is always tricky and we can't "be there" to see it directly, we can only guess from afar. But maybe you can avoid the things that bit us

Cheers,

Terry

Thomas

Sun May 07 2006, 12:10AM

Registered Member #120 Joined: Fri Feb 10 2006, 07:07AM
Location: Westchester New York
Posts: 83

I'll display my driver circuit and psu. I'll make a pcb of it, but I would like to insure first my values are correct. My PWM is a bit of a variation between Steve's SSTC feedback design with Richie's PWM and some with some additional things and adjustments I added.

[Edit: Go here

to learn how to post pictures.]

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