Tesla Coil Build Thread

[Atomicrox]

I don't think an SCR could switch quickly enough for a SSTC. If it's rated 50Hz you may be able to push a little further but not into the 100's of kHz. I've seen some work on OLTCs (offline tesla coils), which are similar to what you described: the silicon only switches the "interrupter" (usually 100-400Hz) and not the resonant frequency, doing what the spark gap would on a SGTC. I don't have a link ready but it should be easy to google. I do remember they being very complicated but with those parts available it might be worth a shot!

 

[benmwv]

50hz isn't the rated frequency, that's just the max rep rate for an 8000A peak. I will have to dig through some datasheets, but I'm sure it can handle a few khz at least if we are talking <1000A.

If It doesnt work out for a TC, then I'm pretty sure I can make a badass induction heater from it :eg:

OK so it looks like there is 1us delay on the turn on, with a maximum of 1000A/us current rise. 100us Turn off time.

It is rated for 8000A for 10ms with a max of 50hz rep on that. So actually the maximum it can handle is 1800V 8000A at 50% duty cycle 7.2MW.

So no more than 10khz, realistically probably 7-8.

 

[Atomicrox]

Whatever you build with it please don't forget to post here XD

 

[benmwv]

I might just go for an induction heater. Or both, I have a couple of them.

I could use a work coil measuring 12in dia x 12in long with 12 winds of copper tube = 30uh. Then use a 47uf cap to make the resonant frequency about 4khz. I can just use 20 2.2uf film caps or whatever.

By my calculations this would pump out around 2kw at line voltage. It would increase a lot when heating metal too because of the inductance changes.

Probably wont build this until like christmas break or something though.

 

[Sigurthr]

Yep, you can make an OLTC from them but that is about it. I guess you could do "single shot SGTC" as well with the SCR as the SG on a DC charged "SG"TC. Would be simpler than a real OLTC which are challenging.

 

[benmwv]

Well, I just looked up OLTC. I pretty much invented it again in my post up there haha, they are exactly what I was picturing.

For the OLTC all I really need is a few caps, an inductor, the SCR, and then a voltage based trigger. So basically a resistor divider on the tank cap going to an opamp.

Am I missing something? Seems much easier than an SSTC.

 

[Sigurthr]

The tank circuit will see ring up and ring down with Vpk of several kV for a 120v mains system if the mains isolation secondary is made resonant. So it isn't quite as easy as sensing the tank voltage.

If you want to do an OLTC the right way do what Scopeboy did:
Tesla Four - proposal
Tesla Four - system diagram

 

[benmwv]

Oh crap that is complex. I wont be building anything like that!

I might just do like you said, basically a one shot sgtc with the SCR in place of the spark gap and trigger it by a simple adjustable 555 timer circuit.

Sig, atom if you had these what would you build? You think the induction heat will work well?

PS i'm gonna try a slayer tonight

 

[Sigurthr]

If I had a SGTC laying around I might try a single shotter, but otherwise I'd build a mean old IH. I still need to make one myself.

 

[benmwv]

How about that, but replacing the lamps with a center tapped work coil. The center going to v+. So by *alternately activating the scrs you allow *current to flow on each half of the coil. Creating a sort of square AC.

 

[Atomicrox]

TBH I don't understand why OLTCs have to be that complex but I trust the few guys out there that have built one have very good reasons. I wouldn't try to reinvent the wheel unless I had very good electronics knowledge, which isn't the case

 

[Sigurthr]

Iirc it is because they use a fully resonant power coupling system so you need a way to repeatedly interrupt the input power when Vres and Ires reach a setpoint. Think DRSSTC but with a resonant mains transformer and no input ballasting only instead of tens of cycles of rise it is running CW for thousands of cycles at a time.

 

[Atomicrox]

Makes sense but couldn't it work more like an ARSG SGTC and just interrupt on a fixed frequency (say 400Hz)?

I suppose some sort of ballasting would be needed to avoid too much power draw, or at least the primary cap should be sized not to be mains resonant.

 

[Sigurthr]

The problem is that in order to get the performance expected you need resonance throughout to generate the massive resonant currents. "Classic" TCs (not solid state modern types) have hundreds of amps circulating in the primary circuit for even 100W or so of power input. For example, my very first SGTC ran from a 225W NST and had more than 600A primary current due to resonant rise. You look at a large SGTC, like a 2kW monster, and you're looking at sometimes 8, 9, or 10kA primary current. I've seen high power RSGTCs destroy 1/4" tungsten electrodes.

Remove the (tertiary) resonance and limit the current capacity and you basically have a DRSSTC.

 

[DashApple]

Well I got another Half bridge IGBT , 1.2Kv @ 75 Amps , But I really don't know if it will run CW at 85 KHz pushing some 3- 4 Kw of power .

Hard switching hell .....

 

[Sigurthr]

You need a phase controller like Ward uses for his QCW stuff so that you can get ZCS, else you're gonna pop it in anything other than a PLLSSTC.