Quick Drsstc question / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Quick Drsstc question

Move Thread

LAN_403

StevenCaton

Thu Sept 03 2009, 12:01AM

Registered Member #1845 Joined: Fri Dec 05 2008, 05:38AM
Location: California
Posts: 211

Assume there exists a drsstc with no secondary coil and toroid. Now say that you inject a 100KHz square wave into the system to get the IGBT's going.
My question is, would the tank circuit current in this situation double on every half cycle of oscillation until something explodes? In my mind, it would have to double (well, almost double, every half cycle)

In real life, we don't see the current in the tank double every half cycle because of the secondary resonator loading the bridge right?

TheBoozer

Thu Sept 03 2009, 12:14AM

Registered Member #1535 Joined: Wed Jun 11 2008, 11:37PM
Location: Northeastern Pennsylvania - USA
Posts: 117

I test my DRSSTC's with no secondary or topload initially and the tank rings up super fast. I do this to calibrate the OCD. I use one of Steve Ward's controllers and the OCD stops it from running away. I'm also using his interrupter.

I've also dropped in large iron cores that were once transformers to load the primary down.

I doubt I've answered your question, but, in my case, without the OCD, yup, it would explode. Once the secondary and topload are installed, it pulls enough away from the primary so the OCD hardly kicks in.

I estimate my IGBT's would blow after about 10 rf cycles with my tank circuit unloaded...

teravolt

Thu Sept 03 2009, 02:29AM

Registered Member #195 Joined: Fri Feb 17 2006, 08:27PM
Location: Berkeley, ca.
Posts: 1111

I am working with Steve Wards controler with OCD and have frequency dubleing happening to. It is the full drsstc minus the secondary. Maby I need a secondary as a tuned load . haven't figured it out yet

TheBoozer

Thu Sept 03 2009, 08:39AM

Registered Member #1535 Joined: Wed Jun 11 2008, 11:37PM
Location: Northeastern Pennsylvania - USA
Posts: 117

Teravolt: You're still having that problem? Can you take a scope snapshot? You don't need a tuned load in there to test it. Here is a photo of my scope shot during OCD calibration. I'm using a 200:1 CT, so it's 10amps per volt on the screen. There is no load in the primary. OCD set to 500 amps...

Rich

Steve Conner

Thu Sept 03 2009, 10:45AM

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

Yes, the tank circuit will ring up until something explodes. (Assuming that 100kHz was in fact the exact resonant frequency.)

The current won't double on every half-cycle though. The voltage increases by something like 4/pi times your inverter's DC bus voltage, and the current increase goes according to this voltage increase and the tank's characteristic impedance.

So in practice, neglecting losses, it goes up by the same number of amps per half-cycle.

StevenCaton

Thu Sept 03 2009, 11:55PM

Registered Member #1845 Joined: Fri Dec 05 2008, 05:38AM
Location: California
Posts: 211

The voltage increases by something like 4/pi times your inverter's DC bus voltage

What voltage? The voltage across L and C right?

So in practice, neglecting losses, it goes up by the same number of amps per half-cycle.

Well, I figured that, but tell me if I am thinking about this correctly.

Once the tank gets pinged with , say, +200V, the current surges up to a level, (lets say 50 amps) dictated by the surge impedance. (which L and C affect) So the large current on that first half cycle has essentially built up energy stored in an electric field (mmc) and a magnetic field (primary coil). But now lets say instead of going to -200V, the bridge doesn't apply a negative voltage. Then, in theory, the tank should ring up to near 50 amps on the negative half cycle. (not quite 50 though because of heat losses and such) Ok, so now assume that the bridge did place -200V across the tank, then it should ring up near 100 amps on that cycle, because we have the negative voltage from the bridge, AND the energy transfered into the system from L and C, which were charged up on the first positive half cycle. That how I'm thinking about this. Is that incorrect Steve?

Steve Conner

Fri Sept 04 2009, 10:25AM

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

It's correct for the first two half-cycles. But if you extend the argument, it goes 0, 50, 100 as you said, then 150, 200, 250, 300, etc.

Not 0, 50, 100, 200, 400, 800, kaboom.

The inverter does add more energy (proportional to Vdc*I) on each half-cycle than the one before, but because E=0.5*C*V^2, it all works out to a constant increment of voltage.

DC bus voltage is what your DC bus caps are charged to: 200V in your example above, but in practice usually 160V, 320V or even 600V for the brave.

StevenCaton

Fri Sept 04 2009, 03:17PM

Registered Member #1845 Joined: Fri Dec 05 2008, 05:38AM
Location: California
Posts: 211

Not 0, 50, 100, 200, 400, 800, kaboom.

Oh LOL, I never meant that. I knew the bridge will always transfer X amount of energy into the tank every cycle. I shouldn't have been saying "double" afterall, because that is only true for the first two cycles.

The inverter does add more energy (proportional to Vdc*I) on each half-cycle than the one before, but because E=0.5*C*V^2, it all works out to a constant increment of voltage.

DC bus voltage is what your DC bus caps are charged to: 200V in your example above, but in practice usually 160V, 320V or even 600V for the brave.

Ok, I've got it now, Thanks Steve.

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