matching MMC and Primary RMS current / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

matching MMC and Primary RMS current

1 2 next

Move Thread

LAN_403

Physics Junkie

Wed Nov 14 2012, 04:33AM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

I am curious as to what an "acceptable" range would be for matching MMC RMS with Primary RMS. I am using the formula from the hvwiki for Primary RMS current. My H-bridge is using the ixgn60n60 minibloc's and from what I have read can be pushed over 1000A. However, this being my first drsstc, I dont plan on testing the limits right away, would like to get the coil working first

. Since these can be pushed over 1000, I use this for my Ipk value. Assuming I always run a 5% duty cycle, RMS current in the primary will be around 187A. I am not exactly sure how to do Irms of my MMC but I assume that it's (#of caps)(Irms rating per cap). So my MMC Irms is 135A (13.5 Irms per cap with 10 caps in the MMC) Should the Irms of the MMC be at least that of the primary or how close should it be for good operation?

Steve Conner

Wed Nov 14 2012, 06:46AM

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

The RMS rating of your cap bank is the RMS rating of one cap, times the number of parallel strings.

Physics Junkie

Wed Nov 14 2012, 08:12AM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Okay thanks Steve! So that makes my cap bank RMS rating 67.5A. Even if I assume primary Ipk is between 800-1000, that still makes an RMS of 150-187. What else could I do to drop RMS current in the primary other than add more caps in parallel or lower duty cycle to like 2%? For instance, if I added 2 more paralleled series and lowered duty cycle I would have cap bank RMS of about 91 and primary of about 94.5. How important is it that i match these ratings?

Steve Conner

Wed Nov 14 2012, 11:21AM

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

There's a third option: exceed the caps' RMS ratings and hope for the best.

I think many of us do that.

The RMS rating is thermal, set by how hot the cap can get before the plastic film dielectric starts to melt. So you can exceed it for short periods of time if you allow cooling-down periods in between. The data sheet may also specify it at a 70 degree ambient temperature, so it will be higher if the caps start out at room temperature, or have forced air cooling.

Physics Junkie

Wed Nov 14 2012, 12:56PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Ah Okay this is good news!

I didnt want to change MMC capacitance which in turn changes Z and then leads to redesigning this and that...phew! luckily where I live it is -4 degrees celsius and fairly dry. So if I'm careful and just operate the coil outside or in the garage, I should have minimal heating problems.

Wolfram

Wed Nov 14 2012, 03:14PM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

187A RMS with those miniblocks sounds completely unrealistic. That's more in the league of a bridge of CM300s or maybe even CM600s.

Physics Junkie

Wed Nov 14 2012, 05:51PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Wolfram, I was kind of thinking the same thing. It doesnt sound right. I am just going by the formula on the wiki as stated above. Even if I limited current to say, 500A, I'd still have primary RMS of about 93.5 and thats with 5% duty cycle. Do you know of a better way to calculate RMS currents? I hope I'm doing the calculations right in the first place!

Marko

Wed Nov 14 2012, 07:35PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

If I remember correctly, the RMS current is an equivalent of DC current that would do the same power dissipation as your pulsed waveform on a 1-ohm load.

For a square wave current of 500A and 5% duty cycle you'd get 12.5kW of power, which equals to 111.8 amps dc current. However, I think you may be getting unrealistic results for the following reasons:

- Assuming this is for a drsstc, a 5% looks like rather high duty cycle. Something like 1-2% (100bps, 100-200us) seems kind of more reasonable,

- You may have used the peak current value as if it was a square wave, which may not correspond to reality too well due to complex waveform shape of the drsstc primary current, while you should actually look for the RMS current during your burst. It's a current that rings up over time, and is also sine, so to first approximation it may be more like half of your peak current.

And finally, it's worth to note that rms current may not be a good measure of losses in constant-voltage-drop devices like igbt's: for this purpose it's actually often more preffered to use average current!

So for 500A peak and 5% duty cycle, the average current is just about 25A. Count in the increase in Vce drop at 500A to get a more accurate result :)

Marko

Physics Junkie

Wed Nov 14 2012, 10:51PM

Registered Member #7267 Joined: Tue Oct 16 2012, 12:16AM
Location: Detroit, Michigan
Posts: 407

Thanks Marko! Using the average current definitely gives better results here. I'll try with lower duty cycles too and see what works best. Thanks for the tip

Marko

Wed Nov 14 2012, 10:59PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Hi Physics junkie

As an extra note - when calculating losses for your capacitors, wiring, coils, etc. the rms current is the better measure because these are all resistive components; for the igbt's, though, average current often makes a better projection.

On the other hand, if you're using CDE capacitors, they are very sturdy regarding rms current rating. I've ran them up to 20A per cap and they performed very well.. forced air cooling will likely allow for even more!

Cheers,

Marko

1 2 next

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