No limits to DRSSTC / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

No limits to DRSSTC

first previous 2 3 4 5 next last

Move Thread

LAN_403

Hazmatt_(The Underdog)

Fri Mar 17 2006, 01:31AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

I'm really interested in what happens here because I have some IGBT's I want to put to work too. 1200V 300Amp-er's as a matter of fact, each on their own heatsink.

My bet is that during operation when you come up to full power it will work for a while but when an arc is drawn the foil to the leads of the caps will absoltely vaporize!

Have you talked to the manufacturer about the application and the terminal currents?

Dr. Drone

Fri Mar 17 2006, 02:17AM

Registered Member #290 Joined: Mon Mar 06 2006, 08:24PM
Location:
Posts: 1673

Hazmatt_(The Underdog)

Fri Mar 17 2006, 03:43AM

Registered Member #135 Joined: Sat Feb 11 2006, 12:06AM
Location: Anywhere is fine
Posts: 1735

Don't get me wrong, I really want to see it succeed! It's just that I'm considering the physical limitations of the device's construction.

And might I add, I saw your site and goddammm you have a lot of money!
I'm thankful to have a P.T. but that pig power supply floored me. And all the pretty lexan, jeeze. Maybe when I have a job I can start to do these things. For now I have to rely on myself and lack of budget.

HV Enthusiast

Fri Mar 17 2006, 12:51PM

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

wrote ...

DRSSTC put a lot of stress on the MMCâ€™s, so if I can hold this machine together, it should help us determine how far and how much power can be applied to poor old MMCâ€™s.

I don't believe that your 48kVA DRSSTC will shed any additional light regarding the use of MMC capacitors in DRSSTCs other than the fact that you will prove a very large MMC array can be used. As far as individual MMCs go and how much electrical stress they can withstand, that can be determined with almost any DRSSTC. Remember, that your 48kVA DRSSTC uses an MMC with many different parallel strings, so even though the total system power is 48kVA, the individual stress on each MMC is probably quite low and i'm willing to bet that some of the smaller DRSSTCs out there using just a single string of caps will put more stress across each individual MMC capacitor than your set-up.

Dan

Steve Conner

Fri Mar 17 2006, 02:06PM

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

Hazmatt: Most of our DRSSTCs nowadays have an active limiting circuit that measures the peak primary current and starts cutting out the drive if it gets above the limit you set. It does usually kick in when the coil arcs to ground. Provided Chris sets the limiter to a value less than the peak current rating of his cap bank, nothing should vaporise.

You typically install this limiter on your DRSSTC when it gets big enough that the cost of blowing the IGBTs is more than the labour cost of building the limiter and getting it working.

Dr. Drone

Fri Mar 17 2006, 04:28PM

Registered Member #290 Joined: Mon Mar 06 2006, 08:24PM
Location:
Posts: 1673

Steve Ward

Mon Mar 20 2006, 05:43AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

My bet is that during operation when you come up to full power it will work for a while but when an arc is drawn the foil to the leads of the caps will absoltely vaporize!

Have you talked to the manufacturer about the application and the terminal currents?

Where have you been? A few of us have tested these IGBTs to these extremes, and they perform as expected. I personally have used the powerex CM300 type IGBT to 1500A peak. The CM600 is simply 2 parallel CM300s. Assuming relatively good current sharing among dies (which we suspect the current sharing is good enough) the CM600 should handle the 3kA that we are planning on just fine. Ive done a 9kVA DRSSTC system now and understand it well, so the 48kVA project should just mostly be scaling, along with some other design issues.

Just to prove their durability even further, Jimmy ran his CM150 IGBTs at 1700A peak

, no problem.

I'm really interested in what happens here because I have some IGBT's I want to put to work too. 1200V 300Amp-er's as a matter of fact, each on their own heatsink.

While putting each IGBT on its own heatsink seems great from a thermal point of view, its really bad for the extra inductances that it may add. Watch out if your buss connections start getting longer than a few inches... Of course, you may be using small heatsinks, in which case you would need 1 per IGBT.

Steve Conner

Mon Mar 20 2006, 01:23PM

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

I can second that. When I was running my OLTC2 with its big toroid and the 32uF tank capacitor, I must have been putting at least 3kA peak through each brick, and they lived

Powerex IGBTs seem to be pretty overbuilt, and I'm betting Hooper's big coil will do something spectacular.

I think the most likely problem he'll run into is that the huge amounts of EMI from these giant 25ft sparks and monster ground strikes will upset the electronics and the resulting bad drive will blow the IGBTs. My OLTC2 finally died by striking its own wiring harness, causing an EMP that trashed the driver board. (even though the wiring was on the floor underneath grounded chicken wire, and the input to the driver board was transformer isolated inside screened cable)

That can be avoided by careful design and layout of the wiring, and commissioning the system carefully with a gradual ramp-up of voltage and power while keeping a close eye on the waveforms. If anything suddenly changes as the power is ramped (except for breakout) it's cause to stop and investigate. Optical fibre interconnects are nice at this kind of size and power level, and Finn Hammer worked out a cheap and simple design for one that I particularly like.

HV Enthusiast

Mon Mar 20 2006, 01:40PM

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

I agree with Steve.

If you can, I'd try designing your whole electronics control using as high as voltage logic as possible.

If you can use all 4000 series logic at 18V (or similar), this would greatly improve noise immunity. In fact, even today with the big high votlage transimitters, we still use 4000 series logic for most of the control circuitry for this reason.

Also, if you can, I would recommend using isolated heatsinks (at least isolated from ground), especially at those peak power levels. Remember, that there is a capacitive junction between the IGBT device and the heatsink, and if you have the device pulsing with high peak currents / voltages, you can induce some serious noise loops in the system this way. Just something else to look at to help with noise immunity and your control circuits.

Marko

Mon Mar 20 2006, 01:58PM

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

Best idea is to lock everything in a big metal box, as I did on my small DRSSTC (ATX SMPS box).
Maybe not nice as that beautiful arcylyc but very useful and can also look good if we build small wooden case around it.

If heatsinks are grounded (to real ground) it is far better idea than leaving it hang, it becomes heavily capacitively coupled to coil and make noises, coule to IGBTs, board...

At least faraday cage like in Steve Ward's coils seems also good idea.

first previous 2 3 4 5 next last

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