SISG-based Marx? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

SISG-based Marx?

Move Thread

LAN_403

J. Aaron Holmes

Fri Jun 23 2006, 12:07AM

Registered Member #477 Joined: Tue Jun 20 2006, 11:51PM
Location: Seattle, WA
Posts: 546

I was thinking earlier today (while I was supposed to be working

) that one might make an "off line" Marx generator (OLMG) using SISGs instead of air gaps. Can anybody think of why that wouldn't work? For example, the following schematic shows how I imagined this might be put to use as the power supply for a small OLTC:

...however the Marx portion might have other uses. Using Terry's existing SISG design, less a couple SIDACs, it seems like getting 900V out ought to be doable. I'd worry about more than that (and certainly anything over what any single IGBT could handle) because it seems like the speed of the SIDACs couldn't otherwise keep up with the rise time after the first SISG fired, and IGBTs would end up getting overvolted and ruined.

If this seems doable, does it matter whether some SISGs have lower firing voltages than others? Or could they just each use one SIDAC for 300V. If firing order needed to be enforced, one could obviously use one SIDAC on the first, two on the second, etc. My illustration above suggests a 900V output with three 300V SISGs, using a voltage doubler as the first Marx stage. That idea probably deserves some scrutiny too!

Regards,
Aaron, N7OE

Marko

Fri Jun 23 2006, 12:30AM

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

It can be actually easily done, but I don't know how much use can anyne have of some short few-kilovolt pulses.

For SSTC driving there was an topology researched by steve ward based around marx but as he said in the end DRSSTC's put it to shame forever.

Terry Fritz

Fri Jun 23 2006, 02:52AM

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

Hi,

I don't see any problems with the SISG "relativly low voltage" Marx. At high voltages, we worry with false triggering... Best to set the gate drain resistors right (use a ten turn little trimmer pot!!!!) so the timing works out good.

It is all pretty unknown territory... But the physics seems fine...

Cheers,

Terry

J. Aaron Holmes

Fri Jun 23 2006, 03:20PM

Registered Member #477 Joined: Tue Jun 20 2006, 11:51PM
Location: Seattle, WA
Posts: 546

Terry Fritz wrote ...

At high voltages, we worry with false triggering... Best to set the gate drain resistors right (use a ten turn little trimmer pot!!!!) so the timing works out good.

Is the timing really critical for a three-stage 900V Marx like this, though? If one stage did false-trigger, you shouldn't be frying any 1200V IGBTs, right? What I lack a sense of is the importance of firing order. I suppose if this thing is "off line", then the reason for guaranteeing that the first stage fired first would be to keep the big voltages out of the supply line, yes? In which case, perhaps I should put a single SIDAC in the first stage, and two each (for 600V) in each of the last two stages.

Firkragg wrote ...

It can be actually easily done, but I don't know how much use can anyne have of some short few-kilovolt pulses.

For SSTC driving there was an topology researched by steve ward based around marx but as he said in the end DRSSTC's put it to shame forever.

The "use" would be for a small OLTC-ish coil without a primary transformer. With proper selection of components, it seems like one ought to be able to get 60bps at 900VDC. That seems "ok" for a small OLTC. I did a bit of poking around for solid-state Marx designs last night and saw one based on IGBTs that produced a large multiple of something close to the IGBT's rated voltage, which seems like it would indeed require very precise timing to ensure that all the IGBTs fired at precisely the same time. Else they'd be vaporized. The thing that attracted me to this SISG-based LV Marx was that it appeared that timing was *not* critical, at least w.r.t. keeping the components from meeting terrible smoke-releasing ends.

Regards,
Aaron, N7OE

Terry Fritz

Fri Jun 23 2006, 08:40PM

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

Is the timing really critical for a three-stage 900V Marx like this, though? If one stage did false-trigger, you shouldn't be frying any 1200V IGBTs, right?

The IGBTs can't fry. If one opens up, it will just retriger and close again. The timing should not be a concern at all here.

the reason for guaranteeing that the first stage fired first would be to keep the big voltages out of the supply line, yes? In which case, perhaps I should put a single SIDAC in the first stage, and two each (for 600V) in each of the last two stages.

They should all fire at the exact same time and I am not sure you can do much about that. Perhaps if the first stage had a lower gate resistor say 50 ohms instead of 100. Then that IGBT would turn on faster compared to the others.

Cheers,

Terry

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