3-phase bridge diode modules / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

3-phase bridge diode modules

Move Thread

LAN_403

Quantum Singularity

Mon Apr 28 2008, 03:56PM

Registered Member #158 Joined: Sun Feb 12 2006, 09:53PM
Location: Central Ohio
Posts: 282

Just wanted to get some verification here before I blow anything up...

A 3-phase bridge module, when used as a single diode between the +/- terminals effectively has 2x the voltage rating and 3x the current rating, correct?

For example I just got this on ebay:

and plan to use it as a protection diode for my coilguns that I want to get up to 900-1200V charge on. Its rating is 800Vrrm and 1200A Ifsm and 6000A I2t.

So then theoretically when used just across the +/- it would be 1600V, 3600A surge and 18000A I2t, correct?

HV Enthusiast

Mon Apr 28 2008, 05:37PM

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

wrote ...

So then theoretically when used just across the +/- it would be 1600V, 3600A surge and 18000A I2t, correct?

Nope - sorry. The rating across +/- is 1600V, 1200A surge. (Basically two diodes in series)

It would only multiple by three (current rating) if you could absolutely guarantee each diode conducts perfectly. But in reality, due to differences in Vfwd (and other characteristics) between diodes, only two diodes will actually be pulling the majority of the current. And the problem worsens with temperature.

For example, you might have your three phases conducting (due to Vfwd mismatch) at 85%, 15%, and 5% current. But the junction on the 85% is going to heat up more than the other two phases. As temperature increases, Vfwd will decrease, and therefore suck even more current (percentage wise) thus leading to a thermal runaway condition where that one diode phase is carrying practically all the current.

So if you use as a diode, only assume one phase of that diode will be conducting.

Quantum Singularity

Mon Apr 28 2008, 07:23PM

Registered Member #158 Joined: Sun Feb 12 2006, 09:53PM
Location: Central Ohio
Posts: 282

I see what you are saying. But seeing how it is all one module I'd think the diodes are pretty well matched unless they are damaged? If its true then there is no sense of me stacking a whole bunch of (the same) SCRs and clamping diodes all in parallel in my current setup I've been using? What about the conduction time stacking in series, if one diode begins to conduct a little faster than the other than the other would see more voltage so one couldnt really add the voltage ratings either, or is that not as pronounced as the different current division between them?

I'm curious, has anyone done any experimenting paralleling diodes/thyristors, and monitored them under pulse discharge with a multi channel scope to really see how equally (or I guess unequally actually) that they split the current?

HV Enthusiast

Mon Apr 28 2008, 09:48PM

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

Diodes and SCRs should never be paralleled for this reason. Therefore, either get bigger diodes or a bigger SCR. Its especially worse in pulsed applications. Put a slug of current through it, the diode with even a slighter less voltage drop will conduct while the other just sits there. Especially since high di/dit will see high transient temperature rises which cause transient runaway.

Just do the test yourself. Get some diodes, run some current through a bunch in parallel and make some measurements. Very easy to do.

Quantum Singularity

Mon Apr 28 2008, 10:32PM

Registered Member #158 Joined: Sun Feb 12 2006, 09:53PM
Location: Central Ohio
Posts: 282

Dr. Shock wrote ...

Diodes and SCRs should never be paralleled for this reason. Therefore, either get bigger diodes or a bigger SCR. Its especially worse in pulsed applications. Put a slug of current through it, the diode with even a slighter less voltage drop will conduct while the other just sits there. Especially since high di/dit will see high transient temperature rises which cause transient runaway.

Well I suppose I got enough SCRs then to use in many a coilgun since I've been paralleling at least 3 or 4 in every design. Perhaps I'll take some of the cheaper ones and test em out just using 1 instead of the whole bank. I was just real nervous about just one holding up to a 3kJ+ coilgun, and it did seem pretty common here on 4hv to make such a bank, so perhaps folks before just arent speaking up that this isnt necessary.

Dr. Shock wrote ...

Just do the test yourself. Get some diodes, run some current through a bunch in parallel and make some measurements. Very easy to do.

Well I'd like to, and my previous induction launchers like I said used several in parallel, but I dont see how I could measure anything without a fancy scope since were talking pretty short pulse times here.

Steve Conner

Tue Apr 29 2008, 08:57AM

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

Did someone tell Jason Rollette that paralleled SCRs don't work? He used them in his railgun for years

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