Interupting xenon flash current ?

The flashtube is roughly a constant-voltage device once fired, with maybe a bit of negative resistance. So there is energy lost in the cap esr and the wiring, and maybe adding the igbt just moves the losses around and spreads out the pulse.

I thought igbt's latched on at high currents, but obviously the series guy turns off.

John

There are "short circuit" rated IGBTs that can take hundreds of volts while full on for a few microseconds.

Best regards, Spehro Pefhany

--
"it\'s the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

that in modern units IGBTs are

to flash a tube and measured

current peak to be about 1900A and am

about 6us and still at 300A after

reasonably priced/sized IGBT to

but the only IGBT I have is one I

My test setup isn't all that

about why it takes 20us from the

but that just seems like and

15us for the current to reach

fairly large, 80mm long and 8mm dia.

On SCRs and other devices made for hogh current and relatively fast circuits like this, one should know the I*I*T rating; something that used to be published about 30 years ago and "vaporized" maybe 5 years after. I guess that the I*I*T rating of that spec was low...

material about 1" wide and 5" long

just measured the inductance to

to measure a 3000A pulse in an

around 8ms if I remember

tube. The scope gnd, the charge

connected the scope probe tip to the

5" long shunt????? I bet you are measuring L*dI/dt. 0.3uH with ~1kA/ less_than_1us will easily give you more than 1V. I'd use bunch of chip resistors in parallel (e.g.: ?name=P5.0TCT-ND) or current transformer. Layout is not a trivial thing when you are dealing with the stuff like this. Just my $0.02 :o)

Yes, if we assume Vcap starting at 275 volts, and 60 volts across the tube, about 3.5 ohms of flash- capacitor esr should do it. Hmm, I'd expect a 22uF photoflash capacitor to have lower esr than that.

However they are rated, most IGBTs, in my experience, simply current limit, much like MOSFETs, at very high currents. I'm not aware that they latch up. But, given the typically rather high limit currents, if the voltage drop is very high and lasts for long, destructive die temperatures can occur. Perhaps something appearing like a latch-up can be part of the high die-temperature process. For my high- current IGBT designs, I've used gate drivers that include a Vce over-voltage-detection gate-shutoff circuit, to protect the IGBT. Once I disconnected that circuit, because it was tripping, improperly I thought, but then a huge IGBT module I was using immediately failed. It turned out an insulation- failure short in the output cable was responsible. I promptly reconnected the fault shutoff and only then replaced the IGBT. The shutoff will remain.

Fairchild Semi has this nice app note on IGBT in series with the tube and a 555 timer test circuit for the pulse width

then there is this piece of marketing drivel from power electronics magazine

MIT thesis

enjoy,

I do xenon as part of my livelihood at the University.

Steve the strobemeister.

It seems that all of the IGBTs designed for strobe use are limited to

150 Apeak, suitable only for small on-camera type flash tubes under 20 Joules, and I see the apparently typical CY25BAH-8F flash IGBT is rated for a 5000 flash lifetime.

I just hauled a big flash unit from a ~1970 copier breadboard in from the barn to see how they did it. This flash was designed to fire around 5 times per sec continuosly 24/7 for about a year at around 10 or 20 J per flash IIRC; it has a 5" long 1/2" dia tube in a fan cooled lamp housing and a 19" rack mount power supply with a big ferroresonant transformer and bridge rectifier charging a 4" x 4" x 7" capacitor with inacessible labeling. The capacitor does not discharge very much per flash. In series with the tube is a 0.1 ohm wirewound power resistor with 27 turns of edge wound metal strip 1.25" dia 7" long, I calculate about 4 uH. This resistor drops enough voltage to quench the arc very quickly after it fires, but without any adjustment posssible.

I might be able to fire it up and measure the flash current sometime.

Hey John, do you have a core material recommendation for that 100:1 CT idea?

Regards, Glen

I noticed that to. Even in the TO247 pkg it's hard to find one good for more than 300A.

If you do, please post your results here. I'd sure like to know.

If there is no absolute truth then nothing can be known.

After I get everything mounted and shorten all the wiring as much as possible. I'll wind the transformer and see what happens. I went thru some junk and found a current sense transformer. The data sheet with it says it's 1:200 and to use a 200ohm resistor. It also says 35A max. It has large hole in the center to run a wire thru. Any idea why it is limited to 35A? I don't want to run into that problem when I make my own transformer.

Mike

If there is no absolute truth then nothing can be known.

35A --> 175mA in the winding. It's probably made with fine wire (around 28AWG).

A 200 ohm resistor sounds absurdly high.

What V.s is it rated for?

Tim

-- Deep Fryer: A very philosophical monk. Website @

possible.

found

to use

to

into

No, they go to 10kA and higher. My circuit went to 1.5kA and 1.2kV = 1.8MW. Look for IGBT modules made by Powerex, Mitsubishi, Eupec, ABB and others.

Jim Thompson snipped-for-privacy@My-Web-Site.com posted to sci.electronics.design:

So basically you are commutating the flashtube with the SCR instead of another SCR?

Mike snipped-for-privacy@comcast.net posted to sci.electronics.design:

Mike. Re: your sig, see Godel's theorem. But don't give up it is just a theorem.

Spehro Pefhany snipped-for-privacy@interlogDOTyou.knowwhat posted to sci.electronics.design:

My tongue would hurt after that.

Mike snipped-for-privacy@comcast.net posted to sci.electronics.design:

There is no calibration needed for a current transformer, just the terminating resistor. Just use big enough core to not have a saturation problem. A company called "Ion Physics" makes them to handle many kA in sub microsecond square topped pulses. The rest of the instrumentation must be up to the task as well.

John Larkin snipped-for-privacy@highNOTlandTHIStechnologyPART.com posted to sci.electronics.design:

No, they commutated the flashtube with an SCR. Google SCR commutation.

Pearson makes CT's that work over a million:1 frequency range, down to

0.15 Hz. Beautiful stuff.

John

Yes, I've seen units that used a second flashtube.

John

That mag is pitiful. I may rant on it one day soon.

John