Nice enough cap. We were in a hurry, couldn't find anything locally, so we just made one right then and there, late one night. It was less trouble making it than just the nuisance ordering one, plus zero delay.
Being mylar it wouldn't stand continuous duty, but gave years of faithful service in Ol' Sparky.
Cheers, James Arthur
You'd be best of examining the performance of the circuit you have, with an eye on defeating specific behavior. Have you actually recieved parts back for this condition on warranty? They might benefit from failure analysis.
Keep in mind that few commercial suppression circuits will serve indefinitely. Most components have a wear-out mechanism - this is often stated specifically in their datasheet as current or joules x number of events.
Even fuses, presented in partial, unavoidable parallel surge paths, will degrade incrementally for each instance.
RL
Thanks. Spark gaps are certainly cheap enough.
Chris
It looks like there is another manufacturer of similar-sized IGBTs that might do the job (but they don't have datasheets online so I can't check the surge current capability and turn-on time, and anyway I would of course prefer the one that you have already tested!).
They also make a SCR that seems to be able to handle the di/dt and over 4kV:
I still think I had better not build one for now.
Chris
If you're looking for surge capacity (and getting more bang for your buck), check out articles from some of the particle accelerator labs -- such things like IGBT-pumped magnetic "compressors", shock lines, etc. (saturable reactors making sharp, high energy pulses).
I think the general idea is you can pulse them at whatever they'll handle; trouble is, as soon as they stop handling it, Vce rises (desat) and it explodes in 10 microseconds of those conditions, instead of 1000. :) Tricks for increased speed and power include higher gate voltage (20-30V instead of 15) and overshoot (goosing it to 30-60V for a few hundred ns to speed up the internal gate voltage node).
Under such conditions, relatively slow fall (turn-off) time may also be desirable, as it absorbs more inductive energy as conduction loss rather than avalanche destruction (and provides time for your external snubber to take over). That also depends if you're turning off sooner or later than the energy storage element can provide (which might be useful to truncate a 10/1000 network's decay, making it into an 8/20 pulse instead?).
Tim
-- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
Hi,
I forgot to say we have used this one :
Please take a look at the test circuit. we have done it exactly as it is designed (capacitor is 4uF/2KV) and instrumented it with LabView. After one week or so of free running (1 C discharge every 5sec) it finally blow ! 1500? i remember :( ... meanwhile the mitsubishi and abb IGBT modules are rock solid when operating with the same conditions (during a month!) and it runs at this very moment for an electromechanical relay production line.
Habib.
Thanks!
Years ago, I worked on a Blumlein triggered spark gap nitrogen laser, used for material science studies. I have heard that it still works.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.