Blow out! :-)

[KLR650]

Pushed this effort a little too far ;-)

Glad I was several feet away when it went lol

[Pano]

That is not a happy transistor!

[adason]

that capacitor looks even less happy, maybe thats what it all started...you breach the cap, you get dc where you do not want to, there goes transistor

[tvrgeek]

Ah grasshopper, you must strive to keep the smoke inside the wires.

[KLR650]

Originally Posted by tvrgeek

Ah grasshopper, you must strive to keep the smoke inside the wires.

But Master, I leave no trace on the rice paper! lol

I think I exceeded the power dissipation - worked fine 100V p to p and 500Hz but shortly after increasing to 1KHz it went (I'm driving a piezo stack). Exciting! :-)

BTW, I'm running it off +120V and -20V - hope this hasnt got anything to do with it?

[Nico Ras]

Originally Posted by KLR650

But Master, I leave no trace on the rice paper! lol

I think I exceeded the power dissipation - worked fine 100V p to p and 500Hz but shortly after increasing to 1KHz it went (I'm driving a piezo stack). Exciting! :-)

BTW, I'm running it off +120V and -20V - hope this hasnt got anything to do with it?

Put a schematic up maybe we spot something. I may be mistaken but is an array of Piezo drivers not extremely capacitive?

[jaycee]

In the style of Mortal Kombat.... TOASTY!!

[wg_ski]

Originally Posted by KLR650

I think I exceeded the power dissipation - worked fine 100V p to p and 500Hz but shortly after increasing to 1KHz it went (I'm driving a piezo stack). Exciting! :-)

BTW, I'm running it off +120V and -20V - hope this hasnt got anything to do with it?

Most TO-220 transistors have absolutely horrible SOA above about 30V VCE. They may be only good for 5 or 10W (or volt-amps) driving a purely reactive load at those rail voltages. I'd start with something bigger, run the drivers at high bias and use base stoppers to avoid parametric oscillation (the burst-of-oscillation-on-part-of-the-cycle phenomenon).

It may have also gone into cross-conduction and then the current goes completely out of control. Speed differences between NPN and PNP show up with a vengence driving a capacitor.

[KLR650]

Thanks for that.

I dont have a diagram to hand to post but I'm using a IRF630/9630 in a simple design - the bias set by a variable resistor but with a NTC thermistor in parallel to offer some thermal control of the bias. I'll try a more robust output pair and maybe have a low value resistor in series with the output for some isolation. Good fun though!