Peak power rating for 1N3349 Zener Diode

Hello,
I am looking for the peak power rating for a 1N3349 Zener diode rated for the 180V, 50W. The Zener is being used to discharge an inductor and the discharge time is ~1ms. On all of the datasheets for this Zener diode it says it is rated for 50W, which I believe to be continuous. The normal operating current through the inductor is ~0.7A, so this means that the Zener diode will have to be able to dissipate 126W as soon as the supply is disconnected.
Although the Zener diode is rated for 50W, is this an issue due to the very short discharge time?
Thanks,
Nick

Greetings,

You’re correct in understanding the 50W figure as a continuous rating. When it comes to pulsed conditions though, one can’t meaningfully quote a figure without also specifying the waveform. One can find figures for 8/20us and 10/1000us standard waves given for transient suppression diodes, or 8.3ms half-sine waveforms for devices targeting mains rectification, but I can’t say I’ve ever seen quotes for a 1 ms inductive discharge.

Thus, one’s left to guesstimate and test. Obviously the time-averaged dissipation in pulsed fashion can’t exceed the continuous value, but the extent to which a person can concentrate the power into small pulses is limited by the ability of the silicon die itself to absorb the energy in each transient without overheating, since as pulse width declines, so also does the opportunity for heat to transfer out of the die and into the package.

Offhand, running at 2.5x the continuous figure for a millisecond is not a thing I’d expect to be problematic so long at the application frequency is limited. If a person were talking about applying 50 megawatts for a microsecond once a second (50 watts average…) now that would be more problematic.

Thank you for the response.

In the application I am considering, its actually an inductive load. Specifically, the Zener diode is going to be used to disengage a clutch at a faster rate. The frequency will be very low, maybe disengaging the clutch once every couple months. Modeling the waveform may be tricky…

Would this application change anything?

That would certainly qualify as a “limited” switching frequency. If anything, I’d guess the diode in question might be a bit over-sized for the job.