Mike Knudsen wrote:

> Recently one of the solenoids burnt out and in the process went
> short circuit.  This in turn caused a huge current in the driver
> which subsequently charred and sent a plume of smoke into the air.

Often the reverse occurs, where the driver transistor shorts out
and lets the smoke out of the solenoid!  Either way is bad.

Yes, most people, including engineers, don't think of where that
inductive kick of the solenoid's sudden current stopping goes.  The
energy in the solenoid's field collapses suddenly, causing a big
voltage spike that breaks down the transistor, eventually destroying
transistors not specially built for such use.

It is easily stopped by adding diode directly across the solenoid.
The polarity of the diode is such that isn't a short.  This will slow
up the solenoid release by about 10 milliseconds.  This is usually a
negligible amount.  There are ways of reducing this time if it bugs
you, such as placing a resistor of about three times the solenoid
resistance in series with that diode.  The diode rating can be about
anything, as they have to only carry the 1 amp for that 10 milliseconds,
so even a quarter ampere diode would do, and the voltage rating is only
the solenoid voltage times a safety factor of two.

Will Herzog