David wrote:

One note on voltage drop of selenium rectifiers: Rectifiers
were constructed by placing a set of plates is series. The
size of the plate determined the current rating. A new
rectifier had about 1.5 volts drop per plate at rated
current. Each plate could withstand about 45 to 50 reverse
volts. Most designs pushed the reverse voltage right to the
limit. As the device aged, the forward drop would increase,
power dissipation would rise, and the stack would eventually
fail with a very obnoxious smell.

I would have guessed pushing the current to the limit was more
dangerous (at the same power dissipation). The common
radio set with up to 100 mA seems to have been a safe haven
for seleniums as long as the electrolytic caps in the power supply
held or the fuse blew early enough?

Regards.
H.