PMOS in parallel with NMOS

Sounds like a screen grid resistor to me, which normally keeps the screen grid at say about 200-300 V. An internal short in the pentode or a fault in the socket could connect the resistor to ground, causing

However, it is important to read the small print how the power dissipation of high power resistors, not to mention any metal case transistors are measured. At least for transistors, it is measured at

The _repetitive_ voltage peaks in an AC waveform can cause high I²R dissipation and even high average power dissipation for instance in a volt meter voltage divider resistors, compared to the average (not RMS) voltage.

However, in a guitar amplifier with typically a separate switch for the anode voltage (while the heater voltage is kept on during coffee/bear breaks) only cuts the current flow through the anode voltage choke, which of course may cause quite high voltage peaks. This voltage peak (and possibly peaks caused by cutting the mains voltage) may be lethal to any semiconductor component, but it definitively does not destroy any resistors, since this is a _non_repetive event, unless of course the resistor is operating past its maximum ratings in normal condition.

Paul

Hi Mike and Paul,

No! Sorry!... My fault. I have left out the rest of the story.

It was not just a humble screen-resistor. The resitor in question was part of the B+ power supply chain that booth feeded the power-amp and the pre-amp. It's task was to drop the voltage from 370V (poweramp-voltage-side) down to 200V (preamp-voltage-side). Current sinks were two pairs of 12AX7 triodes. Each of them had a screen-resistor of roughly 100K and was working at ~~ 1mA. The cathodes where biased around 100V when idle (full swing. what you would expect in an audio amplifier).

Overall the 10K@10W resistor had to drop 170V at 4mA (worst case), e.g. just roughly 0.7 watt. 10 time overrated.

The reason for such a massive overrating was *just* the inductive kick-back that happends during power-off. It was impressive for me to find out how much energy a fist-sized trasformer near saturation can emit if the magnetic field breaks down. Holy S*it!

The job of the 10K resistor was two fold. First off it had to drop 170V during power-on, but had to current-limit the reversed voltage-spikes during turn-off as well. This saved the life of the preamp-side filtering capacitors (they drop down to easy 50V during shut-off).

The poweramp-side of the amp however has to deal with the majority of the current. If *that* resistor wouldn't be there most likely the preamp-tubes would have fused inside. The poweramp-tubes take that easily because the arching (milliseconds off) happends at the tube sockets, not inside the glass, and there is not enough energy inside the coil to back-bias the "big" filter caps.

Told me a lesson, and after I have analyzed what happends during turn-off I have learned deepest respect for for the pioneers of electronics. Tube stuff ain't for sissies.

Otoh I have some transistor book from the 60th here. Written in the pioneers days. It contains a shitload of circuits that from todays perspective is good for a good laught. They just took tube-circuits, divided all volages by ten and replaced all triodes by germanium PNP's.

Seems like the fact that tubes distort in squares law (like fets) and bjt-trannies do the exponetial thing wasn't well known back then.

Nils