[sexton_mallard]

A NOS UL84 turned up and the selenium rectifier was put out to pasture with a silicon diode and 30 Ohm 7w wirewound resistor to bring the HT on the anode of the UL84 to about 210-215v. The hum?

Hummmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm!

I had no choice but to isolate the mains electrolytic can, a small 3-in-1 Plessey can on top of the chassis next to the UL84.

I really did not want to dig this out as there is a cramped nest of wires and the output transformer underneath. Lack of space and room under the chassis to allow the wirewound resistor to radiate heat safely rules out placing the 3 new electrolytic caps beneath the chassis so a re-stuff is my only practical option. The Plessey can is slightly too small so I may restuff a taller can that came out of a Grundig set after checking for clearances in the case.

With the can out of the chassis disconnected, It appears the 32-40-40 can has shorted internally with a 40-32mfd internal short so to give the effect of a 72mfd. There was no short to the case/earth/ground. The can overall is not leaky as it held a charge well. (One test with a solid state rectifier was to flick the power on and off after 1 sec before the valves heated up and conducted, not very accurate but the can discharged slowly from about 280v over 10 15 minutes on all terminals). I know two were strapped with a resistor so the short would not reveal itself until the can was removed. My component tester showed 77mfd on two sections, and my multimeters showed a dead short across two of the the terminals.