If a valve has a separate base rather than pins on glass or wire-ended (B8D, B5A etc) don't assume it's faulty till you have removed the base.
I had bought a 1D5 from an ebay supplier who checks the filament and I was sure I had checked it before giving feedback a few months ago. I decided to test it and the other valves used by my Ever Ready 5017 prior to repair/restore. A multisection capacitor is in the B5 location for the U4020 rectifier valve and the field coil speaker replaced by a normal one, both I think about 1965 or 1966 on this 1935 radio.
Anyway, the heater was open circuit. So I decided to see if it Really dead
or just a corroded wire.
I soaked the base in boiling water which made glass come loose. I used my SMD heatgun and tapping to get solder out of pins. Then fresh solder (for the flux) and solder sucker on each pin. The base then came off.The heater works!
The Anode wire was about 1.5cm too long! It must have caught on assembly originally. I discovered a shorter insulated centre tap wire which I added extension to for experimenting as there is an unused pin.
The pin order is unusual on B4 / B5, it's 1 3 2 4 (clockwise from separate pin, bottom view).
The valve was Anode - CT - h1 - h2 - Cathode
on the Pinch (quite visible, but use a meter to check heater connections). I added some soft sleeving to all the wires and assembled the base (without glue) and soldered up the wires not too tight. The Anode wire I pushed back a little and cut off 1cm.Reassembled view from "behind" valve label.
Testing is here
I'll add a bead of epoxy or ersatz flour & lacquer later after the radio is restored.