[morzh]

Hi Terry,

The big one definitely! It's a mess to take apart and clean the tar. After that it's a piece of cake. Keep the value of the input cap the same the others aren't as critical.

Meaning the can assembly? Ugh....dread at the thought Those are 1uF and 0.13uF.
Are the 1.0uF really electrolytics? They show "+" sign on the sch.

There are a few bakelite block caps and a metal can with a few .25 or .5 in it if memory serves me correct.

They are all .05uF (50nF) except one that is 0.01uF (10nF), also one fat box combining a 250 Ohm resistor and the cap.

I measured two of those yesterday (I disconnected one terminal to isolate the cap from the rest of the sch): one gave me 100nF i/o 50nF (not sure what to attribute it to), and another one was direction-dependent and kept giving me 75nF in one direction and 350nF (0.35uF) in another, the measurement sometimes for no reason would get down to 120nF (0.12uF).

>Actually, it was my intent today to measure the capacitance of those by de-soldering the surrounding wires. I wonder myself how they have survived these 80 years

It's the leakage that is problematic.

This may be leakage indeed.
But, considering 80 years, and I almost don't hear any 60Hz buzz, and the music is decent quality, the distortion clearly being the Amplifier type (one of the tubes I think is a bit off) - they survived remarkably well, still being somewhat capacitors, and not just bad resistors!

If you think of it, today's electrolytics at full rating (ripple curr. / temp ) are guaranteed for 2000 hrs - it is a pain to pick a good one today for a power application.

Q: The Tube 24A, #3 (the detector tube) has the Plate Voltage in the table equal to 35V. In fact it is 110V-120V.
The rest of the tubes have correct plate voltages, plus-minus 5% or so.
The plate voltage is acquired through a T-filter of two resistors (100K and 500K) and one capacitor (0.25uF) from the "J" voltage point (lower voltage Field Coil contact). This one being 136V also provides the Plate voltage for the 1-st Audio (tube 27) and it is correct (120V or so). The only explanation I have is that the tube is not conducting proper (large enough) current and so presents to little a load, and so does not drop the voltage enough.

This to me may be the result of the lower than needed screen grid voltage on the 1-st audio 24 tube, but it is not listed anywhere - in the table it is vacant.
The voltage is obtained from the "F" node which measures at 12.3V and then filters through 250K resistor and 0.25uF cap, becoming 9.2V at the grid.
I wonder if the cap is leaky and lowers the voltage more than it needs to be.
Well, seems like I explained it to myself and now am gonna go and measure the voltage without the cap.

Any thoughts?