I just replaced the filter capacitors in this 1956 portable. I've noticed the volume will get softer after a few minutes of playing. Sometimes, it picks right back up and sometimes not. You can still hear the station even after the volume dies down.

I tried replacing the disk capacitors around the 1U5 tube, but the problem is still there. The 1U5 plate voltage is only 15 volts, half of what it should be. I even tried a different 1U5 without success.

The original selenium rectifier is still in place; I'm only getting 95 Volts instead of 115.

Thanks!

If you can get meter readings across the selenium rectifier in both directions (rectifer disconnected), it is leaky and should surely be replaced. This will damage electrolytics and cause low voltages. Otherwise, if it does not give meter readings in both directions, it still sounds like it is weak. You can either replace it with a modern diode and perhaps a 50 or 100 ohm resistor in series, or purchase a N.O.S. selenium rectifer off of eBay. If you use a modern unit, place resistance in series with it until the plate voltages are at or 15% above specified values. Filament voltages should never exceed 1.7 volts per tube (1 volt tubes). If your radio employs any 3 volt tubes (the kind of tube with a divided filament such as 3V4 or 3Q5), keep this in mind when checking voltage here. Usually, since many of these radios have the tubes strung in series, a check at one or two tubes will tell the story for the rest of them. Never unplug any of the tubes in radios like these while they are in operation. The power supply can build up dangerous voltages that will burn out all of the filaments when the tube is re-inserted. You will understand this when you realize that the required voltages are achieved through loading and resistors. Unload the resistor and the voltage will go up.

With disc condensers, leakage is not usually as much of a problem. With my older portables, which use paper condensers, a problem I frequently have is that the volume will die down after 15 or 20 minutes of operation, and will become somewhat distorted. Leakage even in the millions of ohms through the audio bypass condensers will cause this problem. Check all audio bypass condensers, even at the volume control. Check tone condensers as well. Using your most sensitive resistance meter setting, with the condenser disconnected, if the needle swings up at all (a hare's breath or less), it should fall back to exactly where it started. Anything less in these extremely high impedance circuits will cause trouble. Be sure that when you test a condenser that you disconnect one of its two leads. Do not touch both of the meter leads with your fingers at the same time or the conductivity of your fingers will throw off the reading.

Thomas