I am sure the short answer is 'any number of ways'.
I have a new NOS tube (RCA produced 5U4G from a reputable dealer), and everything was working great. Had it on on Thursday, no issues, shut off the amp. Turned it on the next day and nothing. The heater element did not turn on. Pulled it, cleaned the pins, nothing. Put in a different rectifier and it was fine.
Guessing that last heat cycle killed it? Going from hot to cold, something probably broke internally a few seconds after I powered it down and the metal started to contract?
I have a new NOS tube (RCA produced 5U4G from a reputable dealer), and everything was working great. Had it on on Thursday, no issues, shut off the amp. Turned it on the next day and nothing. The heater element did not turn on. Pulled it, cleaned the pins, nothing. Put in a different rectifier and it was fine.
Guessing that last heat cycle killed it? Going from hot to cold, something probably broke internally a few seconds after I powered it down and the metal started to contract?
One big thing to do is abide by the first capacitor size limits if you want the rectifier to last a long time. Choke-input is even better.
Rectifier tubes can still die from vacuum loss or possibly poor solder joints at the pins.
It is usually worth re flowing solder on the pins in case case of a dead rectifier.
Rectifier tubes can still die from vacuum loss or possibly poor solder joints at the pins.
It is usually worth re flowing solder on the pins in case case of a dead rectifier.
The filter cap size should have nothing to do with a failed filament in the 5U4G valve.
canonken, did you cross-check that the 5U4G heater/filament is an open circuit using a meter? Have you checked the mains ACV in your house, and what heater voltage the amplifier supplies at the valve base/socket?
Are you able to look through the glass and see the welded heater tabs and the top 'loop' of the filament? There is a photo on page 37 of link for reference for 5U4G. If the filament is really open then you could perform an autopsy and break the glass carefully and use a resistance meter to determine where the 'break' is.
https://dalmura.com.au/static/Power%20supply%20issues%20for%20tube%20amps.pdf
canonken, did you cross-check that the 5U4G heater/filament is an open circuit using a meter? Have you checked the mains ACV in your house, and what heater voltage the amplifier supplies at the valve base/socket?
Are you able to look through the glass and see the welded heater tabs and the top 'loop' of the filament? There is a photo on page 37 of link for reference for 5U4G. If the filament is really open then you could perform an autopsy and break the glass carefully and use a resistance meter to determine where the 'break' is.
https://dalmura.com.au/static/Power%20supply%20issues%20for%20tube%20amps.pdf
Last edited:
Hmmmmm. Started with a good tube. Pins 2-8 which I take to be the heater loop. Immediately showed continuity and <1 ohm, as expected.
Then put the leads on the bad tube. and depending on the position I hold the tube, I get different readings! When upright it might show out of range...or maybe 10,000 ohms for a moment. When I turn it upside down, I sometimes get something closer to 1 or 10 ohms for a second. Continuity beeps on and off.
Guessing an internal filament is broken, and in different positions it might be making a connection, or a weak one, or something else. Either way...it is not happy or acting how it should compared to the good tube.
To answer your voltage question - this is a amp I purchased from a vendor and it is intended for use with this tube, I do not know the voltage or other characteristics, I just know it is designed to work with and compatible with this tube (and many others are running it).
Then put the leads on the bad tube. and depending on the position I hold the tube, I get different readings! When upright it might show out of range...or maybe 10,000 ohms for a moment. When I turn it upside down, I sometimes get something closer to 1 or 10 ohms for a second. Continuity beeps on and off.
Guessing an internal filament is broken, and in different positions it might be making a connection, or a weak one, or something else. Either way...it is not happy or acting how it should compared to the good tube.
To answer your voltage question - this is a amp I purchased from a vendor and it is intended for use with this tube, I do not know the voltage or other characteristics, I just know it is designed to work with and compatible with this tube (and many others are running it).
I CAn say some RecentlY built (last 20 years) amplifiers can be hard on rectifiers for no good reason.
A blue flash inside the rectifier happens if you put a whole lot of capacitance without a choke or some series resistance.
I once looked at an Eico power supply where someone replaced the capitors. Complaint was that it seemed to eat perfectly good rectifiers. Seems it was too easy to replace the first 40 mF with 330mf. I got to see a rectifier have a blue flash on the inside about 30 seconds after power up and the filament go open. I just replaced the rectifier with a two 1n4007s in series with 1M resistors across them for balance. No need to worry about the rectifier sound on a regulated bench supply.
A blue flash inside the rectifier happens if you put a whole lot of capacitance without a choke or some series resistance.
I once looked at an Eico power supply where someone replaced the capitors. Complaint was that it seemed to eat perfectly good rectifiers. Seems it was too easy to replace the first 40 mF with 330mf. I got to see a rectifier have a blue flash on the inside about 30 seconds after power up and the filament go open. I just replaced the rectifier with a two 1n4007s in series with 1M resistors across them for balance. No need to worry about the rectifier sound on a regulated bench supply.
Any comments about cathode conduction and filter capacitance is a 'red herring' and has nothing to do with the OP's issue imho.
canonken, you could advise the 'reputable dealer' of your problem - it may have been checked for ok operation, but as you indicate the orientation of the valve may have masked a cursory check.
There is also a chance that the internal fault is within one of the pins (2,8) but not in the evacuated valve itself - that would require a soldering iron and solder to flood the inside of the pin and possibly make an ok connection. Visual checking of internal welds may be difficult to confirm, as it sounds like a weak connection is being made, so it may be a crack through a weld that is difficult to see through the glass
canonken, you could advise the 'reputable dealer' of your problem - it may have been checked for ok operation, but as you indicate the orientation of the valve may have masked a cursory check.
There is also a chance that the internal fault is within one of the pins (2,8) but not in the evacuated valve itself - that would require a soldering iron and solder to flood the inside of the pin and possibly make an ok connection. Visual checking of internal welds may be difficult to confirm, as it sounds like a weak connection is being made, so it may be a crack through a weld that is difficult to see through the glass