Capacitor smokes!

[davidnow]

Hi,
after completing the stereomourII with all the upgrades, i have listened to it for a while although i knew the DC filament supply voltage was a bit too high.
Today i decided to mesure it again in order to post the values to get instructions for resistances change (hence my previous post).
I followed the manual step by step, installed the 2A3 tubes (and took off the 12AT7s), plugged the amp and wait few minutes for the voltage between A1 and A4 to stabilize (2,7V), then mesured voltage between D1 and D4.
While i was waiting for voltage to stabilize, the capacitor 20uF between 15 and 16 began to smoke quite a lot from its cross at the bottom (opposite side of leads). I unplugged the amp immediately.
The amp has played flawlessly for hours before i decided to mesure it once more. I am wondering what mistake i made (take the 12AT7 off?) and what i have to do to repair (just change the capacitor)?
Thanks for help
David

[Paul Birkeland]

The 2A3 tubes have two fat pins and two skinny pins. If you push hard enough, you can force the fat pins into the holes for the skinny pins.  When you do this, the filament inside the tube will be connected between B+ and the cathode resistor, thus creating a short between the 400V power supply and the cathode resistor.  The cathode bypass cap is rated for 250V and will blow under these conditions.  The 2A3 that was used in that socket under these conditions will also be destroyed in the process and need to be replaced.

If, by chance, you were trying to measure the DC voltage between pins 1 and 4 with the tube improperly plugged in, this voltage will also seem higher than it should be.

[davidnow]

Thank you very much Paul for your answer. Actually i don't think i made a mistake because i just unplugged the 12AT7 to mesure DC filament supply and let the 2A3 as they were when the amp played music. i realize i might not be clear enough : the amp played quite well, i unplugged the 12AT7s to mesure filament supply and i did not unplugged the 2A3s.
David

[Paul Birkeland]

If the 220uF cap at 15 and 16 smokes, it is because the 2A3 isn't plugged in properly.  I wouldn't run the amp anymore if that cap is blown.  There's certainly the possibility that something else around that cap smoked, but this is an issue that is common enough that we are familiar with the circumstances that create it.

[Paul Birkeland]

I would also add that a much less likely scenario to blow that cap would be the 2K/3K cathode resistors not being adequately soldered and falling out of the circuit electrically, just leaving the cathode bypass cap in place.  This will also blow that cap right up in a hurry!

[davidnow]

I unsoldered the 2K/3K resistances, one has the ceramic broken. Both resistances are cut, i mean there is no current that can pass through (sorry for my English). That could be the origin of the capacitor blow? But how these resistance can also be blown up?
David

[Paul Birkeland]

If those resistors are blown, that's further evidence that the 2A3 has been plugged into the socket incorrectly, and the filament of the 2A3 has shorted B+ to those resistors.  This is absolutely what has happened, your DC voltages will probably be way off, and you will have a 2A3 that's well on its way out.

[davidnow]

I respect your advice but as i did not move the 2A3 from playing position (and it played well) to mesure the voltage, there is something i don't understand...
The DC filament supply voltage was 2,65V before the capacitor blew out and i unplugged the amp.

Do you think it is enough to replace the resistors and capacitor and then proceed to voltage check out?
David

[Paul Birkeland]

Can you post some photos of the build?

With the blown resistors and capacitors that you have, the filament voltages that you're measuring are not useful, as half the amp is likely not working properly.

[davidnow]

I mentioned the voltage before the capacitor blows in case this would help to understand what happened, you wrote that the voltage should be way off if the 2A3 is not properly installed.
Here are some pictures.
Thank you very much for your help
David

[Paul Birkeland]

What is the value of the writing on the green resistors? 

Yes, that cap has blown because it was exposed to too much voltage.  The resistors blew for the same reason.  When you get a new cap installed and new resistors installed, you will likely have some voltages that still aren't correct, and this would be consistent with a 2A3 that's damaged. 

You need to fix the blown cap and resistors before messing around with the filament voltage.

[davidnow]

It is written R13 5% 3W

[davidnow]

I guess i may order a new 2A3 tube also...

[Paul Birkeland]

Yes, I would recommend a replacement.  If you look at the 2A3 tube and the shiny part of the glass is white, that's another possibility that could do a lot of damage that's a bit harder to predict.  I had a Stereomour come in recently for repair that had this issue, and the cathode resistors had certainly seen more heat than they normally would.

It is 100% OK to run the amp with your good 2A3 on the side that has a good cap/resistors and no tube in the other 4 pin socket. 

[davidnow]

Hi,

1- i feel stupid but unfortunately i pulled off the 2A3 and don't know now which tube is good or bad. I can't see any white part on the glass either. Is there a way to recognize the good one?

2- are the JJ 2A3-40 a good choice to replace the Sovteks?

Thanks again
David