My Build - Switch On Looms!

[Shawnobi]

So I've finished the wiring (thanks to all those who have helped with my previous posts).

It's based around an old Selmer O/P tranny and choke. Circuit consists of EF86 > 12AX7 feeding 2 6V6 pentodes with 5Y3 rectification.
It's my first build and it was a bit of a deep end job

So apart from checking and rechecking the wiring, is there anything I can do to double check before powering it up with all the valves in?

[Loogie]

If you don't already have one build a light bulb limiter.

[Shawnobi]

Ignore post.

[Shawnobi]

OK, so some dodgy rectifier wiring fixed. The next problem is the cathode bypass resistor and cap smoke as the amp starts to warm up. So there is a current inrush for some reason.
Any ideas?
Scheme attached:


Also one 6V6 is hot enough to cook with, the 2nd (Attached to the failing bypass cap) only warm.

Mods - Could you perhaps move this thread to the debugging forum?

[g1]

Switch the 6V6's around, does the overheating stay with the tube or with the socket?
If the problem stays with the tube then it is a bad 6V6 and that is why the cap is going bad (over-voltage).
If the problem stays with the socket, then there is a problem with wiring/circuitry.

[Shawnobi]

I need to pick up some more caps.
Until then I noticed that on the scheme on the lower 6V6 there is a wire coming off (poss pin 5?!? - doesnt say) it runs through a 470K and ties into ground and the speaker side of the output tran. Seems I missed it. Could that cause this issue do you think? I'm not quite sure what that's doing. Some kind of feedback signal for the control grid??

[Enzo]

It is the grid return for that tube. it is drawn a little unusually in that they drew a connection in both sides of the grid, but it is still pin 5.

Nothing cosmic, it is just a resistor to ground. Doesn't matter that the speaker is also grounded.

Without that resistor your power tube grid is floating, and yes the tube will run away.


The power tube cathode resistor that burns up, what wattage resistor are you using?

[Shawnobi]

The resistor was originally a 0.25W I've upped it to a 5W. It still melted the cap though. If I reinstate the 470K from pin 5 to ground you think that will solve the problem? Can hopefully swap a new cap in this evening.

[Enzo]

Whether it solves all the problems or not, you need that 470k grid return. Chances are it will stop the tube from running away.

The cathode bypass cap needs to be replaced, but for now you can just remove it. The amp will work without it. It will affect the gain and tone response some, but you don;t need it to see that everything else is working or not. If you have one, great, but if you don't have that cap handy, you don;t need to stop working on the amp.

[Shawnobi]

Whether it solves all the problems or not, you need that 470k grid return. Chances are it will stop the tube from running away.

The cathode bypass cap needs to be replaced, but for now you can just remove it. The amp will work without it. It will affect the gain and tone response some, but you don;t need it to see that everything else is working or not. If you have one, great, but if you don't have that cap handy, you don;t need to stop working on the amp.

Thanks Enzo. I'll probably pop another cap in as it's a cheap warning sign that something isn't right if it melts again! If it stays good I'll feel comfortable leaving it switched on and taking some measurements. I'll report back as soon as I've done it.

[Enzo]

I suspect your cap melted because of the red hot resistor burning next to it.

[Shawnobi]

I suspect your cap melted because of the red hot resistor burning next to it.

The original one failed with the resistor. I then replaced with a 5W and a new cap. The new cap bubbled out it's insides as well, but new resistor seemed fine!

Will see what happens with the 470K fitted

[Enzo]

Oh...

If not from heat, then caps fail from reverse voltage or over voltage. Is it wired correctly? I am thinking that the excess tube current would cause abnormally high cathode voltage, but I wouldn;t think enough to burst a cap, but maybe...

[Shawnobi]

Fitted the 470k resistor, new cap and new 250 ohm resistor but same issue. 250R starts to smoke as the amp warms up. Looks like it didn't take the cap this time but prob would have done.

[Enzo]

OK, remove the power tubes and get out your meter. With no power tubes, measure voltage on pins 3,4,5 of both sockets. Following the schematic from post #4, whatever your B+ voltage is, you should find it at pin 3 of each. The voltage at pin 4 of each ought to be almost the same. Pin 5 of each ought to read zero volts DC. If the pins 3 and 4 voltages are missing, it won't make the current high, it will make it low. But if pin 5 has a positive voltage on it, it will push the tubes screaming into excess current.

Power off. Now measure resistances. From the one pin 5 to ground, you should get the 470k ohms you just installed. From the other pin 5 to ground we ought to see about 477.5k to ground. Do not measure the resistors themselves, measure from the socket pins to ground. And just for fun, measuer pin 8 or each socket to ground, expecting about 250 ohms. I am not worried that the 470k will measure 460k, I am worried that we will find one of those paths open, or if a wrong part was installed, a gross error in value. Trust me, it happens all the time that someone misreads a color code or grabs the wrong part from a bin.

If you are having trouble with that resistor burning, you should clip your meter across it first, then watch the voltage when you power up. I am guessing it will be a lot higher than supposed to be. So you have two power tubes. Pick one and install it in one of the sockets, then power up. Does the voltage across the cathode resistor go nuts or stay at reasonable level? or at least does the resistor start to bake or does it sit there politely? Now remove it and try the other tube in the same socket. Does it behave the same or different? If they both act the same when used by themselves in the same socket, then either they are both bad in exactly the same way, or they are both OK. Now take one of the tubes and put it in the other socket. Does it work the same there or does changing sockets change the results.

What we just did was test each tube and each socket in a very crude way. You either have a bad tube (or two), or there is some problem with one or both sockets. We just need to find out which of those is the case.