Need ideas

[ricach]

What gremlin am I fighting here.......
Here's my situation with a particular preamp tube/socket:
12AX7 is split between channel 1 (V2-B) and channel 2 (V2-A) (switching, not physically two inputs)
V2-A works 100% and V2-B works 60% of time. Wiggle the tube in the socket and V2-B kicks on (and off). Heater is suspected because tube glows brightly when both channels working, only half bright when just one channel works. As well as sound fades out and in.

Ok so cleaning tube sockets makes no difference so I change out tube with new one(s) and it requires lots of wiggling to get tube in a position where either channel works at all. 3 different tubes have been tried. Yet when I pop the original back in, it works most of the time. WTH?

[Enzo]

So the tube socket is not making good contact with the pins on the tube. It may be electrical "dirt" you missed, or it could be the female pins have lost their grip. Tighten them. Or replace the socket.

[ricach]

I use Caig DeOxit stuff and it's always worked wonders cleaning 'electrical dirt'. Usually I wouldn't hesitate to just replace a socket, but in this case that's a monumental task so tightening the pins looks to be the next viable option. Although the pins feel tight in general, I should realize that doesn't necessarily mean they're all tight. . Thanks.

[ricach]

After tightening the sockets and recleaning them, the amp worked great for a weeks worth of practicing. But then it started fading out and back in. Its the same preamp tube, but wiggling it does not change anything. It works for a few seconds then fades out, then in a couple seconds fades back in. It keeps this up all the time now. The heaters are glowing the same 'glow' throughout so its not the elements cooling and heating. But the rate of 'fade' is identical to tubes warming/cooling. It doesn't stay on long enough to generate so much heat to expand/contract contacts, or circuit traces. Or does it?

UPDATE:
Ok, switched tubes around, now back to the 'wigglin' problem, but no fading in and out. If I swapped tubes back, then back to fading problem. This is weird.

[olddawg]

Is this a hard wired or pc mount socket? Sounds like you need to trouble shoot to me. Take voltage readings in failure mode. You could try resoldering everything because the intermitancy may not be the socket. Wiggling the tube may just be stressing something in the general area. A scope would be helpful for this problem also.

[ricach]

PC mount. I've already done the troubleshooting route once and thought I fixed this. Used a scope and while in failure mode it was obviously the one preamp tube causing the problem. And even a pot seemed to trigger weirdness. I even used some cold shot to determine any heat related issues. With some resoldering and cleaning of the socket and pots, it worked flawlessly for a couple days. So I put it back in the cabinet and played it for another week before this happened. Which I believe is just part of the original problem it had.

I've swapped around more tubes and have a combination that works flawlessly right now. Go figure.

[Enzo]

You have an intermittant connection somewhere. If the socket itself is not loose or dirty, then the solder to it is suspect, if the solder has been redone, then we look for a cracked circuit board trace. A solder pad may have broken off from the end of a trace, but the edges still usually touch, so it sometimes works. All of that is consistent with your symptoms. Pushing on a tube, a control, a resistor, or most anything also applies some flexing to the board, so moving even an unrelated part can still cause the symptom to change.

Troiubleshooting will always get you to the problem. We find what voltage or signal is going away and then find where along the path that is disappearing. If the heaters are going out on a tube, but not the others, then there are only so many places it can be. A 12AX7 heater has three connections. Most times they are wired for 6v, so pins 4 and 5 are wired together. In 12v operation, pin 9 is left unused. We look at each end of that heater and see if the pins at either end are losing the supply voltage. If we find one that is, we trace to the previous point where it was good and look for the break in between those two points. Or if we know A and B are wired together but their continuity is intermittant, just wire a jumper across and be done with it.


You could indeed have a bad socket. Pins "feel tight"? You can't feel one loose pin. Are any of the female contacts spread? Your socket could be broken. I have run into this a few times, though it is rare. The female pin is a little metal flag rolled into a circle to grip a tube pin. The "staff" of that flag goes down through the socket body and comes out the other side as a solder tab or pc pin. Those pins can snap in half inside, so the female part still works and may be as tight as you like, and the solder pin is still soldered in nice as can be, yet the operation bepends upon the two halves of the brokebn pin touchinjg each other.

[Enzo]

You have an intermittant connection somewhere. If the socket itself is not loose or dirty, then the solder to it is suspect, if the solder has been redone, then we look for a cracked circuit board trace. A solder pad may have broken off from the end of a trace, but the edges still usually touch, so it sometimes works. All of that is consistent with your symptoms. Pushing on a tube, a control, a resistor, or most anything also applies some flexing to the board, so moving even an unrelated part can still cause the symptom to change.

Troiubleshooting will always get you to the problem. We find what voltage or signal is going away and then find where along the path that is disappearing. If the heaters are going out on a tube, but not the others, then there are only so many places it can be. A 12AX7 heater has three connections. Most times they are wired for 6v, so pins 4 and 5 are wired together. In 12v operation, pin 9 is left unused. We look at each end of that heater and see if the pins at either end are losing the supply voltage. If we find one that is, we trace to the previous point where it was good and look for the break in between those two points. Or if we know A and B are wired together but their continuity is intermittant, just wire a jumper across and be done with it.


You could indeed have a bad socket. Pins "feel tight"? You can't feel one loose pin. Are any of the female contacts spread? Your socket could be broken. I have run into this a few times, though it is rare. The female pin is a little metal flag rolled into a circle to grip a tube pin. The "staff" of that flag goes down through the socket body and comes out the other side as a solder tab or pc pin. Those pins can snap in half inside, so the female part still works and may be as tight as you like, and the solder pin is still soldered in nice as can be, yet the operation bepends upon the two halves of the brokebn pin touchinjg each other.

[ricach]

One thing I left out was that I can pull, push, wiggle, twist, no matter what i do the amp doesn't skip a beat with this tube. Put other tubes in there and it's back to the wiggle problem. I compared the tubes and no visible difference in the pin sizes. However, it is an ECC83, which is what the schematic calls for, the others are 7025's.

Enzo, I know I should put it back on the bench, but I think I will play this amp till it fails with this tube. This is a very difficult chassis to remove, lol.

[Enzo]

We had a problem with some JJ tubes some long time ago, the tube pins were coming consistently slightly undersize, so the tubes didn;t get reliable contact in the sockets. Pin sizes are nominal, so some can be ever so slightly fatter than others. And thus make better contact.

The tube "type" has nothing to do with it. ECC83 is simply the Euro number for 12AX7. 7025 is also just a 12AX7, except the original 7025 had a more tightly controlled noise figure - it is a less noisy 12AX7. Current production 7025 are just 12AX7. COmpanies that sell all three are just doing marketing. Point I am heading to is that a "wrong tube" in the socket may perform less well, as in lower gain, higher noise or something, but that difference will not cause the "wiggle drop out" symptom.

There is a limit to how much one can tighten up a socket. And some sockets just don't lend themselves to it. In the old days, a TV guy might find a tube with loose socket pins. He MIGHT be abloe to tighten the pins, but to do so, he'd have to pull the entire TV chassis out to get at it. An old TV guy trick for a loose socket was to stick just the very tips of the tube pins into the socket holes, then give the tube a slight twist. This made all the tube pins angle slightly. Now when pushed into the socket, those angled pins shoved against the sides of the socket pins and thus made more contact. SO if you have a loose socket that one tube seems happy in, it may be that that tube has some pins at a small angle or with pins a little kinked, so they make better contact.

I don't recommend that old guy trick, because as much as it might save some time and effort on his service call, it also helps spread the socket pins even more. Sort of a pay me now or pay me later thing.

[Enzo]

We had a problem with some JJ tubes some long time ago, the tube pins were coming consistently slightly undersize, so the tubes didn;t get reliable contact in the sockets. Pin sizes are nominal, so some can be ever so slightly fatter than others. And thus make better contact.

The tube "type" has nothing to do with it. ECC83 is simply the Euro number for 12AX7. 7025 is also just a 12AX7, except the original 7025 had a more tightly controlled noise figure - it is a less noisy 12AX7. Current production 7025 are just 12AX7. COmpanies that sell all three are just doing marketing. Point I am heading to is that a "wrong tube" in the socket may perform less well, as in lower gain, higher noise or something, but that difference will not cause the "wiggle drop out" symptom.

There is a limit to how much one can tighten up a socket. And some sockets just don't lend themselves to it. In the old days, a TV guy might find a tube with loose socket pins. He MIGHT be abloe to tighten the pins, but to do so, he'd have to pull the entire TV chassis out to get at it. An old TV guy trick for a loose socket was to stick just the very tips of the tube pins into the socket holes, then give the tube a slight twist. This made all the tube pins angle slightly. Now when pushed into the socket, those angled pins shoved against the sides of the socket pins and thus made more contact. SO if you have a loose socket that one tube seems happy in, it may be that that tube has some pins at a small angle or with pins a little kinked, so they make better contact.

I don't recommend that old guy trick, because as much as it might save some time and effort on his service call, it also helps spread the socket pins even more. Sort of a pay me now or pay me later thing.

[km6xz]

Enzo is right, pin diameter varies with brand(not label but actual manufacturer name, since the hundreds of tube "brands" come from only 1/2 dozen actual manufacturers).
It seems that a lot of guess work and time is being spent when a little testing in-circuit would pin-point it quickly. ANY change in signal should be accompanied by observable parameter changes. When the signal fades out, for example, does the tube static plate current change, or is it only changing when a dynamic signal is present?
It really sounds like a tube socket with worn loose receivers but it could be from other causes, all of which would give evidence of the fault to metering or scope displays. Transistors, tubes and ICs do not need testers or to be pulled to test, they will give all the evidence needed to prove they are good or bad by their operating characteristics in-circuit.