Loose contact inside potentiometer

[drautzburg]

On my Phonic mixing console the headphone pot is erratic. Sometimes I cannot attenuate the left channel. I reverse-engineered the circuit around the pot and it is a standard inverting opamp design. The pot's one end is connected to GND and abviously this GND connection is cut sometimes. It is not a "dirty middle contact" problem. It is also not a problem with the PCB, because I connected the GND pin to ground using a wire and this didn't change a thing.

So the disconnection must be inside the pot. Even when it works and I turn it down, the resistance between the middle contact and GND is a couple of Ohms as opposed to 0 Ohms on the other channel. I could live with that.

Problem is, I cannot easily replace the pot. This would require removing 100 pot knobs, removing two other PCBs and the PCB in question would still be connected to another PCB via 20 soldered short wires and what not. This would probably cause more harm than good. I can only access the pot from the soldering side and I can wiggle the knob.

The pot looks like an ALPS RK12L1230C0K to me:



When the error occurs I can usually "fix" it by pushing the knob down. I also resoldered all pins and this also fixes the problem temporarily. The strage things is, that once it works it works for several hours and I can use it, turn the knob and all without any indication of a loose contact. . It is as if it "clicks in" into normal mode and stays there. As if a spark temporarily welds the disconnection, but I cannot see why there would be enough electrical power to create a spark strong enough.

Can anybody explain why the problem "clicks" into non-existence?

Does anybody know how these ALPS pots look on the inside and could give some guidance for fixing it in place?

 

[shrtrnd]

Don't have a good answer you'll like. It does sound like the pot is your problem.
How sealed is the pot. Any possiblility of using a 'SAFE FOR PLASTICS", spray contact cleaner in there, to see if that helps until you actually break-down and replace the pot?

 

[drautzburg]

I don't know how sealed it is, I can only see the soldering side. What's puzzeling me is that it is sometimes sufficent to tap one pin with a screwdriver to make the error go away. As if the electcrical signal of the screwdriver fixes it. It is this super-sensitivity, I don't understand. But your explanation is void of any myths and therefore probably correct. And you're right: I don't like it.

 

[shrtrnd]

If you've resoldered the leads, and you're sure the circuit-board trace is not broken near the pot, it sounds like the pot itself is on the way out.
If the inside of the pot is accessible, you may be able to spray it with the thin tube that comes with a can of contact cleaner, and extend its life.
No guarantee, but what I'd try in your predicament.

 

[drautzburg]

But wouldn't that just fix the wiper and not the GND connection? - tell me otherwise.

I had it open today in a state where the error occurred. I just touched one of the pot leads with a screwdriver and the problem went away? Can a loose contact behave like this? It was as if the "signal" from the screwdriver flushed away the problem.

 

[shrtrnd]

I'm not there to twiddle with it. I'm trying to help with diagnosis based on the informtion you report.
Now you're describing something like a static electricity build-up that you're discharging to solve the problem, which would seem unlikely.
This is my input.
You've either got an unreliable potentiometer/physical connection; or you've got a probelm eslewhere in your circut, and your experiences with the pot are merely a symptom of some other problem.
Have you checked any actual ground connections to the pot, or the circuit the pot is connected to?
Does wiggling the pot itself change anything when the symptom occurs?

 

[drautzburg]

Yes, wiggling the pot when the symptom occurs I can often "fix" it. But sometimes it is really stubborn and no amount of wiggling can get it back to work. The ground connection on the board is not the problem because grounding it with a wire didn't change things.

I checked the specs of this pot on the ALPS website and they give limitations of soldering temperature, duration and the number of times you can resolder it (twice). They also say, that there should be no solder on the mounting side (with double-sided boards) These things seem to really sensitive.

 

[shrtrnd]

I know you don't want to do it, but it sure sounds like you're going to have to replace the potentiometer to eliminate the problem.
If you find replacement pot, I'd order a spare or two, so you have them, when this problem occurs with another pot on your board.
That's the only solution I can offer at this point.