Hi, I'm trying to repair an old synth...I think I have found the fault, but need your experience to be sure. I'm sorry, but this will be a long post.
So here's the problem: powering on the synth, and maxing the volume you can clearly hear a low note playing.
The synth is equipped with six voice chips which have VCA and VCF. Basically the tone generation is continuous, it's the VCAs job to keep the six voices shut down when no notes are played.
The CVs that control the VCAs are mixed with other CVs, and a 4051 demultiplexes the six VCA CVs which go to the voice chips. Just after the 4051 outputs, and just before the VCA pin of the voice chips there are capacitors connected to ground...in between there is a 1MOhm resistor.
I checked with the oscilloscope the outputs of the 4051 demultiplexer, and noticed that 5 VCA CVs signals are almost flat (just a few mV), the sixth signal is wrong: has greater amplitude, and lots of spikes...it should be flat like the others.
Could be the 4051 the culprit? Are they known to fail? A guy some time ago told me to desolder 1leg of the 1mOhm resistor to be sure that's the 4051 and not the voice chips itself...what do you think?

 

It would be better if kept your postings regarding this matter to the same thread. Here http://forum.allaboutcircuits.com/showthread.php?t=73771

 

Sorry, I opened a new thread because the matter was no more Opamp related...is there a way to change the title of the old thread?

 

Instinct says it's more likely to be the voice chip. Not usual for one output of a 4051 to fail, if it functions correctly otherwise.
Disconnect one end of 2 of the 1M resistors, and use short wire links to reverse the connections temporarily.
Does the fault appear to move to the other voice chip?

(BTW, what model of synth are you repairing? .If it's a 6 voice, and it's a Roland -- suspect the voice chip! )

 

Hmm, interesting! It would be a good and easy test to do before the 4051 replacement. I don't know how to do the short wire links...alligator-clips? Maybe on the resistors end, but on the tiny pins of the voice chips I doubt it could work.
BTW, the synth is a Roland Jx-8P. The voice chip is the IR3R05 (custom Roland chip)...they are less prone to failure in comparison to the 80017A on the Roland Juno's.

 

If you de-solder one end of each of the 2 1M resistors, just solder tack a short piece of wire to each unsoldered resistor leg, and cross wire the other ends into the solder pads the resistors have been removed from. No need to try and solder to the pins.
I must admit when I read '6 voice' I automatically assumed a 80017A fault.. Not the case, obviously. I don't know how easy it would be to get a replacement IR3R05 chip, if you need one?
I'm guessing there aren't 'clones,' like there are for the 80017A assemblies?....

 

Ok, now it's clear...once unsoldered 1 leg of the 1st resistor and before desoldering the other one and crossing wires, I could check the output pin 12 of the 4051 - it's the one connected to the voice that stays on. While the connection between those two is interrupted, if the signal on the 4051 is still "strange" there's something wrong with the 4051, or into the common input it receives (this would be really hard to troubleshoot)

BTW There aren't clones of the IR3R05 as far as I know. THEY are quite rare but with some effort can be found 2nd hand on Internet for around 35€...

 

I'd just pop a new 4051 into its socket, or if no socket then it doesn't take long to unsolder the IC and put a socket in and a new IC.

I have not seen a 4051 fail but they are not that common, in TVs and VCRs they use a 4 quad analog switch to switch the audio in the AV inputs, and they fail on a regular basis. If quad analog switch ICs do fail in appliances it's likely that an eight analog switch IC can fail too.

 

Ok, so I desoldered 1 leg of the 1mOhm resistor, basically the output voltage from that 4051 channel is still weird (lots of spikes where it should be flat like other channels are).
The voice chip makes a lot of noise, but that's because once desoldered the resistor the chip has its VCA pin floating...once grounded the voice is perfectly silent.
So my suspects are still on the 4051, I ordered a new one...when it arrives I'll know if I'm right.