Crackling sound caused by flux residuals! Is that really possible?

[Sören_Marodören]

I have just recently repaired a Technics RS-1500U Reel to Reel Tape Recorder.
It was recapped by its owner and worked well for some time before crack and pop sound suddenly started to appear. The owner has only used high quality electrolytic capacitors and is highly skilled in soldering. So the crackling sound was a surprise for him.
The attached video (Crackling_sound.mp4 stored in Cracklin_sound.7z) shows the sound from the headphone connector on an oscilloscope and how it sounds like.
When I got the tape recorder I started to look at the usual culprits like switches, potentiometers and capacitors. But the fault was not related to them. And due to the nature of the problem (the problem was most frequent after power on) it was very difficult to find the origin of the problem.

After some time the problem was isolated to the headphone amplifier. Everything looked visually ok on the board and in-circuit measurements did not indicate any faulty components.  The next step was to replace one component at a time to see when the fault disappeared. At the end all components (transistors, resistors, caps and inductors) were replaced by new – but the problem was still there. (See Schematic_Headpohone_Amp.JPG for the schematic on the headphone amplifier.)

At this point there was only one component that was not replaced, the PCB. But I have never heard of PCBs that causes this problems. Suddenly I realized that there could be one more common “component”, flux residuals. Nope, the board was not cleaned by me or the owner during our work. I have never had problems with flux residuals before, as long as it looks nice and clean. (If the residuals look dark or if there are plenty of them I clean the area with IPA.)

Therefore I started to clean the left channel of the headphone amplifier, and the problem disappeared immediately. To be sure that my test setup was still working I switched over to the right channel and could hear that the problem was still there. This time I cleaned the right part of the amplifier and eliminated the problem at that side too.

In the picture “Half_cleaned_Board.jpg” you can see how it looked half way through the cleaning where left channel was cleaned (right side if the board) but the right channel was still to be cleaned. As you can see there is not much residual on the right channel.
This board was manufactured in the early 80´s and the original residuals are still on the board. When the board was recapped, new flux residuals were mixed with the old and probably created a mix that was susceptible to breakthroughs that resulted in the noise. If someone has a theory (or practical experience) on how this can happen I hope you can describe it in this thread.

During my work on this tape recorder I looked for hints and suggestions on Google. But could not find any that pointed me to clean the board. Instead there were plenty suggestions on potentiometers, switches, capacitors and other traditional components. I did see a few mentioning that semiconductors can cause this problem. Now when I have seen what flux residuals can do I start to suspect that the root cause might not been the semiconductor itself. Probably they replaced the semiconductor and cleaned the board, and solved the problem.

[Gyro]

The solder joint on the solder side resistor near TR231 looks very dodgy.

[drvtech]

I've had precisely this problem with flux trapped under an MLCC in a Mackie amp. Like you I had hunted high and low for the problem but eventually cured it with a squirt or two of IPA.

[David Hess]

I have heard of uncleaned flux causing that problem but never had an opportunity to investigate it myself.  I would normally suspect a cold or cracked solder joint first which can be tested by flexing the board.

[S. Petrukhin]

You probably used a defective flux that is conductive. For the high-resistance input of the amplifier, it was sensitive and caused excitement. The old flux is most likely rosin - it is harmless. On the japanese PSB never washed the flux at all. 

[Berni]

Yep flux becomes slightly conductive after it has absorbed enough dirt or moisture

I never seen it become conductive enough to act as a short on a power rail or anything, but it can definitely leak some microamps between components, enough to upset sensitive analog circuity. With enough voltage and dirt it could even leak miliamps.

I seen multiple cases where a board started working after cleaning the flux off it. But its usually from things where some hand soldering or rework was used. The paste used for SMD soldering generally does not cause problems since there is so little of it and the kind of flux matters too.

[Le_Bassiste]

https://www.eevblog.com/forum/metrology/very-slow-dmm-offset-settling-humidity-effect/msg2643525/#msg2643525