1
\$\begingroup\$

I bought a NIB (new in box) cassette recorder on eBay recently but have had problems with the audio from the start. The issue was that touching the dials could cause the audio to crackle, go quiet, or cut out entirely.

It took some time and a lot of triage to narrow the issue down to something unexpected. The problem is with the screw that connects the PCB ground to the case. The reason it appeared to be the dials was because the PCB was being nudged slightly against this screw (it was very confusing to locate this problem!).

Here is an overview of the PCB:

enter image description here

This is the point at which the PCB ground plane connects to the metal inside (the housing is all plastic):

enter image description here

The issue is when I add the screw here and begin to tighten it. Screwing it in enough to achieve continuity between the case and the ground plane does not create any noticeable distortion. But if I continuing to tighten it then the audio will begin to crackle, become very quiet, and then cut out entirely. This only affects the audio signal; the tape continues to play as the motor spins.

enter image description here

I tried adding a steel washer to ensure a good contact with the ground plane but it made no difference with the previous description:

enter image description here

I made sure that this was not a fault of the PCB due to pressure or flex. I flexed the board manually and applied pressure to the location where the screw would be with my finger as well as the back of my plastic screwdriver. The audio is perfect without the screw being present but this also remove all continuity with the case:

enter image description here

What I don't understand is why this only happens if the screw is tightened. Simple continuity with the case does not cause the audio to cut out. For example, the tape is playing just fine as I bright the ground plane with the case here:

enter image description here

This issue occurs whether the cassette player is powered by batteries or the DC barrel jack. The player also has an option to be powered by AC with what I think is an AC/DC converter inside (not visible in the photos).

I could use a non-conductive screw or a plastic washer to "fix" the problem but then the case is no longer grounded.

I am interested in a solution that solves the problem safely but I also want to u the audio cuts out only when the screw is tightened and not just from achieving continuity with the case.

I am new to electronics. Shouldn't the case always be connected to ground for safely in the event that it somehow becomes live?

UPDATE

I added a small piece of heat shrink between the screw and the ground plane contact to test for sure whether or not this was a mechanical issue. I was able to tighten the screw fully without any of the previously described issues. This does not appear to be a problem with mechanical stress around the area. Continuity between the screw and the case was confirmed as well as no continuity between the screw or case with the ground plane of the PCB when the heat shrink was used.

enter image description here

UPDATE

The answer provided by JRE indicated the cause of the problem. Sure enough, there was a large filter capacitor on the reverse side of the board that was just barely touching another metal standoff. The degree of which it touched probably changed as I handled the board but it would always short itself when the ground screw was fully tightened.

I bent the capacitor back and the problem went away. It was probably some kind of filter capacitor that connected up with the speaker; grounding it caused the audio to cut out.

I'm glad that I tested it only with batteries or my fused power supply connected. I guess that means that the chassis could have been live if I'd used the AC adapter?

\$\endgroup\$
5
  • \$\begingroup\$ Still sounds like a mechanical issue to me. Inspect the traces and solder joints (even cracked components) with a microscope if possible. \$\endgroup\$ – rdtsc Jul 28 '20 at 12:37
  • \$\begingroup\$ I have inspected the traces, solder joints, and even removed and reapplied several "suspect" connections at various pads with no improvement. Flexing the PCB with the screw removed and poking at various components with my spudger does not cause any problems. I used an Amscope with WF10X/20 lenses for a very close inspection. \$\endgroup\$ – Zhro Jul 28 '20 at 12:43
  • \$\begingroup\$ What is on the other side of the board near the screw? \$\endgroup\$ – Tom Carpenter Jul 28 '20 at 12:54
  • \$\begingroup\$ If due to "quality" of ground, this would be a very unusual failure mode. The only electrical difference between a "snug" screw and a "tight" screw is a tiny change in resistance and inductance. Try a star-washer: should make better connection with less force. If so, very odd. \$\endgroup\$ – rdtsc Jul 28 '20 at 14:47
  • \$\begingroup\$ @Zhro: AC adapters are usually isolated. Line voltage on one side, low voltage on the other, galvanically separated in the middle by a transformer. With a properly functioning AC adapter, the low voltage side will not have any connection to the high voltage side - it is safe to touch things connected to the low voltage side. \$\endgroup\$ – JRE Aug 27 '20 at 5:30
3
\$\begingroup\$

Your PCB is attached to a metal plate. There are components on the PCB between the PCB and the metal plate.

Very likely there is a component that is just a little too tall. As long as the screw isn't completely tight, there's a tiny gap between the component and the plate. If you tighten the screw really tight, the component touches the plate and causes a short to ground.

Take out the PCB, turn it over, and look for parts that stand up higher than the others. See if those "too tall" parts can touch the metal plate.

\$\endgroup\$
2
\$\begingroup\$

Obviously, it isn't the electrical connection to ground that's causing the problem. Instead, the physical flexing of the PCB is either causing an open or a short somewhere nearby. In older equipment, it would not be surprising to find a hairline crack in the PCB itself or one of the nearby solder joints.

Or, it could be that the tip of the screw itself is touching something that it shouldn't — possibly a wire whose insulation has been abraded.

\$\endgroup\$
2
  • \$\begingroup\$ Flexing the PCB or applying force using a screwdriver (as mentioned in the question) does not trigger the problem. I don't see how the PCB could flex from this screw either. It's just ground plane in that area and it doesn't make a difference if I distribute the force with a washer. \$\endgroup\$ – Zhro Jul 28 '20 at 12:18
  • \$\begingroup\$ The screw enters a tall metal standoff on the reverse side of the PCB. It is completely enclosed. \$\endgroup\$ – Zhro Jul 28 '20 at 12:32
0
\$\begingroup\$

Sounds like an issue with the potentiometers the dials are manipulating. The sound issues you are having are common symptoms of a dirty or failing potentiometer.

\$\endgroup\$
3
  • \$\begingroup\$ The cassette player is brand new old stock and never used. They were also treated with DeoxIT FaderLube to help with any oxidation. There is no issue with turning, jiggling, or otherwise manhandling of the potentiometers when the screw mentioned in the question is removed. What appeared to be a problem with the potentiometers was a result of movement of the PCB at the site of the screw. \$\endgroup\$ – Zhro Jul 28 '20 at 12:41
  • \$\begingroup\$ If you are certain the screw and grounding is the issue, then check the solder points near the screw. Old through-hole solder joints can break down over time and appear fine until you look very closely. Check where the legs of the components come through the bubble of solder and try lightly applying pressure to each leg. See if anything moves. \$\endgroup\$ – David Jul 28 '20 at 12:57
  • \$\begingroup\$ This is not an issue with the amount of pressure exerted by the screw but rather how well it makes contacts with the ground plane. \$\endgroup\$ – Zhro Jul 28 '20 at 13:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.