Brymen BM235 InEr issue

[Sebaxtian]

Hello! I have a Brymen BM235 multimeter that worked very well for over 5 years, but as of lately I keep getting InEr error and beep from multimeter even when the probes are not connected. I saw here some threads where it says that is an issue with the firmware, where if you put the switch in Auto LoZ and then quickly move it to another position, the message InEr appears for a second. Mine does this also but the InEr message appears even if i move the switch from Off position to Volts or Ohms. It stays for a second and that's it. I thought it's a problem with the traces on the board and opened it up and checked. The board was a little dirty, I cleaned it with alcohol and greased the traces. I put the board back and after powering it up, I had InEr error message displayed on every position of the switch, with the probes disconnected. After a few cycles of power up and shut down, the error disappeared and hasn't come up since then. I made a photo of the board and uploaded it here. You can see there are two traces that have the green paint washed off of them. Could this be the problem and if it is, how can I fix it?
Thank you.

[sonpul]

If you put the selector on A ; mA or uA is also displayed by InEr? If yes, then you need to check the sockets, in them the contact is divided into two independent parts, a short circuit or leakage is possible.

[Sebaxtian]

Thank you for the reply. The InEr message and beep was present on all positions of the switch, but like I said, after a few On/Off switches, it went back to normal. I also cleaned the sockets and checked them for short-circuit and they are fine. The multimeter works now and the issue didn't manifest itself. But I wonder what caused the issue and maybe there are users here that had this problem and find a fix for it.

[Sebaxtian]

This is a closeup of the traces where the green paint is faded

[Fungus]

This is a closeup of the traces where the green paint is faded

Calling joesmith...

Are they they really relying on solder mask to prevent the switch from making contact? That doesn't seem believable to me.

[ebastler]

Are they they really relying on solder mask to prevent the switch from making contact? That doesn't seem believable to me.

Maybe it does not matter whether or not the switch contacts the trace in that position, since other contacts remain open? They could easily have routed the connecting trace differently if this were critical.

What has me wondering more is that lonesome piece of trace on the inner circle at. approx. the 8 'o clock position, which is apparently not connected to anything. Did that tear off from the main trace at some point?

Edit: Annotated picture added. The trace fragment in the black circle is the one that has me wondering.

[TomKatt]

What has me wondering more is that lonesome piece of trace on the inner circle at. approx. the 8 'o clock position, which is apparently not connected to anything. Did that tear off from the main trace at some point?

Isn’t that a few via holes to the other side?

[Sebaxtian]

That piece seems is in the right place, doesn't wiggle or anything and I didn't found any loose pieces when I took the multimeter apart. Hopefully this issue is not present in the BM789 version, because I just order it, exactly prior to this one giving out errors. Maybe this is a sign... 

[Sebaxtian]

That small piece doesn't have a connection/hole to the other layer of the board

[ebastler]

Ok, that lonely piece of trace is also visible in Dave's teardown (6:34 timestamp):



But the marks on the green soldermask are much less pronounced, if they are present at all.

[Sebaxtian]

But the marks on the green soldermask are much less pronounced, if they are present at all.

Maybe it hasn't been used as much as mine.

[Fungus]

Are they they really relying on solder mask to prevent the switch from making contact? That doesn't seem believable to me.

Maybe it does not matter whether or not the switch contacts the trace in that position, since other contacts remain open?

I guess not. The switches make contact between adjacent tracks so if both contacts touch the same track, that's OK and if there's nothing to the side then that's OK too. Contact pairs circled below.

It just looks bad...

[floobydust]

OP I would clean the inside of the banana jacks. Jack detect is a high impedance circuit using the half-shells and any leakage currents there between them will give troubles.
I use IPA on a Q-tip and stuff it in there, give it a spin to clean out crud, then let it dry out. One of yours (V?) looks black at the top compared to the others, so might be grease or oxide something from your test leads contaminating them.
The two-pin header to the jack board has the mA, A jack detect signal. Make sure it's clean on the jack board as well.

The Brymen rotary switch I think is a poor design. It shows amateur PCB layout and the wiper is cave man tech, I think they could do much better. It also shows in new models  if you look at the traces there, a few mistakes.

[Sebaxtian]

OP I would clean the inside of the banana jacks. Jack detect is a high impedance circuit using the half-shells and any leakage currents there between them will give troubles.
I use IPA on a Q-tip and stuff it in there, give it a spin to clean out crud, then let it dry out. One of yours (V?) looks black at the top compared to the others, so might be grease or oxide something from your test leads contaminating them.
The two-pin header to the jack board has the mA, A jack detect signal. Make sure it's clean on the jack board as well.

I already cleaned the banana jacks with IPA and checked the continuity on the split banana jacks, and it was ok.

[floobydust]

It seemed to get worse after you took it apart and cleaned it, then got better, is that right? I think it just needed to dry off, some IPA has added water.

What lube did you use on the rotary switch? That topic will start another war here. How did the rotor wipers look, I use a paper swipe to clean it.
I thought they used mineral oil on the PCB. I use DeOxit F5 with PPE with no problems, although it has little lubricant in it.
Grease is a bit much - lithium salts conduct a little, silicone grease reacts with nearby components.

I thought the BM235 has had issues with the rotary switch needing a spacer, and the input jacks might crack. Not sure but just to alert that it could be a mechanical problem.

[Sebaxtian]

It seemed to get worse after you took it apart and cleaned it, then got better, is that right? I think it just needed to dry off, some IPA has added water.

What lube did you use on the rotary switch? That topic will start another war here. How did the rotor wipers look, I use a paper swipe to clean it.
I thought they used mineral oil on the PCB. I use DeOxit F5 with PPE with no problems, although it has little lubricant in it.
Grease is a bit much - lithium salts conduct a little, silicone grease reacts with nearby components.

I thought the BM235 has had issues with the rotary switch needing a spacer, and the input jacks might crack. Not sure but just to alert that it could be a mechanical problem.

After the cleaning, the board had enough time to dry, more than half an hour, so I think that wasn’t a problem. To grease the traces I used some grease from the rotary switch, I took a bit from there and spread it on the board, I used just a bit.

[J-R]

Based on the information provided, my 2 cents is that this was only an input jack contamination issue.  The bottom of the jack can collect quite a bit of environmental junk as well as metal shavings, and the delay after cleaning seems reasonable.  Further flushing/cleaning might be a good idea.  And of course both the A and mA uA jacks are able to contribute to the issue.


Are those traces partially worn through or is it just the coating?

I'm firmly in the add lubricant to contacts camp after taking apart a few DMMs and seeing the shocking amount of wear and metal dust.  I used SuperLube a couple times and never had any issues, but now I think I've settled on CRC 2-26.
 Either way, I just apply tiny amounts to only the wear points and it really smooths the selector operation.  My perspective is I'd rather have to clean dried gunk off a board 20 years from now instead of trying to recreate worn contacts/traces.