PCB's becoming conductive is such a frustration. I like your point to point fix, but it is a TON of work. Historically, that's the way it once was done, and worked for decades.
Some PCBs show an effect called "Conductive Anodic Filament" (CAF) which reduces the isolation resistance. But it is observed especially at higher voltages. Maybe the PCB contains some impurities leading to a migration of copper ions.
Actual experience: Equipment was used at a seaside location with humid salty air, and salt laden moisture was absorbed into the PCB. Cleanings had some effect, soaking and baking had more effect, but the only permanent solution was to replace the PCBs and apply conformal coating. The PCB material was that tan porous "consumer electronics" junk though.
My first thought was bad vias without proper clearance to the ground plane. It looks like almost every trace goes through a via (on the back of the board anyhow). Really poor PCB quality at any rate. If you do end up redesigning the PCBs and post the files online, it may help out some other people that have those same devices. Because with three bad examples on your bench alone, there a probably thousands (?) of faulty units out there. Either way, glad to see you posting a lot of videos recently. Longtime fan of the channel!
Or maybe a poor quality substrate that absorbed moisture? I.e. if you stuck one of the boards in your oven at like 80-90°C for an hour (or left it in a bag with desiccant for a few days), would the inter-trace resistance go down.
It looks like defective or poor quality PCB material, or perhaps the PCB material was contaminated with something conductive during manufacture. In my line of work I came across circular multi-pin connectors from a well known manufacture that had defective/conductive rubber grommets, in the few megohms to tens of megohms like your PCBs. The rubber had a manufacturing defect so such things can happen. I like the idea of making replacement PCBs, there must be many more units out there that need this repair. Your point to point wiring turned out nicely but no doubt time consuming.
Once we used a _no clean_ Flux on a job that required high reliability. It was a nightmare. OC insisted the no clean Flux needed not to be cleaned. It turned out this specific Flux likes to soak up humidity, and then it becomes quite conductive. We measured on unrelated signal trace resistance as low a one kilo Ohm. Try again cleaning the boards, perhaps with Isopropyl in an ultrasonic bath.
maybe the PCB's absorbed moisture somehow? Depending on what kind of material they were made of, I suppose it could be a possibility. If you put them in oven to get the moisture out then....
Or use a hairdryer, at least to see the if it changes anything. Maybe your "free wiring method" is the way to go, on a cold winter night. Happy New Year!
Good point. I'd try drying them in an oven and if that fixes the problem, then I'd spray them with heavy coats of conformal coating to keep them dry permanently.
@@laoluu I'm not sure that a hairdryer would do much. I'd follow the procedure for bake drying surface mount chips before reflow soldering to prevent them from popcorning (where moisture in the epoxy package turns to steam during reflow and blows a chunk out) - usually it requires baking for several hours at temps over 100C to drive all the moisture out.
My guess is that there is moisture in the circuit board.Try to remove all connectors and cook it in an oven at like 140c see if that works ,I've never seen this one before Good find.
Wow! makes me think of carbon loaded plastics, like all this black around, but you tested the PCB with nothing else connected I think (will run it again) so these huge resistances are an artifact of the PCB manufacture. Can you stick a bare PCB in the oven 100 C overnight and bake it ? But again the big change with the wired version was the black connectors so I dunno....but personally I would hit Kicad and get one made ! surely less painful than wiring that lot up.
My first hypothesis was hygroscopic flux as well, as I've had it happen before in my multimeter. But this issue must be the related to the boards themselves. PCBs do absorb moisture as well. At the assembler where I work we dry PCBs before reflow soldering if the package has been open for too long. Obviously these boards are just wave soldered, but the same moisture absorption can happen over time, I guess. I've never seen a device fail because of conductive boards before, though. Suggestion: Put them in a small, temperature controlled oven at 60-70C (140-158F) for 24 hrs. See if it makes a difference.
@k1mgy Most, if not all PCB manufacturers are located in China. If we order from a company in our own country, they are just a middleman. PCBs that are made for the commercial market get vaccuum packed with a humidity indicator, just like SMT part that have a MSL. Due to the reflow process any moisure in a component or PCB can rapidly expand and destroy it. PCBs can delaminate. As I said I 've never heard of a finished product with boards becoming so conductive. It is the only theory here, though.
Maybe the flux wicked up the conductors and is on the other side of the PCB as well and was protected during your cleaning by the connectors? I would try cleaning the board you removed the connectors from again with regular soap and water, if that does not fix it I'd try acetone and if that does not fix it maybe put it in an oven for 24 hours and if that still does not work I'd design new PCB's, very interesting video :)
I have had a LED dimly lit for no reason, but after a PCB wash it fixed it. Don't know why your wash didn't work, unless it's something insolvent with what you washed it with, or the PCB is maybe porous and absorbed something - maybe possible on the edges, but not very likely in the centre -.
I would not try to dry the PCBs in an oven because it is quite likely that the problem will reoccur. In my opinion, there are two solutions: replace the PCBs completely with wire connections or have new, high quality PCBs made.
I would try to unsolder everything, then a bath of pure acetone, preferably with ultrasound, then a bath of pure alcohol, and let it dry in the oven at 60ºC for 2 hours.If this fix it, after resolder everything, put some uv top coat.
Make new PCBs, once you’ve got them you can repair as many as you need in very little time, maybe even just make your own and not bother with that junk at all.
i think to design a PCB for them is a good idea as they are garbage and the used market is probably going to be flooded of that model used but almost brand new not working so your friend if he need more can get them for almost nothing plus the price of the repair
If you have a scrap board available: remove the solder resist on a channel and see what happens. Looks a bit like some bozo put a can of 'anti-static' paint in the machine that applied the solder resist?
Are you seeing the same resistances across a dismantled bare board? Maybe better make your own boards? You can get xlr type rck cases, maybe make your own?
PCB's becoming conductive is such a frustration. I like your point to point fix, but it is a TON of work. Historically, that's the way it once was done, and worked for decades.
Some PCBs show an effect called "Conductive Anodic Filament" (CAF) which reduces the isolation resistance. But it is observed especially at higher voltages. Maybe the PCB contains some impurities leading to a migration of copper ions.
We missed your meticulous investigation on fixing electronic test equipments.
Actual experience: Equipment was used at a seaside location with humid salty air, and salt laden moisture was absorbed into the PCB. Cleanings had some effect, soaking and baking had more effect, but the only permanent solution was to replace the PCBs and apply conformal coating. The PCB material was that tan porous "consumer electronics" junk though.
My first thought was bad vias without proper clearance to the ground plane. It looks like almost every trace goes through a via (on the back of the board anyhow). Really poor PCB quality at any rate.
If you do end up redesigning the PCBs and post the files online, it may help out some other people that have those same devices. Because with three bad examples on your bench alone, there a probably thousands (?) of faulty units out there.
Either way, glad to see you posting a lot of videos recently. Longtime fan of the channel!
Or maybe a poor quality substrate that absorbed moisture? I.e. if you stuck one of the boards in your oven at like 80-90°C for an hour (or left it in a bag with desiccant for a few days), would the inter-trace resistance go down.
Maybe take the unpopulated pcb's and put them in a oven for a while and see if it dries out any moisture in them
It looks like defective or poor quality PCB material, or perhaps the PCB material was contaminated with something conductive during manufacture. In my line of work I came across circular multi-pin connectors from a well known manufacture that had defective/conductive rubber grommets, in the few megohms to tens of megohms like your PCBs. The rubber had a manufacturing defect so such things can happen. I like the idea of making replacement PCBs, there must be many more units out there that need this repair. Your point to point wiring turned out nicely but no doubt time consuming.
Once we used a _no clean_ Flux on a job that required high reliability.
It was a nightmare.
OC insisted the no clean Flux needed not to be cleaned.
It turned out this specific Flux likes to soak up humidity, and then it becomes quite conductive. We measured on unrelated signal trace resistance as low a one kilo Ohm.
Try again cleaning the boards, perhaps with Isopropyl in an ultrasonic bath.
Tough soldering job - well done!
I guess they used cheap PCB material, might be low quality / bad CEM 3.
The problem could be in an inner layer if it's a multi-layer board. It would be interesting to see it x-rayed.
no inner layers here. simple 2-layer board
maybe the PCB's absorbed moisture somehow? Depending on what kind of material they were made of, I suppose it could be a possibility. If you put them in oven to get the moisture out then....
oven would be my first attempt. Second would be scan and re-draw them in kicad, and get new ones made
Or use a hairdryer, at least to see the if it changes anything.
Maybe your "free wiring method" is the way to go, on a cold winter night. Happy New Year!
Good point. I'd try drying them in an oven and if that fixes the problem, then I'd spray them with heavy coats of conformal coating to keep them dry permanently.
@@laoluu I'm not sure that a hairdryer would do much. I'd follow the procedure for bake drying surface mount chips before reflow soldering to prevent them from popcorning (where moisture in the epoxy package turns to steam during reflow and blows a chunk out) - usually it requires baking for several hours at temps over 100C to drive all the moisture out.
My guess is that there is moisture in the circuit board.Try to remove all connectors and cook it in an oven at like 140c see if that works ,I've never seen this one before Good find.
If the PCBs in the breakout boxes are bad, what about the PCBs in the mixer itself?
For sure I'd contact the manufacturer (if possible) and just point them to this video...very disappointing and ever more so frustrating!
Wow! makes me think of carbon loaded plastics, like all this black around, but you tested the PCB with nothing else connected I think (will run it again) so these huge resistances are an artifact of the PCB manufacture. Can you stick a bare PCB in the oven 100 C overnight and bake it ? But again the big change with the wired version was the black connectors so I dunno....but personally I would hit Kicad and get one made ! surely less painful than wiring that lot up.
My first hypothesis was hygroscopic flux as well, as I've had it happen before in my multimeter. But this issue must be the related to the boards themselves. PCBs do absorb moisture as well. At the assembler where I work we dry PCBs before reflow soldering if the package has been open for too long. Obviously these boards are just wave soldered, but the same moisture absorption can happen over time, I guess. I've never seen a device fail because of conductive boards before, though. Suggestion: Put them in a small, temperature controlled oven at 60-70C (140-158F) for 24 hrs. See if it makes a difference.
"PCBs do absorb moisture as well."
CHINA.
@k1mgy Most, if not all PCB manufacturers are located in China. If we order from a company in our own country, they are just a middleman.
PCBs that are made for the commercial market get vaccuum packed with a humidity indicator, just like SMT part that have a MSL. Due to the reflow process any moisure in a component or PCB can rapidly expand and destroy it. PCBs can delaminate.
As I said I 've never heard of a finished product with boards becoming so conductive. It is the only theory here, though.
Maybe the flux wicked up the conductors and is on the other side of the PCB as well and was protected during your cleaning by the connectors? I would try cleaning the board you removed the connectors from again with regular soap and water, if that does not fix it I'd try acetone and if that does not fix it maybe put it in an oven for 24 hours and if that still does not work I'd design new PCB's, very interesting video :)
Bad PCB process. this is not repair. This is REMAKE!!!
I have had a LED dimly lit for no reason, but after a PCB wash it fixed it. Don't know why your wash didn't work, unless it's something insolvent with what you washed it with, or the PCB is maybe porous and absorbed something - maybe possible on the edges, but not very likely in the centre -.
I would not try to dry the PCBs in an oven because it is quite likely that the problem will reoccur. In my opinion, there are two solutions: replace the PCBs completely with wire connections or have new, high quality PCBs made.
I would try to unsolder everything, then a bath of pure acetone, preferably with ultrasound, then a bath of pure alcohol, and let it dry in the oven at 60ºC for 2 hours.If this fix it, after resolder everything, put some uv top coat.
PCB moisture contamination => Put the PCB into ventilated oven and test. If wrong change PCB
Can this be described as anything other than shoddy design and manufacturer?
Make your own box and call it FeedbackNoLoop 😁
Perhaps contact the manufacturer or go with another one. This is not acceptable.
Make new PCBs, once you’ve got them you can repair as many as you need in very little time, maybe even just make your own and not bother with that junk at all.
Replacement PCBs
i think to design a PCB for them is a good idea as they are garbage and the used market is probably going to be flooded of that model used but almost brand new not working so your friend if he need more can get them for almost nothing plus the price of the repair
If you have a scrap board available: remove the solder resist on a channel and see what happens. Looks a bit like some bozo put a can of 'anti-static' paint in the machine that applied the solder resist?
Incredibly poor PCB quality. I think that is the cheapest and easiest way to go. Probably not the quickest however.
Hate to say it, but you get what you pay for. Tell them to buy hand-wired boxes that work, instead of these 'modern' failures-in-waiting.
Are you seeing the same resistances across a dismantled bare board? Maybe better make your own boards? You can get xlr type rck cases, maybe make your own?