I have a PCB I'd like to repair - but the whole thing on the component side is potted in what appears to be epoxy resin. Anything that could dissolve it while leaving a chance of identifying the components? I don't have a circuit otherwise I might have just knocked up a new one - it's not terribly complicated.
Dichloromethane is the 'canonical' solvent. Also nitromethane - as used in model glow engine fuel. There is also the 'hammer and chisel' method.
here's a thread (about depotting a hifi amplifier module) from which you might be able to scrape a bit more info.
I once turned epoxy to soft putty using steam. Almost all components would survive that ok, given a long dry out afterwards.
NT
Urgh, can't stand it when manufacturers do that. I've had some luck in identifying passives in-circuit via the solder side and then grinding the resin away to reveal details of semiconductors, but it's risky - very easy to erase markings.
Do you know what the module's *supposed* to do - i.e. can you not ditch it altogether and rig up an alternative that does the same job?
Excellent - I'll give it a try.
It is subject to - I'd say - severe impact energy.
Yes - I'm guessing it's the output device that has failed, but would need to identify it.
I can buy a new module for 60 quid - but it's not worth it. So anything other than a small amount of work needed to fix it will mean the whole lot's junk.
Not a damned thing you can do. Its easier to repair a chip...
Nitromethane wont touch epoxy. Which is why we use it as a proofer against nitromethane containing fuels..
Dont hold your breath. I've *cured* soft epoxy using a warm oven..
Not much. "Potting compound" is such a useful substance that it's generally used to both pot components, and to make much of those components in the first place. It's very difficult to extract things in a way that leaves recognisable forensics.
Where I've done this before and known broadly "what" but not "what value" I've usually ended up baring the solder-side enough to see the circuit, then slicing the component side apart into anonymous black cuboids that I could measure the values of individually. This takes about as long as you'd expect...
If you're just replacing an output stage, then simply guess at what a good design would be and try it. Bipolar or VFET, + or -? Hfe isn't usually critical enough to care. Swap a whole matched pair or H-bridge if you have to.
Being a (somewhat ex-) aeromodeller, I have wondered about this myself. However I definitely read it originally in an old copy of either aeromodeller and/or RCM&E, as well as having seen it referred to since. You'll see that the thread I referred to also independently mentioned using nitro. Perhaps potting epoxies are different? I would be interested to know more...
J^n
Nitromors or polystripper will do.
But IGWS will strip painted components if it gets to touch them.
Derek
No. That black stuff is a bugger. Best bet? Chisel. Been there, done that.
D'oh! That's why they encased it in epoxy ;-)
Cripes, many years ago went down this path. Basically destroyed dozens of "ranson holding" items to extract the underlying circuit. Hacksaws and chisels worked best.
Oh, the ransom holding company had gone bust, taking a couple of million quids worth of IP with them. Fair game. Sorted.
Al.
40 years ago we used a product to dissolve potted epoxy. We mainly used it to take apart epoxy potted transformers so that we could rewind them. It would dissolve fingers etc and had to have a layer of water over the top in its container. It would dissolve almost anything except for metal and glass, and of course it took the paint off the components. We were able identify the components and test them and manufacture our own circuits, which we potted in epoxy.
I can't remember the name of the product but this site may help:
Indeed, and last year I used boiling water to set some quickly. Its one apparent contradiction I've never gotten to the bottom of. All I can say is it worked. This was in the 90s when I knew little about the stuff. If anyone can explain, please do!
NT
Oh dear. But I'm not surprised.
The solder side is clear, and there are only a handful of components. But I'm not going to waste much time on it.
I'd guess it's an Triac.
To make you feel happier many modern car common rail turbo diesel engine management computers are potted... and I'll let you figure out how much the dealers want to charge for a recent model replacement. It isn't pretty.
Try comp.sci.electronics... or one of the military forums... A10 is good for frustration relief... weeds... energy salesmen...
Google "potting compound removal". First hit has tips for mechanical removal and links to chemicals used for removal.
MBQ
A short exposure to heat will set epoxy quickly.
Prolonged exposure to dry heat will not do anything to set epoxy.
Prolonged exposure to heat+moisture will destroy set epoxy (particularly the rapid-set type).
So simmering overnight would probably solve Dave's problem (but I wouldn't want to bet on the functionality of the PCB afterwards).
Man at B&Q coughed up some electrons that declared:
The firm I used to work for had to do a bit of depotting (on their own kit, to check for manufacturing quality etc).
There is a specific depotting chemical, but it does tend to turn the compound rubbery rather than wash it of. So the use of a sharp scalpel will still be required, though it is easier.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.