Is it possible to liquify epoxy?

[daveyk]

DC/DC converter: 12 volts to +400volts.  Epoxied in a small case (entire device is not very big, 40mmx50mmx15mm). Pealing the metal case away (except on top, it is one solid block of grey epoxy).

Is there anything that can dissolve the epoxy, but not destroy the circuit?  I would suspect the Germans epoxied an electrolytic cap in that epoxied box.  I have one the last ones I could find on order for $403 and that's if the vendor can actually find what they show as in-stock.

[TheMG]

Soaking in acetone (at least a day) will usually soften the epoxy, just enough to be able to pick away at it but there will be a lot of manual work involved. Unfortunately it will also soften and damage some parts like the shrink wrap on electrolytic capacitors, epoxy coated disc capacitors, etc.

There are some stronger, much nastier solvents that will do a better job, not sure if any of those will actually liquefy the epoxy though, and they are not commonly available. Very likely to also damage some other stuff of course.

[daveyk]

Unless I can make my own boost converter to go from 12volts to 400volt.  Now much current draw from it, low duty cycle use. The 400volts is used to charge a cap which in then quickly discharged with a FET.  So it would not take much of an inverter.  Main use to get rid of the epoxy might be to reverse engineer it.

[DavidAlfa]

Worth reading:

https://www.tedpella.com/technote_html/Reworking_Cured_Epoxy.pdf

DCM seems to be the most aggresive solvent for epoxy, can be sourced easily.

[connectTek]

AU $4.79  31%OFF | DC-DC high voltage boost module ZVS capacitor charging electromagnetic gun 45-390V 780V adjustable voltage stabilizer
https://a.aliexpress.com/_mrEbkgg

[connectTek]

I use these for my railgun.
Good and cheap

[fzabkar]

Unless I can make my own boost converter to go from 12volts to 400volt.  Now much current draw from it, low duty cycle use. The 400volts is used to charge a cap which in then quickly discharged with a FET.  So it would not take much of an inverter.  Main use to get rid of the epoxy might be to reverse engineer it.

Sounds like a camera flash or capacitor discharge ignition.

[jpanhalt]

DCM = dichloromethane aka methylene dichloride, methylene chloride and others.  It is a component in some paint removers for epoxy and "all finishes" but VOC restrictions have limited its availability.  You need to check the MSDS/SDS to be sure.  It is extremely volatile and is flammable.  Chloroform might be a substitute and is non-flammable.

Heat will soften epoxy.  You may need to get to 140° or 150°C, but some versions soften at much lower temperature.  I would use  heat first and try to peel it off.

[daveyk]

Thank you!  I ordered several of each type to try. 

Dave

[daveyk]

Thank you; will do.  The chemicals sound scary.  I do have PPE, but want to avoid them.  I will use the heat gun method outside.  Raining hear today, but let's see what the weekend brings.

[magic]

Surprisingly, some of the shittier epoxy resins give in (become brittle) to ordinary vinegar after a few days of soaking.
I recall some success destroying FR4 laminate with either formic acid or DCM (I tried both, don't remember if both worked or which was better).

Boiling hot nitric or sulphuric acid is the only way to be sure. But it will also strip all ICs and transistors naked.

[magic]

BTW, is it really that hard to cobble together some piss poor, low current, low efficiency boost to 400V?
Every photographic flash contains something of that sort inside.

I think this is the real problem here.

[TimFox]

DCM = dichloromethane aka methylene dichloride, methylene chloride and others.  It is a component in some paint removers for epoxy and "all finishes" but VOC restrictions have limited its availability.  You need to check the MSDS/SDS to be sure.  It is extremely volatile and is flammable.  Chloroform might be a substitute and is non-flammable.

Heat will soften epoxy.  You may need to get to 140° or 150°C, but some versions soften at much lower temperature.  I would use  heat first and try to peel it off.

I used dichloromethane once to remove adhesive from ceramic tiles that had been covered with carpet tiles (a brilliant idea from the previous owner in both bathrooms).
Following the warnings on the container, I opened the windows and used a floor fan to blow fumes away, but I still suffered horrible headaches after the operation.
It did remove the adhesive, however.

[daveyk]

"BTW, is it really that hard to cobble together some piss poor, low current, low efficiency boost to 400V?
Every photographic flash contains something of that sort inside."

Well, if the DC/DC converters that connectTek found for me on Alliexpress do not work, I would love to make my own boost supply.  I am not sure where to start to make a +400volt boost supply.  That could indeed be fun.  I am all open for suggestions and what size coil is needed and the suggested switching IC.

[KG7AMV]

I just went thru this same search without any luck because I had an epoxy coated (black blob type epoxy) on a board.

If it was my board I would have slid it under the fiber laser and seen what it would do but was not my board. 

I ended up scraping and chipping slowly for hours. So if you find something I am all ears.

[jpanhalt]

If you just want to destroy the epoxy, rather than dissolve or soften it, nitric acid will etch it.  There are probably other ways to chemically burn it off, but nitric acid used to be used for decapping.

One thing to try that is a bit less aggressive would be an organic sulfonic acid.  Alkyl sulfonic acids are used in some of the really effective silicone caulk removers.   They work by denaturing the polymer.

[eliocor]

[...] I am not sure where to start to make a +400volt boost supply.  That could indeed be fun.  I am all open for suggestions and what size coil is needed and the suggested switching IC.[...]

https://www.analog.com/media/en/technical-documentation/application-notes/AN118fb.pdf

[KG7AMV]

If you just want to destroy the epoxy, rather than dissolve or soften it, nitric acid will etch it.  There are probably other ways to chemically burn it off, but nitric acid used to be used for decapping.

One thing to try that is a bit less aggressive would be an organic sulfonic acid.  Alkyl sulfonic acids are used in some of the really effective silicone caulk removers.   They work by denaturing the polymer.

Don't want to destroy the substrate 

[JacobPilsen]

daveyk should use x-ray first.
And then use a milling machine.

[jpanhalt]

Don't want to destroy the substrate

I suspected that.  Hence my first suggestion of heat.  Otherwise, you are left with a situation similar to removing one coat of paint from another coat with the same chemistry.  There are only two ways to do that I know: 1) Elbow grease and very fine sandpaper; or 2) Careful laser ablation.  In fine finishing, that is sometimes done on purpose.  The first coat can be a primer of a different color than later filler coats until the desired flatness is attained.  Then a finish coat is applied.

[garrettm]

Look up "epoxy glass transition temperature." This is a phase transition point where the epoxy goes from hard to soft/rubbery. A pressure cooker (filled with water) can be used to heat the epoxy to the transition point so you can scrape it off with minimal damage to the board. I'm sure there are some videos of doing this on youtube so you can get a better idea on how this works.

[N2IXK]

Another recommendation for DCM.  Doesn't liquefy the epoxy, but softens/swells it into a crumbly consistency than can be carefully picked away.  A soak for a few hours or overnight is needed.

Evaporates VERY readily, so use it in a closed container large enough to hold whatever you are depotting.   Will absolutely destroy many plastics, and remove markings like the color stripes on resistors, so be careful.

[Ground_Loop]

I knocked this up in about 10 minutes. 5vdc, 1KHz square wave switching the FET and I got an unloaded 573vdc out.   With a neon bulb load I was able to get a bit over 400vdc with some tuning and bumping the supply to 12vdc.   The inductor is 23mH, cap is 4.7uF, FET is IRF820, diode 6A10.  Of course anything really usable would need some actual design effort.





[ Specified attachment is not available ]