Epoxy potting

[DialtonePickups]

John explained that some materials are conducting vs. insulating, and we need one that insulates because of the active circuit + testing while wet. Wouldn't that apply to any active circuit? So throw an active, running circuit in water and it's going to break, which (I'm learning) is pretty much what we're doing when we encapsulate with a water-based or otherwise conductive material and then turn it on. We dont have to test while wet, but I like to because I catch any problems before the epoxy cures and can save the case if needed that way. If we wait until the epoxy cures, we have no way to fix it. We have a huge pile of dead pickups, I'm trying not to add to it.

I'll have to check out the plumbers tape!

[big_teee]

It's an extra step, but what if you wax potted first?
Had some dried wax residue on your Active before Epoxy.
The wax should not be conductive.
I make some pickups where I wax pot twice.
Once on the bobbins, then again after assembly.
You might wax pot, and leave a little in the cover, let cool, and then Epoxy.
However it sounds like you already have things worked out.
Just a another thought.
T

[Joe Gwinn]

Thanks! The stuff we are using now seems to be working perfectly, but we'll take a look, free samples can never hurt.

You don't want pickups coming back under warranty, but it's hard to do a lifetime test - takes too long. So how to test to ensure reasonable lifetime?

The standard answer is an accelerated aging test. There are many kinds, and a whole learned literature [agrt], but it's actually simple to do. Make five or ten sample pickups, with indelible serial numbers, and proceed to test. The trick is to combine thermal shock with high-temperature operation, all with power applied if the pickup is active. For the low temperature, the kitchen freezer. For the high temperature, a toaster oven set to 160 F. These temperatures are easily achieved by leaving the pickups in a car either overnight in Minnesota during the winter or during the Day in Tucson during the summer. Move from hot to cold and back to hot once per day. For instance, into the freezer at 9 PM, into the toaster oven at 9 AM, and so on, for a month.

If none of the test articles break from this test, you are good. If there are failures, dissect them and figure out what failed. Update the design and/or process, and try again.


[agrt] For the record, the vast learned literature goes under the name "accelerated reliability growth testing". Lots of fancy math, mostly to figure out how fast the units under test are aging so one can tell how long to test to ensure that a required lifetime will likely be met.

[DialtonePickups]

You don't want pickups coming back under warranty, but it's hard to do a lifetime test - takes too long. So how to test to ensure reasonable lifetime?

The standard answer is an accelerated aging test. There are many kinds, and a whole learned literature [agrt], but it's actually simple to do. Make five or ten sample pickups, with indelible serial numbers, and proceed to test. The trick is to combine thermal shock with high-temperature operation, all with power applied if the pickup is active. For the low temperature, the kitchen freezer. For the high temperature, a toaster oven set to 160 F. These temperatures are easily achieved by leaving the pickups in a car either overnight in Minnesota during the winter or during the Day in Tucson during the summer. Move from hot to cold and back to hot once per day. For instance, into the freezer at 9 PM, into the toaster oven at 9 AM, and so on, for a month.

If none of the test articles break from this test, you are good. If there are failures, dissect them and figure out what failed. Update the design and/or process, and try again.


[agrt] For the record, the vast learned literature goes under the name "accelerated reliability growth testing". Lots of fancy math, mostly to figure out how fast the units under test are aging so one can tell how long to test to ensure that a required lifetime will likely be met.

Thanks for the tips!! We did a little bit of that with this testing we did: https://www.facebook.com/media/set/?...1124589&type=3

But I think we need to do more. I really appreciate your suggestions!

[Joe Gwinn]

We did a little bit of that with this testing we did: https://www.facebook.com/media/set/?...1124589&type=3

The drop tests are not that hard for a potted assembly to survive. In WW2, they shot potted electronics with glass vacuum tubes (the VT fuze) out of howitzers at 50,000 G acceleration.

As for the hot and cold tests, good start, but not enough hot-cold cycles, and probably the soak times were too short. Thermal mismatch is usually the biggest issue, and will cause the assembly to tear itself apart.

Hot Humidity can also be an issue. The way to test this is during the toaster sojourn, have the pickup in a plastic bag with a wad of wet paper towel.

But I think we need to do more.

Probably the easiest way to make the test more severe is to make the low temperature lower and the high temperature hotter. It's hard to make the low temp much lower than 0 F with ordinary home appliances, but if you can go hotter than 160 F (check component ratings), the pickup will age faster.

The classic way to test semiconductors in metal cans was ten cycles of alternately plunging the test article in boiling nitrogen at -196F, then silicon oil at 700 F, then nitrogen, and so on. This is every bit as severe as it sounds. The amazing thing is that it is possible to build packages that will survive this abuse.

The other test is to keep some test articles at high humidity for a year. This involves a few mason jars with a water-salt slush at the bottom and some kind of shelf to keep the pickups above and out of the salt slush. Put the mason jars somewhere warm. The salt slush will keep the relative humidity at 75%. Other kinds of salt will maintain higher or lower. Search for salt test of a hygrometer.

[DialtonePickups]

The drop tests are not that hard for a potted assembly to survive. In WW2, they shot potted electronics with glass vacuum tubes (the VT fuze) out of howitzers at 50,000 G acceleration.

As for the hot and cold tests, good start, but not enough hot-cold cycles, and probably the soak times were too short. Thermal mismatch is usually the biggest issue, and will cause the assembly to tear itself apart.

Hot Humidity can also be an issue. The way to test this is during the toaster sojourn, have the pickup in a plastic bag with a wad of wet paper towel.




Probably the easiest way to make the test more severe is to make the low temperature lower and the high temperature hotter. It's hard to make the low temp much lower than 0 F with ordinary home appliances, but if you can go hotter than 160 F (check component ratings), the pickup will age faster.

The classic way to test semiconductors in metal cans was ten cycles of alternately plunging the test article in boiling nitrogen at -196F, then silicon oil at 700 F, then nitrogen, and so on. This is every bit as severe as it sounds. The amazing thing is that it is possible to build packages that will survive this abuse.

The other test is to keep some test articles at high humidity for a year. This involves a few mason jars with a water-salt slush at the bottom and some kind of shelf to keep the pickups above and out of the salt slush. Put the mason jars somewhere warm. The salt slush will keep the relative humidity at 75%. Other kinds of salt will maintain higher or lower. Search for salt test of a hygrometer.

This is really helpful, thank you Joe!! I'm planning on starting another round of testing soon. You have some really good suggestions here.

[David King]

I would make the internals modular. Separately encapsulated coils and electronics so if one goes bad you can replace it but keep the case which must represent a considerable investment on it's own. We should all be thinking in terms of cradle to cradle on all our products. Much more likely to recycle the copper and aluminum if you can pop the thing apart with just a couple of screws. Future generations will undoubtedly curse us for our profligate abuse of their resources.

[DialtonePickups]

If we could find a way to separate the case + internals, we would love to. Especially in our earlier days (not that we're that old yet), we broke a lot of pickups, and they are useless once they are encapsulated. But, I don't think we can. Our parts are machined by a small shop in Oregon and no two cases are exactly the same. It took us a long time to find a shop that was a good fit - their parts might be a little off sometimes, but they respond to phone calls and don't require us to order thousand of parts at a time. So, back to epoxy - if we encapsulate them separately, we'd risk the internals not fitting each individual case. I bet there is some way to do it, but not that we have figured out yet - at least not easily. We are so maxed out on time/money that the thought of taking another thing on right now seems intimidating. But I agree that we, along with any other individual/business who reasonably can, needs to think about sustainability. When we can afford it, we will revisit. In the meantime, I'll keep reusing plastic bags & paper towels during assembly, trying to buy only recycled material for our packaging, etc. I know it's not much but it's a start!

[Jason Rodgers]

Speaking of wax, could you coat the inside of your case, like a releasing agent on a mold? If it tested ok, everything would be still be sufficiently impregnated and immobilized with epoxy; but if something went wrong, the guts could be popped out and tossed, and the cover could be reused. You have the eponymous dials, of course, and I don't know how those fit into the scheme.

[Joe Gwinn]

The standard answer these days is to deconstruct the failed units, and figure out what's happening, and fix the process so that doesn't happen again. Now of course when a monoblock potted assembly fails, it may not be so obvious what the root cause was.But if you run an unpotted pickup through the thermal cycles and nothing fails, but when the same electronics are potted and the thermal cycles cause failure, the first place to look is epoxy shrinkage and mismatch of linear coefficient of thermal expansion. This part comes from industrial experience. One way to tell is to conformal coat the the electronics uin urethane rubber before potting. If this prevents the failures, the problem was mismatched coefficients of epoxy shrinkage. It may be that the permanent solution is the urethane layer.

Anyway, maintain faith - many people epoxy pot electronics, and it works. But it may require some thrashing about to figure out what to do and not do.