5

I am considering placing a typical PCB in a high-pressure oil environment. Right now it is just a "crazy idea", but I am wondering what would prevent this from theoretically working. Notionally, consider a motor controller consisting of some passives, a microcontroller and servo driver. All COTS.

The oil in use is non-conductive. I believe the core of my worry is about pressure on components. I am looking at about 300psi. Specifically, I would imagine some components on the PCB would not be tolerant to pressure (electrolytic capacitors). Some would be (SMD resistors, IC packages, etc.)

To be clear, the motivation for this idea is not related to cooling. It is simply convenient for my end application to be embedded in oil. (I tagged it with "cooling" because the setup is similar.)

Lastly, I have thought about possible references I can pursue for this. I know that some high-performance systems use phase change cooling within pressure vessels. I couldn't find any references on this, but I'm an EE and not physics. Perhaps there is some information out there on design of these systems?

Alex Hirzel
  • 325
  • 2
  • 9
  • I think there is no simple answer. A lot depends on size. 300psi is about 21 kg/cm^2, some bigger IC packages may just break and damage die inside. – Kamil Dec 05 '14 at 17:24
  • Does it need to be at 300psi? I recall an experiment in oil-cooling a PC system that worked for a while and then impurities / contaminants seeped into the coil (from dirt, components on the board, etc.) and it started to play up, so your non-conductive oil may not stay so long-term. – John U Dec 05 '14 at 17:43
  • I did this sort of thing a few years ago. Had a PIC24xx, some LTC and ADI parts and a bunch of passives making up a data acquisition system which would operate at about 100-120m underwater. Mounted everything inside a steel cylinder and filled it with 'transformer oil'. I avoided anything 'squishy' like electrolytics and derated other passives wherever I could. Used ceramic resonators instead of quartz crystals - but that was more of a gut-feel decision. – brhans Dec 05 '14 at 17:45
  • I have seen this done. I know a researcher who put electronic tags on diving marine mammals. They put them in cylinders, fill the cylinder with oil or wax, cap off the cylinder, and then pressure test in a small barometric chamber. The ones that pass go to the field. The only problem you would likely have would be with components that have sealed air spaces. These might will be crushed. I agree with John U about the contaminants. But there are special liquids that are designed for this application. Fluorinert, I think, is one. You could also consider using epoxy. – user57037 Dec 05 '14 at 17:50
  • Epoxy, wax, or just a good potting compound would seem like a better suggestion than oil. – John U Dec 05 '14 at 18:00
  • I understand this is done for underwater UAVs, you might want to look into their experience. Gut feel is that potting in epoxy is probably similarly hard on parts except perhaps ones that can buckle (HC49 crystals for example). At least it's easy to test- and relatively safe - since the materials are non-compressible. You could easily test at 1000 PSIG with a small hydraulic pump. – Spehro Pefhany Dec 05 '14 at 18:14

1 Answers1

3

Your biggest problem will be with sealed air spaces inside components. I know that EPROMS were commonly regarded as likely to fail under high pressure. More up-to-date examples would be quartz crystals and oscillator modules. Power transistors in TO-3 cans would have an enclosed space, as would some kinds of small transistors in metal cans (or op-amps in metal cans, if you remember those).

John Honniball
  • 2,829
  • 12
  • 10