I'm investigating ways of making a capacitive sensor that will operate in very moist environments, potentially even submerged. What I need is a way to make some conductive plates in a specific arrangement that are electrically isolated from each other and from the environment.

One of the options I'm looking at is to use a PCB. I would put the plates in the proper arrangement on both sides of an FR4 board using polygon pours and then cover the whole thing with soldermask. There would be no components on the PCB at all. The entire board, except for the external contacts (outside of the moist/wet sample), will be soldermasked.


  1. Will the soldermask block the moisture effectively, preventing the plates from forming an electrical contact even in the case of prolonged submersion?
  2. Will the FR4 remain dry? Or will it soak up water, leading to an electrical contact between the plates?
  3. If this is not the case, is there anything I can treat the board with to make it waterproof?

Let's say for the sake of argument that I will be using seeedstudio's Fusion PCB service.


Yes FR-4 itself is waterproof ("near zero water absorption" according to Wikipedia).

  • \$\begingroup\$ Ah right, wikipedia. I should definitely have thought of that. \$\endgroup\$ – drxzcl Nov 4 '11 at 22:26

Seems hardly worth conjecture... Just have a couple prototypes made for the sake of experimentation. I've used many such services, but my currently favorite option is DorkbotPDX (good quality, very inexpensive, relatively fast turnaround). Then come back here and answer your own question for the benefit of all :).

  • \$\begingroup\$ Excellent plan. Thanks for pointing me to Dorkbot btw. Their pricing structure (per square inch) is much more suited to this project than seeed's (preference for square boards), as I plan to have a long and thin board. \$\endgroup\$ – drxzcl Nov 4 '11 at 22:32
  • \$\begingroup\$ I went ahead and designed a probe and got it into Dorkbot's nov 7 order. I'll post back when I receive the things and have had the opportunity to test it. \$\endgroup\$ – drxzcl Nov 7 '11 at 9:12
  • \$\begingroup\$ @DrX awesome! looking forward to reading about your experience! \$\endgroup\$ – vicatcu Nov 18 '11 at 16:24
  • 1
    \$\begingroup\$ Currently running the soak test! More info at zerocharactersleft.blogspot.com/2011/11/… I will update the question with my findings as soon as the beast has been soaked enough ;) \$\endgroup\$ – drxzcl Nov 23 '11 at 23:38

Water + circuit boards are a nasty combination. You can get any of the following situations:

  • trapped moisture underneath components (the space between SMT caps/resistors and the board effectively forms a very thin cavity that acts as a capillary)
  • moisture + surface contamination (leftover flux, dust, whatever) forms a high-impedance (megaohm range) parasitic conductive path
  • DC voltage + water = electrolysis that can cause migration of metal ions + corrosion

I would encapsulate it somehow -- not sure if conformal coating is waterproof (my guess not), but you could use plates that significantly protrude above the circuit board, and encapsulate it in waterproof resin (with just the plates protruding)

  • \$\begingroup\$ I'll add it to my question, but just to be clear, there would be no components on the PCB at all, and everything except for the external contacts (outside of the moist/wet sample) will be soldermasked. \$\endgroup\$ – drxzcl Nov 3 '11 at 22:54
  • \$\begingroup\$ Thanks for the pointer to conformal coating. I'm sure a layer of epoxy would do wonders. \$\endgroup\$ – drxzcl Nov 3 '11 at 22:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.