I have a design that uses the MPR121 capacitive sensor. It's a very simple board with 10 "slider electrodes". Standard 2 layer construction with FR4 .063" thick.
The board interfaces with a microcontroller on another board, which processes the touch info. The touch slider board is meant to sit under a piece of wood that is roughly .5" thick.
During development/testing I used a 1" thick text book as a surrogate for the wood (I didn't have the wood casing at the time). The slider works fine and I'm able to process touch through the book.
When I try to put the slider behind the wood panel, however, I'm not able to process touch info and the ADC values of the MPR121 are totally unaffected by touch. The wood has a curve to it and a somewhat rough surface such that there is no way for the slider to sit flush on the wood. I'm guessing that the issue here is that the air gaps (2 mm in some places) between the slider board and the wood are killing my measurements. In particular the "dielectric coupling" is shot.
How do people mechanically interface rigid capacitive touch boards to irregular surfaces? Is there a dielectric gel? Adhesive? Just buttress the board against a compliant material with desired dielectric constant? Ideally I'd like to be able to remove the board and tweak it as necessary so epoxy isn't a good short term solution, though it would be OK once everything is validated.
EDIT:
I found polyimide foams on mcmaster carr, which seem like they might work in principle though I suppose if it's a foam it will have air inside of it, which may or may not be problematic. It's adhesive backed which is nice so theoretically I could mount the adhessive to the PCB and shove the assembly against the structure. Anyone have experience with/comments on such a solution? http://www.mcmaster.com/#polyimide-heaters/=z9aq3i
EDIT 2:
This TI document refers to two different 3M transfer tapes — 467MP and 468MP. I ordered some on amazon and will give it a shot. Maybe I can use one or two layers to fill all the gaps. I'll report back when I've tried them out.
Best Answer
I ended up using IPG carpet tape during testing. I used that tape because it's pretty easy to remove and not as permanent as the 467/8MP stuff appears to be. Long story short the transfer tape solved my problem. The board is 6" long and over the length was able to conform +/- 1 mm in the z dimension with the help of the transfer tape. The transfer tape did most of the conforming, no doubt. I also added some mechanical features to help push the board against the surface, which helped the board/tape conform to the surface. Problem solved -- capacitative touch is easily measured through the wood. I'll probably switch to the 467MP for fun now that I know that transfer tape seems to solve the issue (at this point the more permanent adhesion of the 467 is desirable. I'll update if I notice any difference in performance).