-1
\$\begingroup\$

I've run across an interesting problem with a device I'm building. It seems that when I put my PCB into its plastic enclosure, I can create some signal simply by tapping the case. What makes this predicament even more interesting is the fact that when I put a silicone protective casing atop the plastic, the effect manifests itself differently.

This is a picture of what my signal looks like when I tap on the casing.

Signal produced when tapping on case. That initial negative peak and its shape is common amongst my different attempts at tapping.

Could this be a mechanical issue perhaps?

In addition, here is a schematic of my sensor circuit pre-amp if it might provide additional clues. The use of such high resistors for my biasing circuit is necessary because my sensor produces very small signals. Of course the downside to this is that I am less immune to noise... could this be what is causing my "tapping" issue?

enter image description here

\$\endgroup\$
  • 6
    \$\begingroup\$ That's ... not a pre-amp ... \$\endgroup\$ – brhans Feb 27 '18 at 17:46
  • \$\begingroup\$ @brhans my apologies if my terminology is incorrect, I simply meant this is what the circuit looks like leading into my first amp. \$\endgroup\$ – JimJammer Feb 27 '18 at 17:52
  • \$\begingroup\$ @Commanderson it's a fairly novel sensor, the basic principle behind it being that when acted upon, it produces a static charge which we are trying to measure. \$\endgroup\$ – JimJammer Feb 27 '18 at 17:53
  • 5
    \$\begingroup\$ Can you post schematics? sensor datasheet? do you have microphonic parts like high-K ceramic caps filtering a high-impedance node? \$\endgroup\$ – peufeu Feb 27 '18 at 17:54
  • \$\begingroup\$ Unfortunately because it is a novel sensor, there is no data sheet. Consequently, I am also unable to reveal TOO many details about its mechanism, which I realize isn't conducive to solving my problem. All I can say is that it has to do with the triboelectric effect, and producing a charge when it is acted upon. @peufeu I have not tried any microphonic parts like that \$\endgroup\$ – JimJammer Feb 27 '18 at 17:59
2
\$\begingroup\$

You can get the same effect by tapping on a DSO BNC input due to piezo-electric ceramics.. even expensive DSO’s!

Replace all ceramic caps used for signal bias with plastic caps. Then coat inside plastic with conductive or static dissipating material tied to 0V. <<1M

We used to generate 200V on ESD charge detectors just by raising a foot off the carpet and retesting. V=Q/C. and changing body capacitance to carpet with a finger tester.

1M is not that high. Try 1G

\$\endgroup\$
  • \$\begingroup\$ Interesting idea to switch from ceramic to plastic cap. Currently there is only one cap that affects the biasing circuit, and it is located right at the 3.3V regulator. Could this really have an effect? \$\endgroup\$ – JimJammer Feb 27 '18 at 18:38
  • \$\begingroup\$ Also, could you further elaborate on the purpose for coating the inside of the plastic, and then tying it to 0V please. \$\endgroup\$ – JimJammer Feb 27 '18 at 18:38
  • \$\begingroup\$ Conductive spray is common for charge shield and EMI. You can test with alum foil but first test it better with finger tap if noise irise time is only fast with tap input it’s microphonic. Silicone won’t help. Epoxy only a bit. Globs of PU adhesive maybe. Use plastic caps \$\endgroup\$ – Sunnyskyguy EE75 Feb 27 '18 at 18:50
  • \$\begingroup\$ Is there a reason why I see this effect more when I have my casing on as opposed to when I don't have my casing, and the effect seems to be minimal or nearly non-existent? \$\endgroup\$ – JimJammer Feb 27 '18 at 19:19
  • \$\begingroup\$ Dissimilar nonconductors in contact get a small voltage (triboelectric effect), and moving them moves the small associated charge... and makes big signals. This is why rubbing a balloon in your hair makes static. \$\endgroup\$ – Whit3rd Feb 27 '18 at 19:59
0
\$\begingroup\$

Rather than fingers, try tapping against the plastic with a piece of identical plastic. If the signal isn't there, then it was voltage from triboelectric surface contact with human skin.

PLASTIC ENCLOSURE. Lack of proper metal enclosure for shielding.

If triboelectric voltages from finger-contact against plastic surfaces is causing unwanted signals, then you're probably also seeing: human shoe-scuffing during winter, motor brush noise (try a Dremel tool), thunderstorms, 27MHz security guard handhelds, cellphones, microwave ovens, Ham radio and AM radio stations, etc., etc., etc.

Suggestion: DON'T FREAKING USE UNSHIELDED FRONT ENDS!!!!!

Or, if you do, then don't complain that your front end is acting like an antenna, and all sorts of weird EM phenomena show up.

\$\endgroup\$

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.