Reversal of TouchRead() value increments

Status
Not open for further replies.
N

No1089

Member
Hi,

I recently built and installed a Teensy that senses a hand on an exhibit. During initial testing and setup the values returned by TouchRead() increased when a hand was present, but the exhibit stopped working suddenly.
Yesterday I went to have a look and found that the values are now decreasing! The hand is isolated from the face panel and the panel is earthed - to avoid static shocks. Nothing changed electrically so I am quite flummoxed as to the reason behind this reversal.

Could anyone give some insight into my observations and perhaps a way to avoid this?

Kind regards,
Chris
 
Thanks Manitou,

I've read through those now and can't find anything directly relevant. My design is earthed (directly at the plug) and the panel 5mm from the hand as well. The weird thing is that the values started going down after initially going up, with nothing changing electrically. And that makes no sense to me, why it would suddenly reverse.
 
Given how cap sensing works, and where I think Manitou was going is how is your audience grounded? Reliable cap sensing requires the hand or whatever doing the touching to be at a consistent impedance with respect to ground. Works great for hand held devices where you are touching the shell but not so much free standing where things like humidity, shoes or the condition of the floor will change how your human antenna behaves. Cleaning the housing can also change how things work as well.

Suggest checking if you get reliable results touching something earthed while manipulating the sensor. If you cannot arrange the design of the exhibit for reliable grounding of users you will need to have some form of running average process to determine what the steady state is and trigger on changes.
 
Thank you GremlinWrangler, from your comment I suspect the carpet cleaning and polishing of the touch surface probably influenced the readings. I hoped the earth on the surrounding panel would negate odd changes. The company wanted an explanation and the cleaning seems like the most likely cause. I will do some more tests and look at coding solutions.
 
Status
Not open for further replies.
Back
Top