Parasitic Capacitive Noise in QC room effecting capacitive touch sensor in product

[oceanjoe]

Hello -

I have a strange problem related to electrical / parasitic capacitance noise associated with using our product inside our factories normal QC room the product having a capacitive touch sensor. A power generator / frequency / converter is being used to rectify the power mains from 240VAC 50Hz to 125VAC 60Hz. The wire that connects the power mains machine to the product is directed above the normal height (~8ft) indoor roof.

When using our product generally (not in this particular set of several factory "QC" rooms) the capacitive touch sensor sensitivity is appropriate. Allowing one or more fingers in contact with the plastic handle to operate the device appropriately.

The problem is when using in this set of particular rooms, the sensitivity is heightened to the point where you can have your hand be 3 or 4 cm away from the handle and it gets activated. But when using rooms with higher ceilings and without the overhead cables from the power mains /frequency converter machine the activation is normal with one finger or more.

Is there any idea that could explain this phenomenon of parasitic capacitance environment cause the sensor circuits to be over-sensitive when the cable is over 8 feet away from the sensor? This is really the only thing that I can see different in the room setups about the cable being overhead and the ceiling height being normal 8 feet in height.

Any advice is appreciated about this strange phenomenon. I talked with the factory and they too are baffled.


Thank You,
oceanjoe

 

[itsmoked]

A human acts as an antenna that receives some of the signal found in rooms with power wiring running in the walls. When the finger gets near a touch pad it will have this signal quite strong on it. You can easily see this with an oscilloscope by turning the voltage setting down and touching the probe tip.

In a normal house this is always present but the magnitude is pretty low. Now in your QC rooms with the overhead wires you've cranked this whole thing way up to an over-the-top problem for your product.

You need to:
1) Shield the overhead power in conduit.
or
2) Shield the product with grounded metal with something like a sheetmetal roof above the testing bench.
or
3) Run the cables not overhead like free-space transmission antennas.

Whatever method you use, first test the space with a scope as described above. Screw with this until you get a solid feel for the existing environment. Hold the scope tip with your fingers right in the area you press buttons. See if the body 60Hz amplitude is uniform so you know what the magnetic field is like in the test bay.

Compare test bays. It may be you have one that doesn't cause this problem as much, note the difference with the scope. You will see a difference if you have different levels of the problem.

Ideally once you've measured the finger voltage you set up a test stand. Some sort of little tripod that holds the scope probe up vertically right in the area you'd be pushing buttons on the DUT (Device Under Test). Then you twist a piece of copper wire around the probe tip pointing straight up. Use maybe the old phone wire size wire. Have 30cm sticking up off the probe tip as a small antenna. Check for the 'received' voltage and then trim the wire down until you get the scope showing the same thing as your finger was showing. You now have a qualitative test for your remediation work. (1,2 or 3 above)

I learned this the hard way back in the day. While in college I made my girlfriend a cool desk lamp that had a transformer, latching relay, power supply, and an amplifier to detect the 'body voltage' of a person touching the large aluminum disk base the lamp stood on to toggle the light on/off. The guts were all visible in the acrylic body of the lamp. Worked great at my house anywhere I set it. This was before anyone had touch products or chips like you can get now. She loved it. She took it back to college and promptly set it on 'college furniture' in her apartment. A metal milk crate... Sitting on the excellent metal antenna it fried as it tried to turn on and off 60 times a second. So sad. Lesson learned. She still married me too.

Keith Cress
kcress -

http://www.flaminsystems.com