How cheap digital verniers work

submited by
Style Pass
2023-02-07 13:00:39

I'm not the first person to look at this - Grant Trebbin has some photos of a teardown, and has done a good job of finding patents like US5068653 Capacitive displacement measuring device with t-shaped scale coatings; Nick Müller has inspected one with an oscilliscope and Anyi Instrument Co have posted some info on how the sensors work.

The green PCB on the left is for the slider. Every eighth bar in the densely striped section is connected together - i.e. the first, 9th and 17th bars are connected together, as are the 2nd, 10th and 18th, and so on.

I measured the distance between the PCBs on the slider and shaft using a cheap feeler gauge; the spacing was about 0.35mm ±0.05mm.

Based on the 600 dpi scan above, the T sections on the shaft and the densely striped section is about 250 mm², and each of the eight banks of stripes has an area of 31 mm². The return strip has an area of about 95 mm². The comb overlaps the ground plane by about 43mm²; while that's obviously the same order of magnitude as the return strip, they don't seem to have been made precisely the same surface area.

So each of the eight stripes has a capaitance of about 0.78 picofarads while the return strip has a capacitance of about 2.4 picofarads - taken in series, that's about equivalent to 0.588 picofarads. That's very small. For example, the ADC on an ATMega (like an Arduino Uno) has an input capacitance of ~14pF. It's also smaller than the input capacitance of my oscilliscope. But some capacitive sensors claim to be able to measure femtofarad capacitance changes - the AD7745 claims an accuracy of 4 femtofarads, and the AD7147 a sub-femtofarad output noise.

Leave a Comment