I'm waiting for some parts in the mail for my next post, and as you guessed, they haven't turned up yet. I only had a day or two to come up with an article, so I was happy when my father came to me with his digital callipers that weren't working. He'd left a battery in them too long and it had leaked, causing so much corrosion that one of the terminals had broken off. He wanted to know if I could repair them. While fixing them I thought I'd take the chance to document the tear down and find out how they work.
As usual, getting to the insides required me to remove a sticker on the back to get at the four screws holding the case together. What I found inside had me perplexed, there was no mechanical connection between the circuitry and the slide mechanism, I had expected something like a rack and pinion gear and some sort of rotary encoder. The PCB had a series of tracks on it that indicated some sort of capacitive sensing but I still wasn't sure.
As I mentioned before, there are a series of parallel tracks on the PCB that indicate a capacitive sensing system. From the vias it appears that every forth one is connected, when you see the other side of the PCB it will become clear that every eighth track is connected. It was still unclear how everything worked though.
|Calliper PCB back|
After a bit of Googling I found a description of how this type of digital calliper works. It's from the Anyi Instrument company in China and I'm going to place a screen capture of it below in case it disappears from their site in the future.
It's simple in concept but complicated in the details. Underneath the ruler markings is another concealed circuit board. It's comprised of a series of conductive fingers that interact with the traces on the PCB to form a variable capacitor that changes as the two surface move over each other. How that's converted to a measurement of length is more complicated.
|Explanation of capacitive sensing system|
Someone has done the hard work of probing the exciters on the main PCB for me. This site about DROs (Digital Read-Outs) by Nick Müller is an excellent resource. Below is an image of what was captured. For all money it looks like a 8 phase PWM signal. From what I can tell, by measuring how well each of these signals is capacitively coupled to the grounded frame, the ic can then use signal processing to sense its position. It is not an absolute sensor, it's relative, this means it needs to be zeroed before each use, but can detect small amounts of motion once calibrated. For those of you more enthusiastic, I've found a patent from the late 80's that describes the system.
Anyway onto fixing the thing. Removing a few more screws freed the PCB from its enclosure.
|Calliper PCB front|
Removing the corrosion from the battery contact was easy with small file. The board was then cleaned with some flux and a cotton bud and the contact re-tinned.
|Corroded battery terminal|
The top of the board is relatively bare. There's a chip on board, a crystal, a battery holder, the exciter traces that connect to the tracks on the other side of the board, there is also the pad where the LCD zebra connector contacts the board.
|Calliper PCB top|
Conveniently there's also a port on the side that allows other devices to read the measurement data. I didn't probe it, but it's well documented and rather easy to read. The writer of Robocombo has done a great job of documenting the protocol. I'll paraphase their work. There's a ground, clock, data, and a 1.5 Volt line. 8 times a second the data is transmitted. The measurement is transmitted as an integer that is 100 times the measurement in a 24 bit block with data being read on high to low transitions of the clock signal. The first bit is a start bit, the 21st bit is a sign bit, and the last three bits appear to be unused. There is also a description of the protocol on the blog yuriystoys that indicates that the 24th bit is a flag for mm/inch mode. In the inch mode it appears to send the data as how many 2 thousandths of an inch are in the reading.
|Calliper data port|
After I reassembled the callipers I found the part of the battery holder that had snapped off. It wasn't too hard to solder it back in place. It was higher than the original design and wasn't as springy. This was overcome by using a slightly thinner 1.5 Volt battery of the same diameter and a bit of folded cardboard above it to keep the battery in firm contact with the terminal below.
|Calliper battery compartment|
There you have it, a functioning set of callipers again.