Electronic Vernier Caliper Teardown
ฝัง
- เผยแพร่เมื่อ 4 พ.ค. 2024
- Teardown of a caliper to the silicon level. Super clever engineering to achieve amazing precision.
Photos of the die at: electronupdate.blogspot.com/2...
How to use a caliper with Ardunio: sites.google.com/site/marthal... - วิทยาศาสตร์และเทคโนโลยี
Awesome topic! These things do amaze me. Even a $5 caliper you whip in and out as fast as you can and it somehow doesn't lose any steps
Agree, it's amazing that I can use a cheep caliper and it just works, and has all this technology in it.
Because all the sensing is going on in hardware, not software!
Because it reads the position off track unlike the older ones that counted bars on the track and had a speed limit.
The patent that explains everything. US5731707. Crude PWM to make psuedo "sinewave" equivalent (FIG2a-g) that moves along the length but constantly flip's a demodulator to avoid negative voltages that goes into an integrator circuit (Fig 14) to accumulate charge every time the next psuedo sinewave. This all results in a sinewave that appears to be moving across the 8 emitters and it's searching for where the capacitive plate that is coupled has no charge because it's balanced by the finges producing a sinewave over the 8 terminals that results in an even voltage. This is cap coupled to the moving bit, back to the large plate. The whole system is basically looking for the "zero" cross. but even with the pwm like psuedo sine wave it's not too acurate. It's original design used a 6bit ADC to get about 10 micron accuracy I believe or maybe 1micron?. Commonly misunderstood to be capsense and it's not, certainly if it did the small relative in-out motion of the slide would throw off readings. The actual capacitance value means precisely nothing, just needs sufficient to couple through. The fun thing is that a 20 dollar Amazon special and a 200 dollar Mitutoyo only differ in physical material quality and low power when screen is off. There are other sensing arrangements Nils Andermo invented that may be more accurate and yet nobody uses them.
Thanks for providing more details
Dropping acid is also my favourite part whenever i take apart old electronics to study them.
😂😂😂
Nice video.
I like the way you immediately get into the details and don't waste time.
So, even digital works on the vernier principle. Great content.
Interesting teardown! I was expecting a little encoder wheel like on a computer mouse, but more precise. The linear design makes more sense.
At first I was sceptical of plastic, but for 5$ you can get pretty precise device. Certainly better than a mechanical ones where you cannot see a damn thing. And very convenient to use.
this is a stunning video! i really enjoy the level of detail and method you go into
Awesome video mate, I've learned a lot thank you! Subscribed :)
Fantastic teardown. I have one of these cheap units and wondered how they achieved such accuracy.
Cool 👍
Seems to me like the capacitive equivalent of the magnetic resolver.
I bet the stuff in the upper right is memory (ROM+RAM) for a small CPU, if even a 4 bit one.
Probably. A lot of chinese manufacturers implement a small cpu core for flexibility and being able to program it. Sometimes they even lay the cpu core out as raw gates
The simplicity of design, incredibly low power of operation, extreme precision (I am able to repeatably get near single micron accuracy from them), and robustness of the measuring mechanism to mechanical shock and wear of these devices is astounding. I am amazed every time I have to use one at work.
I take it you're using the serial interface for that. These calipers only display to 0.01mm.
@toolscientist I have a slightly different type for optics that I think works on the same mechanism
@@Muonium1 a micrometer? Or do you have 0.001mm calipers? Never heard of 0.001mm calipers. I'd be interested to know the brand.
@@toolscientist There are different brands with those calipers. For example "Metav M216.166" or "Sylvac 19-021 / 10-860".
You get them for 100-400€.
0,1mm, the display doesnt even have 2 digits after comma @@toolscientist
another electronupdate video!!!!!!!!
Great video. So the question would be, could you tell absolute position on an XY rail by just creating a PCB that lays on the rail with the wider fingers and then let the board with the smaller fingers just ride along the carriage on the rail? This would be a great way to create a closed loop system, but it doesn't seem like folks do that too often. If you look at linear motors for instance, they tend to use plastic strips along the rail with tiny black lines, and then just pick up light passing through the plastic or getting blocked by the black ink. This leads to a further question that these calipers do seem to know their absolute position, not just relative, albeit you do have to zero them. So perhaps they are just relative and the battery makes them feel like they're absolute because they just always watch for the movement which is why the battery dies on these after about a year.
They almost certainly only have relative position measurement, and remember the absolute position. The trick to a relatively low power use is the low operating frequency. The probing signal is ~8 kHz and if there's a higher master frequency it's probably still just in the tens or hundreds of kHz.
A CNC machine might be electrically noisy. The capacitive signal might be so weak that it could be disturbed by the magnetic field from motors. Maybe it will even have the occasional missed pulse without interference. In a motion system those errors will aggregate, unlike in a pair of calipers that will be zeroed every now and then. Light probably has a higher signal to noise ratio. Maybe it could still be useful for relative measurements, like continuous PID loop tuning, or combined with a coarser absolute positioning system.