3 ms·

I had always assumed that they worked like optical mice (i.e. shine a light at a slightly-textured surface and watch the surface move underneath), but it's much

I had always assumed that they worked like optical mice (i.e. shine a light at a slightly-textured surface and watch the surface move underneath), but it's much simpler than that.

londons_explore 4y ago
Optical mice actually aren't good at precise relative positioning because the way the lens is setup, if they are moved slightly closer to the surface they 'see' more movement. You would imagine that you could use some kind of registration pattern(s) to solve this, but sadly nearly all mouse sensors are made by Avago, and Avago runs their own algorithm on the raw sensor data, and won't let you have the full data stream (you can only receive it via a debug API which only lets you get the data out super slowly). The avago ROM's are built into the sensor, and run a little 8080 CPU with a neat DSP for processing the image data, but sadly the system has too little RAM for any exploit to run your own code on it.
- bloopernova 4y ago
  Oh lord that would be an amazing hack in a story. A super secure facility gets compromised despite using custom firmwares on everything, but the mouse sensor slipped through the cracks.
Cthulhu_ 4y ago
I think you wouldn't get very good sub-millimeter precision with optical; I'm not very well versed in the subject, but infrared has a wavelength of between 780 nm and 1 mm, these things have to be more accurate than that.
- amluto 4y ago
  The little PCB pads are something like 500000nm across, so 780nm is nothing. You don’t need a sharp picture of the thing you’re looking at to locate it.
- pjc50 4y ago
  You can get very high precision with light using a similar technique: cause interference patterns. This lets you do sub-wavelength measurement. See also https://en.wikipedia.org/wiki/Very-long-baseline_interferometry https://en.wikipedia.org/wiki/Very-long-baseline_interferome..., where the wavelengths are typically larger than light.
- mattkrause 4y ago
  A millimeter is 1,000,000 nm, so even if you were wavelength-limited (and interferential techniques let you side-step that), there’s plenty of wiggle room. Heck, you could go down to visible blue light and free up another octave too.
mk_stjames 4y ago
Very large digital scales for digital readouts (DROs) on machines like manual lathes and mills actually do use quadrature optical encoding via etched glass scales inside aluminum housings, similar to the light gate on a scroll wheel of a mouse, but instead detecting fringe pattern light/dark from the etching in the glass scale. The etching is typically 5 microns in spacing, so a resolution of 0.005mm, or sometimes smaller. Because the glass is very stable this can be pretty repeatable over large distances.
- jbay808 4y ago
  Just to expand on that, usually the etched scale pitch is 20 microns, but digital quadrature detection gives you 4x the resolution of the scale pitch, resulting in a 5 um distance between successive rising / falling edges of the quadrature signal. Analog measurement can be used to interpolate to much finer resolution, even 1 nm for good quality optical scales with good electronics.

I had always assumed that they worked like optical mice (i.e. shine a light at a slightly-textured surface and watch the surface move underneath), but it's much simpler than that.

londons_explore 4y ago

Optical mice actually aren't good at precise relative positioning because the way the lens is setup, if they are moved slightly closer to the surface they 'see' more movement. You would imagine that you could use some kind of registration pattern(s) to solve this, but sadly nearly all mouse sensors are made by Avago, and Avago runs their own algorithm on the raw sensor data, and won't let you have the full data stream (you can only receive it via a debug API which only lets you get the data out super slowly). The avago ROM's are built into the sensor, and run a little 8080 CPU with a neat DSP for processing the image data, but sadly the system has too little RAM for any exploit to run your own code on it.

bloopernova 4y ago
Oh lord that would be an amazing hack in a story. A super secure facility gets compromised despite using custom firmwares on everything, but the mouse sensor slipped through the cracks.

Cthulhu_ 4y ago

I think you wouldn't get very good sub-millimeter precision with optical; I'm not very well versed in the subject, but infrared has a wavelength of between 780 nm and 1 mm, these things have to be more accurate than that.

amluto 4y ago
The little PCB pads are something like 500000nm across, so 780nm is nothing. You don’t need a sharp picture of the thing you’re looking at to locate it.
pjc50 4y ago
You can get very high precision with light using a similar technique: cause interference patterns. This lets you do sub-wavelength measurement. See also https://en.wikipedia.org/wiki/Very-long-baseline_interferometry https://en.wikipedia.org/wiki/Very-long-baseline_interferome..., where the wavelengths are typically larger than light.
mattkrause 4y ago
A millimeter is 1,000,000 nm, so even if you were wavelength-limited (and interferential techniques let you side-step that), there’s plenty of wiggle room. Heck, you could go down to visible blue light and free up another octave too.

mk_stjames 4y ago

Very large digital scales for digital readouts (DROs) on machines like manual lathes and mills actually do use quadrature optical encoding via etched glass scales inside aluminum housings, similar to the light gate on a scroll wheel of a mouse, but instead detecting fringe pattern light/dark from the etching in the glass scale. The etching is typically 5 microns in spacing, so a resolution of 0.005mm, or sometimes smaller. Because the glass is very stable this can be pretty repeatable over large distances.

jbay808 4y ago
Just to expand on that, usually the etched scale pitch is 20 microns, but digital quadrature detection gives you 4x the resolution of the scale pitch, resulting in a 5 um distance between successive rising / falling edges of the quadrature signal. Analog measurement can be used to interpolate to much finer resolution, even 1 nm for good quality optical scales with good electronics.