Range Finding at micron/sub-micron? Anyone know a method?

I have a side project relevant to machine building and I just need purely linear range finding, but down to something like the accuracy of a C0 screw at 10µm/m (including repeatability), with a range of at least 1m. Response time on the measurement could be as slow as 1s, though faster is of course always better. I can perform calibrations 'lights-out' over a solid week if need be, so settling times of 1s would still allow sub-micron steps if even measurable.

I'm looking around and it seems that at short distances, laser diodes with built in emission regulators can be used for this with averaging and get down to micron or sub-micron.

This isn't for varied part metrology, I'm looking to do calibration on a clean/new linear motion component and so I could provide a controlled environment for it. I've worked with products in the micron accuracy range before on the design and manufacturing side before as an ME, however I've not been responsible for developing or implementing measurement methods. I know at the workplace I did these designs for we used a CMM that had a 2µm accurate repeatability for $$$$, but that was a full 3D scanning probe system. I just need range finding so I can calibrate from start to finish on a controlled linear motion. I can even make the target, so I can ensure a minimum level of runout on the rangefinder if it can have a focused target, say a small mark or feature on the target with alignment set at 1000mm, then motion profile measured across the full range stop to stop.

In a real machine environment this would be a terrible idea, but in a clean environ with time for settling of vibration, I feel like there should be a (relatively) low cost solution.

Any help is GREATLY appreciated.

Not sure if I fully understand what you are after. If you are looking to look at variation of position over a 1 meter distance, maybe use a linear encoder spar and grab data from the output. A much better solution is use a laser interferometer. Used laser equipment can be had in the low thousands of $ from Laser Sam at; Sam's Laser FAQ: Welcome Page

greif1 said:

Not sure if I fully understand what you are after. If you are looking to look at variation of position over a 1 meter distance, maybe use a linear encoder spar and grab data from the output. A much better solution is use a laser interferometer. Used laser equipment can be had in the low thousands of $ from Laser Sam at; Sam's Laser FAQ: Welcome Page

Click to expand...

Awesome, that's extremely helpful.

Not really sure about this but you may need to consider how your requirements relate to the Rayleigh criterion or principle, something like that. 1 micron is getting near wavelength of visible light, resolution may become an issue.

Both interferometers and encoder scales are getting into the sub-nanometer range nowadays in terms of best theoretical resolution. A Heidenhain LIP21 scale reader head will pick up a 0.03-nanometer step! The scale itself controls the accuracy, that is, deviation from true position, and is listed at 0.125 µm over 5 mm and ±1 µm over a full meter scale. This all assumes lab-grade temperature and vibration control plus amazing guideways, axis drive, and control.

rklopp said:

Both interferometers and encoder scales are getting into the sub-nanometer range nowadays in terms of best theoretical resolution. A Heidenhain LIP21 scale reader head will pick up a 0.03-nanometer step! The scale itself controls the accuracy, that is, deviation from true position, and is listed at 0.125 µm over 5 mm and ±1 µm over a full meter scale. This all assumes lab-grade temperature and vibration control plus amazing guideways, axis drive, and control.

Click to expand...

Yes, absolutely. Hilariously where I worked before was making (in a different business segment) nano-meter scale linear encoders. I don't know why this didn't occur to me.

The issue with the encoders is $/ft vs a laser interferometer starts to get crazy since the glass or laser etched scales have to be ultra-precise as well. The advantage is that after installation they can be very rugged and don't need to be in quite as clean an environ as many of the laser interferometers do (due to the target needing to be clean).

That said, I am thought investigating this with the full understanding a controlled environ will be required for testing. Thankfully I don't need to provide a full clean-room, but all relevant portions of the testing equipment needs to be closed and temperature controlled, which means a lot of sealed liquid cooling throughout the borders of the machine and some decent insulation when everything is closed up. Plus I'm talking about 1m minimum lengths of travel for the calibration, with up to 3m being totally real for the types of things I'll be seeking to calibrate.


I did find, thanks to the first response, that I can get a laser interferometer that can measure a less than perfect target at up to 1800mm ranges with sub-micron accuracy in the $3k-$10k range, used and calibrated. It will be one of those catch as catch can kind of things as is the case with so much used equipment.

I'll price out a linear scale. These are produced frequently through laser etching and then involve a lot of nifty optics and interpolation to achieve the accuracy. I know the scales have to be almost perfectly flat at the time of production, plus installation flatness is critical. The laser interferometers really start looking like the more application rugged option (for me), which is hilarious since they are normally used for calibrating precision equipment, wafer locations, and part standards.

I can also rent a known interferometer, which is common for calibrating optics equipment. One of my co-workers is married to an optics engineer and her employers always rented, never owned, despite doing monthly calibrations on their optics devices. He's going to ask about what the costs were. They were renting a device that is typically $60k-$100k new. I estimate that a 1 day rental is probably on the order of $1k-3k. If I knew everything was going to be plug-in-play on arrival and the entire fixture was ready to go, I could generate >$10k in margin within a day of end product, assuming I had customers for the product. So the rental option seems like a really good approach until business became steady.