Distance sensor range from 0 - 20mm 0.1mm accuracy, is it possible with infrared?

I am trying to detect the position of a screw head if its all the way in or there is some washer the height of the washer is 0.7mm.
Not sure if using an IR device is the best solution for this. Anyone have any thoughts on this?

If the application requires simply detection of the presence or absence of the washer, you may get by with an ultrasonic sensor operating in the high KHz to MHz range. It won't give 0.1 mm sensitivity, but you my get adequate sensitivity from 0.35 to 0.7 mm. Here is one example of a brand I have had some experience with: **broken link removed**

Edit: That link didn't go directly to where it was supposed to go. Try clicking on the model SM650A21600, which operates at 500 KHz and has a claimed sensitivity of 0.35 mm. A 1 MHz sensor would give increased resolution.

John

Thats an interesting problem.

I'm guessing (at present ) IR could be made to do it somehow
but maybe there could be better ways - some more info could help...

i.e. What sort of clearance for access all round - what is each material made of (including what is being screwed into)
Is the washer flat or sprung (much of a gap if not down tight?)
Does the test need manual application or repeated application such as a production line - if so do you
have access to torque measurement during insertion etc Is a visual (computer) inspection possible
(any environment issues? dust water etc)
Is the screw area totally flat (close to engineering flat ?)
In a production environment a host of methods become possible
In a manual application you'd be more limited but then simpler solutions may be possible

Do keep us informed on this one.

As with many questions of this kind, the details matter. E.g. object size, surface finish, speed. If the application is something like production quality control, a certain accuracy overhead will be required for long term stability and general acceptance, so 0.1mm accuracy sounds reasonable. Laser triangulation would be my first choice.

Thanks for all the input.
The application is QC. Its a production it can be manual operated if necessary. The screw is M3 rounded top screw. I am not sure of the based it can be plastic. The screen + washer is manually screwed so 0.7mm +/- perhaps 5%?

I like the idea of Laser triangulation, that was my first choise as well. Using a photo array but not sure of the parts to use. The mechanical mounting will be fixed and calibrated, meaning that once mounted a unit with the washer and without the washer will be plased on the platform to teach my board when its present and when its not. So the solution doesnt need to be exactly 0.1mm accurate even if its 0.5mm accurate it would be good enough. It would be nice to achieve higher accuracy.

Regarding ultrasound at 1MHz, I have seen it give 1mm accuracy but 0.1mm I have not so far seen it done, I could look into it, but with regards to cost the photo array might be a more suitable choice. Think I will build the circuit around a photoarray and see what happens, thanks for all the input.