Insure parallelism of part with lapping jig during lapping 1

[Eeikon]

Hey everyone,

I am working as an electrical engineer and I need some mechanical advise on lapping.

The issue I am facing is that i need to lap a crystal bonded to a rod within a few micrometer of accuracy and I need to align the part (in the center) together with the frame of the jig in order to avoid lapping the rod in the process.
Do insure the parallelism between rod and jig I used to place a plate on top of the jig with a hole on it to use as the jig referential and then measure with a flat gage that the 4 sides of the rod are at the same height. Finally align the jig depending on what the flat gage says until I get it right.

But the accuracy is quite low for the needs (~+-10um) and it takes lot of time to get it right. I added a few pictures of the setup to help understand it and the issues. DO you guys have some ideas on an alternative way to do this? I was considering using something like an auto colimator maybe but there is still the issue of the referential to get as precise as possible.

 

[3DDave]

What comes to mind is using an optical flat and a monochromatic light source. You are looking at around 10,000nm; I expect a lamp in the green to red range, 500-700 nm will be sufficient; that would be 14-20 or so lines across the sample.

There are a number of large flats on e-bay going for cheap due to edge chips, which shouldn't affect your application because you are checking the center. You'll probably want to find a place for calibration to see what deviations the flat does have; some of the listings indicated previous calibrations.

Lamps are more expensive. probably due to keeping the lamp at a constant, known temperature. In this case that isn't so important as long as the temp is steady. Since you are adjusting for parallelism, you don't need to do more than keep tweaking until the line count is low. If the same setup is 10 one day and 12 the next, that should be close enough.

 

[Doug Hunter]

Eeikon,

I recommend talking to a measurement specialist such as Keyence or Ametek. If anyone has a solution, it will be them, but it will not be cheap.

3DDave,

Are you talking about some sort of rigged-up interferometry?


Best regards,
Doug Hunter
Altarium Technical Consulting

https://www.altarium.us/

 

[Eeikon]

@3DDave, thanks! I actually did see some and they do even some optical flat with ring shape (can exclude the crystal from the interferometry measurement but I didn't have much experience with it.

@Doug Hunter I went to check with some companies and the price range for such device is around 20k USD. It's not impossible but not the best also.

I found the LG2 autocollimator from Logitech but I am not sure that it will bring enough precision.


I am even considering some alternative way to directly align both jig and part in one go using gravity and a holder placed under but it seems a bit far fetch.

 

[3DDave]

Hardly "rigged up." It's how many optical surfaces are inspected. One of the companies that comes up is literally named Lap Master, a maker of lapping machines. It may be necessary to add adapters similar in precision to gauge blocks to provide a very flat surface and avoid the crystal or make some other alignment offset compensations.

 

[Doug Hunter]

I also thought about interferometry (my father spent his career innovating in the field at Zygo) but didn't see how the whole set-up would fit inside a standard machine.


Best regards,
Doug Hunter
Altarium Technical Consulting

https://www.altarium.us/

 

[3DDave]

Here's an example:

https://youtu.be/xSYl7q6yKPU?t=407

Using interferometry is different that using an interferometer.

 

[Doug Hunter]

Good stuff, 3DDave. Thanks.

Best regards,
Doug Hunter
Altarium Technical Consulting

https://www.altarium.us/

 

[CWB1]

Is there a reason you're lapping and not grinding?