I recently came across this link. I wonder if we could learn something from it’s design. I know several people are asking for a larger motion range; maybe this can show us something.
Pete.
1 Like
That is a nice design. It’s effectively a flexure Stewart platform. I think the issue in emulating this is that travel comes from a combination of 2 key elements.
- Elastic limit of you flexures (setting your maximum angle)
- Length of the lever arms
Some engineering may be able to squeeze a tiny bit of extra angle out of 3D printed flexures. To get significantly more angle you can move to flexure hinges In the past I have used these cross flexure pivot hinges:
They are really nice, but they are very difficult to assemble.
The other option for increasing travel with the current design is to make it much bigger (theirs is huge!). If we double the size of the microscope we would double the travel. The problem is that travel is proportional to length, and the stiffness of the legs are proportional to the length cubed. So if we double the leg lengths we have a microscope that is 8x less stiff this will cause a lot of stability issues.
I think that if we needed to increase the range of motion from 12mmx12mm to about 16x16mm that we could do that with some careful engineering on a largely unmodified design. This would be 1/3rd longer legs at a penalty of 2.3 times less stiffness in the rigid elements. I think we could probably analyse where this lower stiffness is affecting things and beef up some of those elements.
I think if we want to move onto whole slide imaging ranges of motion, then flexures are no longer the way to go (except for the focus in z). I think that probably a roller bearing stage is going to be the correct way to get that sort of travel.
I was intrigued by the idea of using folded leaf springs in this crazy geometry. Here is a paper describing the design. If I can ever figure out what connects to what, I’ll print one to see how it works.
I did once build a single microscope leg with a printed cross-spring design like your image as part of my exploration of the magnetic encoder. I hoped it would give more out-of-plane stiffness, but it didn’t help the encoder performance. But it’s certainly possible to print them.
This summer I had a couple of high-school students modify a 3D printer (an Ender 3) to replace the print head with a webcam. I wanted to scan 6" silicon device wafers looking for defects. They succeeded quite well. You can’t find an XYZ stage for such a low price (<$200). I haven’t yet explored its performance in detail in terms of position accuracy, but it certainly does this job!
2 Likes