Question about micro-positioning systems

[ChaseRLewis]

So I've recently came into a few dirt cheap piezo actuators. Was considering turning them into a x-y-z positioning stage then maybe a micromill or something. My one issue is I'm unsure of what is best to use for feedback control in the sub-micrometer range.

Only thing I know that accurate is an interferometer. I also wouldn't 100% know how to guarantee that through mechanical means as most other means are temperature sensitive. Any help?

 

[Danger]

I know nothing about this, and yet insist upon having an opinion. Your mention of an interferometer tweaked something in my brain. The first thing that popped into my mind, although I have no idea of how it can be done, is to take constant holographic images and compare them to a reference sample taken before any movement occurs. Would that have the required resolution?

 

[ChaseRLewis]

I know nothing about this, and yet insist upon having an opinion. Your mention of an interferometer tweaked something in my brain. The first thing that popped into my mind, although I have no idea of how it can be done, is to take constant holographic images and compare them to a reference sample taken before any movement occurs. Would that have the required resolution?

That would be using interferometry : P I could but it's more i don't have a lot of experience with optics, I mean what type of optics would I need? Also image comparison is very computationally expensive and would need a camera on each axis for verification which would be fairly expensive. The accuracy of the motors at full step is below a micrometer but I've been reading about how 1 nm is pretty easy to get with closed-loop feedback ... but none mention what they use to get that closed loop feedback ...

Maybe some sort of peizo material that stretches? Not 100% sure on my options.

 

[Danger]

Okay then... and don't forget about that explanation that I don't know what I'm talking about...
Would you be able to get your mitts on something like an atomic force microscope?

 

[tygerdawg]

Sub-micron range is tough to do. Even at single-digit micron range, you must operate in a controlled environment to minimize temperature-induced errors.

Interferometry has merit I think. I've never done it, but my perception of the complexity just scares the bejabbers out of me. I'd first look at some of the various displacement sensor technologies out there. I have experience with laser displacement sensors (Keyence, others), and capacitive displacement sensors (Keyence). Perhaps modern implementations of some ultra-high precision LVDTs may offer something useful. None of this stuff is cheap.

But again, temperature is your Devil. A couple degree's change in temperature will skew your values at that magnitude. Same thing could be said for induced vibrations. I'd hazard an opinion that you'd have to mount this affair on a granite-composite base structure for vibration attenuation.

 

[ChaseRLewis]

Sub-micron range is tough to do. Even at single-digit micron range, you must operate in a controlled environment to minimize temperature-induced errors.

Interferometry has merit I think. I've never done it, but my perception of the complexity just scares the bejabbers out of me. I'd first look at some of the various displacement sensor technologies out there. I have experience with laser displacement sensors (Keyence, others), and capacitive displacement sensors (Keyence). Perhaps modern implementations of some ultra-high precision LVDTs may offer something useful. None of this stuff is cheap.

But again, temperature is your Devil. A couple degree's change in temperature will skew your values at that magnitude. Same thing could be said for induced vibrations. I'd hazard an opinion that you'd have to mount this affair on a granite-composite base structure for vibration attenuation.

Ya I'm planning on doing an air bearing type stand for it. That's what they use for high precision optic tables.

Ya, I'll do temperature control when operating it. No real way around that, and building a box for temperature control would be cheap / easy. Just some cheap peltier devices with temperature feed back.

Micrometer is probably good enough, just was thinking of "how do i get more!".