Questions on choosing a rotary stage 7

[brdf]

Hi...

Do you guys happen to know any good rotary stages that would work in the vertical plane? I searched Physics Instrument, Zaber and New Port etc. but mostly they only specify a maximum load weight so I assume they would only work well in the horizontal plane.

I am designing an automated "goniometer" to make optical scattering measurement. The key part is an aluminum BAR1 carrying a small optical tube assembly (telescope) pointing at ground to collect scattered light. Let X and Y be Cartesian coordinate in horizontal and Z along the vertical up (X cross Y = Z), then BAR1 will rotate around axis-Y in the X-Z plane and my telescope will always point to X=Y=Z=0. Let Theta be the angle the BAR1 makes with axis-Z, and my measurement would require Theta ranges from (-80, 80) deg., with angular resolution about 0.05 deg. Such a configuration would require a rotary stage working in the vertical plane (X-Z) with a load up to ~80Nm (I can add some counter-weight to reduce the torque caused by BAR1's weight and its telescope).

I would be very grateful for your suggestions/recommendations on the followings:
(1) a rotary stage working in vertical plane with angular resolution ~0.05 deg. and a torque load of ~80 Nm;
my BAR1 rotating speed doesn't need to be too fast (b/c I need to record scattering at each desired annular position), e.g., 0.1 deg/s; so worm gear is a good idea?
(2) as the whole optical system would be sitting on one whole bench, for the motor the quieter the better. Should I choose a DC servo or a stepper?
(3) to prevent BAR1 from hitting my incident optics (BAR2 fixed at another angle between (-80, 80) deg), what possible mechanism would be best?

Many thanks!

 

[Compositepro]

Have you considered using a constantly rotating scanning mirror and then correlating sensor data with mirror position?

 

[IRstuff]

There are lots of manufacturers of high load rotary stages; Aerotech is one:

http://www.aerotech.com/products/stages/armsspecs.html

Parker is another:

http://divapps.parker.com/divapps/emn/pdf/RseriesRotary.pdf

TTFN

FAQ731-376

 

[CorBlimeyLimey]

Try re-posting in forum386

Have you looked at

http://www.Thorlabs.com

or

http://www.Standa.lt

 

[lumenharold]

1) We've had very good success with the PI stage in a vertical position. The load was modest and close to the mount so the torque was low. I would give them a call to discuss the application.
2) We were collecting low energy scattered light and motor noise was not an issue. If noise is a problem can you pause the motor to make a measurement?
3) The interface was nice as far as the software (we used labview) so software stops could be used to limit travel. Not sure if that is failsafe enough for you but you could add proximity switches as well.

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[unclesyd]

This one Newark System was advertised on the page of this thread.

http://newmarksystems.com/?gclid=CLGE9f-6hqACFSOjagod9zbfCA

 

[IRstuff]

Just one more comment, counterweights do NOT reduce torque requirements. Added weight ALWAYS increases inertia, which increases torque for a given acceleration.

Counterweights are usually used to minimize linear perturbations that affect angular stability, but ensuring that the CG is as close to the rotation axis as possible.

As for type of motor, a servo is preferrable, since a stepper will inject mechanical impulses into the optical bench. However, if your setup is done with a plausible massive tabletop, the injected impulses should affect angular behavior, particularly if your angular resolution requirements are only 0.05 deg.

TTFN

FAQ731-376

 

[zekeman]

"(X cross Y = Z), then BAR1 will rotate around axis-Y in the X-Z plane and my telescope will always point to X=Y=Z=0. Let Theta be the angle the BAR1 makes with axis-Z, and my measurement would require Theta ranges from (-80, 80) deg., wit"

How can the telescope make a fixed angle to Z axis if it is in the x-z plane. Do you mean it makes an angle to the Y axis.?

I like a stepper for this since you have only modest speed requirements and you can even run it open loop
That means a 1.8 deg stepper would need a GR of 18:1 .and assuming.001" backlash in the final stage, a 5" dia gear would have an error of about .001/2.5*180/Pi= .024 deg.

 

[IRstuff]

Z, thanks for pointing that out.

I don't think what the OP requires can't be done with an off-the-shelf stage. In order to point to a fixed spot on the ground, the center of rotation needs to at ground level.

I think what he's actually asking for is something that looks like a semicircle, whose center is what he wants to point at, and the telescope travels along the semicircle, with the telescope pointed at the center.

TTFN

FAQ731-376

 

[zekeman]

"I think what he's actually asking for is something that looks like a semicircle, whose center is what he wants to point at, and the telescope travels along the semicircle, with the telescope pointed at the center."

IRstuff,
So, are you saying that the Z axis is the axis of rotation and the telescope is an element of the truncated cone of partial revolution?

 

[KENAT]

Not sure what they have in the rotary line but we've used these guys for some optics and stage mount stuff.

http://www.siskiyou.com/

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[IRstuff]

No, the z-axis would go through the TDC of the semicircle from the center of the semicircle on the ground. The telescope is supposed to point to that same spot on the ground while at different lookdown angles. A goniometer stage would produce different lookdown angles, but the displacement of the rotation center above the ground would cause it to look at different spots on the ground.

So, picture a semicircular arch and it's radii, pointing to the semicircle's center, located on the ground. The telescope must align to each of the arch's radii as it traverses along the arch. That's my interpretation of the OP. What that requires is a rotation of the telescope and a translation, so a goniometer won't do the job by itself.

TTFN

FAQ731-376

 

[brdf]

Thank you guys for all your kind replies, suggestions and everything! What a wonderful engineering forum I just stumbled upon!!!

zekeman: IRStuff's answer is correct; the telescope makes an angle Theta with the vertical Z-axis and Theta varies from [-80, 80] degs. To better illustrate my setup, the basic layout can be found in Fig. 1 (page 5221) in the following free online article:

http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-17-7-5217

(just click the "View Full Text: Acrobat PDF")

In the above primitive setup, I record scattering data at each desired viewing angle theta, and manually rotate the gonio bar to the next angular position. My goal now is to use a motor-rotary stage-controller system to control the rotating bar carrying the "viewing tube assembly" in Fig.1.

IRStuff: (1)once I have set up the system, my sample holder will be fixed at the center of the semi-circle and no translation will be needed (this will be a laboratory-based device not field one, sorry I did not point this out earlier).

(2) Also thanks for pointing out the torque thing; yes I did mean to make the CG closer to the rotation axis after adding the counter-weight so that the whole system would be more stable; and such an addition would indeed increase the inertia...I'm digging out my Mechanics notes right now....

(3 in your 2nd reply post, "..impulses should affect..." should be "...impulses should not affect....", right?

Cheers!

 

[unclesyd]

Here are two approaches to your problem. These are concept ideas not a particular recommendation. Both types of indexers can achieve very good accuracy. You need to contact any manufacturer of such machines as to what accuracy is available.

http://sankyoamerica.com/flexible-servo-index_high-accuracy-RA_features.htm

http://www.motionindexdrives.com/high-accuracy-dial-table.php

 

[IRstuff]

You misunderstood what's being translated. The "incident collimator" going from one theta to another requires a rotation AND translation in the system's coordinate system. The typical ellipsometer, from dusty memory, has the viewing assembly and the collimator on arms that are hinged on a y-axis going through the sample stage, and there was some sort of gearing that made the two angles equal. For what you want, you'd actually need two rotation motors with correlated encoders.

3> yes, should NOT, but it depends on the stepper impulse size and how criticalthe angular stability is.

TTFN

FAQ731-376

 

[lumenharold]

The article you referenced has the telescope viewing the sample from normal to 60 degrees. If the OP wants to view the sample from +80 to -80 degrees then won't the translation of the telescope become a limiting factor more so than the rotation? Perhaps translation stages could move the rotary stage vertically and horizontally for half the travel while the telescope rotates to view the sample, then reposition the sample 180 and repeat the test? Basically sweeping the optic in an arc. It would also reduce the required travel if the distance to the sample could be less than the 111cm in the report. Am I missing something?

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[brdf]

Thanks everyone...

IRStuff: Now I see....; in remote sensing often times people only need to measure scattering at a couple incident angles, say 0 deg (normal), 30 deg and 60 deg, in contrast to the denser viewing angles desired; so to reduce cost and complexity I would fix my collimator arm by using some clamping mechanism. But it's good to know your suggestion.

Harold: I guess my (-80, 80) notation was confusing...sorry about that; what I really meant was viewing the sample from normal to 80 deg, with minus angles clockwisely and plus ones counter-clockwisely (or the other way) . In reality there are "forbidden" angular areas where the telescope arm would mechanically interfere with the collimator arm. Also very likely I will end up with 65 deg or so, as the cosine response of my radiometer (telescope) would yield poor signals above 70 deg., no matter how hard I try...

Cheers!

 

[IRstuff]

Had to submit earlier so that I could get to work ;-)

The setup is similar to an ellipsometer, and in my dusty recollection, the viewing assembly(detector) and laser source for an ellipsometer were mounted on radial arms pivoting on an axis going through the top of the sample stage.

In your case, if the source and viewing angles are independent, you would need two rotation stages that could handle large cantilevered loads. Both would parallel and have their centers of rotation go through the sample position. If you imagine arms that goto each assembly in your Fig 1, and imagine rotation stages centered at the sample location, you get the idea. You would need to elevate the sample stage sufficiently off the tabletop to allow the rotation stages to extend below below the sample stage. It'll look like a two-wheeled cart with radial arms extending upward from each wheel when you're done.

Any of the suppliers cited above should work, depending on the actual weights you might have for the two assemblies, and the distance you need from the sample.

TTFN

FAQ731-376

 

[lumenharold]

Here's another question that has popped into my head: why does the sample need to be horizontal and the optics sweep around it in the verticle plane? Why not lay the whole thing down on a nice stable bench?

Second, if the measurements are being taken at fixed increaments, the telescopes might be cheaper than the rotary stages. Why not make a few?

Harold
SW2010 SP1.0 OPW2010 SP1.0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[IRstuff]

I think "fixed" increments meant something like every degree, as opposed to any fine, fractional degree positioning.

Given what he needs to do, it's not clear that horizontal would be an improvement, since that would make the sample sit vertically, blowing it the figurative breeze, and the other rotation stage would have to be mounted and cantilevered about the sample.

TTFN

FAQ731-376