Automating and characterizing an Olympus CHT focus block

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Chris, thank you for the additional info about belt tension.

So, I've just now gone back and rechecked my setup.

Tensioned as described, the belt bells by approximately 1 mm on each side, as measured by displacement from pressing a straightedge against the belt and pulleys.

I tried loosening the belt to the point where it just barely would not slip. That introduced a backlash of approximately 1/8 turn of the stepper motor, such that rocking the motor back and forth within that range did not move the focus knob at all. I know how to deal with backlash, of course, but I had hoped to avoid that phenomenon altogether in going with a focus block setup.

Going back to the original tensioning, I see no backlash. In addition I notice that when the motor is run continuously, there is no variation in the sound as the motor turns.

Going to more than the original tensioning, however, I notice that when the motor runs continuously, there is a systematic variation in the sound depending on where the motor is in its rotation. I speculate that this is due to that slightly off-center pulley that I mentioned, in this case running up against the non-stretchy belt so as to significantly increase & decrease the belt tension on each rotation. That would be a bad thing, for several reasons as you describe.

I gather from all this that there's some range of acceptable tensions, depending on one's tolerance for backlash. That's an interesting aspect that would not have occurred to me before this experience. Thanks for the discussion.

--Rik

Chris S.
Site Admin
Posts: 4042
Joined: Sun Apr 05, 2009 9:55 pm
Location: Ohio, USA

Post by Chris S. »

rjlittlefield wrote:I tried loosening the belt to the point where it just barely would not slip. That introduced a backlash of approximately 1/8 turn of the stepper motor, such that rocking the motor back and forth within that range did not move the focus knob at all. I know how to deal with backlash, of course, but I had hoped to avoid that phenomenon altogether in going with a focus block setup.
Yikes! That's a lot of backlash. I just checked the backlash setting I've been using on my StackShot controller, and it's 128 steps (which as we know, are really microsteps, or 1/16 of an an actual motor step). This translates to 8 motor steps, and since my motor is a 400 step/revolution motor, that's two percent of a turn, amounting to 2 microns of linear motion. To me, this seems like a pretty acceptable amount of backlash. I've never found it to be a problem.

If you compare your timing pulleys with mine (repeating the link for convenience here, fourth image down), it's notable that your timing pulleys are small, compared with mine, and that your belt span is long, compared with mine. I wonder if these factors, in addition to your gearing down on the motor side, contribute to backlash? I'm tempted to suspect that timing pulleys ought to be reasonably large in order to reduce backlash.

It may be worth highlighting that while you designed your approach from scratch, the centerpiece of my build was something I purchased and adapted. I did not change the timing pulleys or timing belt. The vendor originally had several identical motorized blocks (five, if memory serves), and they likely came out of an industrial facility in Asia. My bet is that they were motorized by a custom integrator, and it's fairly likely that the integrator had a body of experience to draw from. So if my motorized block happens to have advantages, this is not due to any brilliance on my part, but some unknown integrator's knowledge.

This is proving to be an interesting discussion that may help us tease out elements of that knowledge that we can use.
rjlittlefield wrote:Going to more than the original tensioning, however, I notice that when the motor runs continuously, there is a systematic variation in the sound depending on where the motor is in its rotation. I speculate that this is due to that slightly off-center pulley that I mentioned. . .
At my original, slightly tighter belt tension, I also noticed a change in sound that had a distinct periodicity. It happened whenever a particular portion of the belt went around one particular pulley (I forget which one). I happen to be a trail runner who switches to a treadmill when the weather requires it, and my treadmill exhibits a very similar sound change when the wide rubber belt's diagonal seam crosses one of the two rollers (again, I forget which roller, drive or drone). My instinct is that our timing belts, like my treadmill's underfoot belt, have a seam, that this seam has physical properties that cause a slightly altered load on the motor as it rounds at least one of the pulleys, at least at certain tensions.

This sound-change has gone away with the looser tension--an observation parallel with yours.
rjlittlefield wrote:I gather from all this that there's some range of acceptable tensions. . . .
Agreed.

Cheers,

--Chris

--edited typo
Last edited by Chris S. on Mon May 18, 2015 12:18 am, edited 1 time in total.

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Chris S. wrote:Yikes! That's a lot of backlash.
Yes. It's a habit of us math guys to push the limits, so if somebody writes "the lowest tension at which the belt(s) won’t slip", I'm very much inclined to try that. With this belt and these pulleys, and paying attention to backlash (which Bando may not care about), the advice is not optimal.
it's notable that your timing pulleys are small, compared with mine, and that your belt span is long, compared with mine. I wonder if these factors, in addition to your gearing down on the motor side, contribute to backlash? I'm tempted to suspect that timing pulleys ought to be reasonably large in order to reduce backlash.
Yeah, there are several things going on. In addition to the ones you've mentioned, my belt is FHT (1mm pitch) versus your MXL (2.03mm pitch). So the belt is correspondingly thinner and more flexible than yours, which I suspect is exacerbating the effect of the belt span.
This is proving to be an interesting discussion that may help us tease out elements of that knowledge that we can use.
Agreed.

--Rik

JGVilla
Posts: 62
Joined: Thu Dec 27, 2007 2:44 pm
Location: Netherlands

Post by JGVilla »

Thank you for all this very valuable and useful practical information. =D>

The following links give a good theoretical background of most factors involved in timing belts and timing pulleys.
You guys are probably familiar with these documents, but for anybody trying to get a hang of it, these documents may be of interest.

SDP SI - Timing belt design and installation suggestions
Also check out the next page, via the link at the bottom of this page.

Gates Mectrol - Timing belt theory
This document does not particularly handle MXL belts, but gives good background information with lots of formulas.

For what it is worth I'm using Gates Powergrip belts (¼" width),
As far as I know Powergrip GT2 belts (see document 1) are not available for MXL pulleys.

Btw, I'm in the process of changing my pulleys to a 30:90 (1:3) setup. Just waiting for the 30T pulley with 5mm bore to arrive.

edited: another document which might be of interest:
SPD SI - Timing Belts
-Jan-

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Labophot-2 stepper motor

Post by johan »

Just for the record, here is another way of attaching a stepper to a microscope arm in order to drive the fine focus using a timing pulley belt.

The arm in this case is a Labophot-2. I used a compression clamp design: 4 bits of children's meccano and 3x30cm lengths of 4mm threaded rod. These were all in my "nonsense box", a box of meccano bits that I kept knowing that one day they'd come in useful for something. That day has come.

The threaded rod is cut in 2 to create 6 lengths that have a nut either end keeping in a double layer of Meccano. Double so it doesn't bend. I cut a square out the centre of two of the Meccano bits using a dremel saw, and the shaft of the stepper sits in that.

The stepper has 8 small pads of sorbothane that sit between it and any scope surfaces to (hopefully) reduce vibration. I have a spare extra arm that will be attached when the timing pulley arrives next week to create a means of tension adjustment (add offset arm into timing pulley path to tighten etc).

Image

The stepper was bought from Cognisys, obviously it is all designed to be set up with Stackshot.
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

completed, and photo shows the third arm used to adjust tension. All seems to work.

Image
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

Post Reply Previous topicNext topic