Questions about linear rail drive gear

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

Moderators: rjlittlefield, ChrisR, Chris S., Pau

skarabaeus
Posts: 10
Joined: Fri May 15, 2020 5:54 am
Location: Austria

Questions about linear rail drive gear

Post by skarabaeus »

Hello friends, I have a question about the drive of linear units.

Does the use of planetary gears make sense. The positioning accuracy should have advantages at high scales.

Next question: a toothed belt drive has disadvantages compared to a direct drive. With a toothed belt, I could gain a lot of strength through the translation. Are there any suggestions on how to eliminate the transmission of engine vibrations.

As far as I know, Stonmaster is the only manufacturer of the Stackmaster who uses screw gears 1:20 (vertical 10kg force). Why does no other manufacturer of macro slides do this and where can you get inexpensive good gearboxes? When I build an indoor stacking system, something speaks against the use of larger motors, e.g. Nema 23 strength actually can't hurt I think. Does any of you happen to be using the jmc ServoStepper?

Thank you in advance for your type, I hope the google translator brings my questions right.

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: Questions about linear rail drive gear

Post by ray_parkhurst »

Hi @skarabaeus,

You have lots of questions!

The problem with planetary gears is basically the same as with most gears...backlash. Microscopes eliminate backlash by working vertically such that the weight of the stage keeps the gears always in contact in one direction and thus minimizing backlash. It's a similar principle to spring-loading. So if you are looking to build a planetary gear-driven system, you should ensure spring-loading or vertical operation to minimize backlash, which can be fairly severe without mitigation.

Toothed belts can give excellent results. I'm not sure about your question regarding engine vibrations in relation to the toothed belts. There should be no more vibration (due to the stepper motor) in a belt system vs a direct-drive, though if the belt is not properly-fitting it can cause its own vibrations and slight slip-related inaccuracies.

I agree with you on larger motors, which are used mainly to get more torque. I have found that there is enough friction in most linear rails that even for vertical systems there is little need for high holding torque. This makes using very small motors feasible. Personally I prefer the motors with the lowest holding torques as these tend to microstep the best. Big motors with strong magnetic systems having high holding torques tend to "lock-in" to their full-step positions unless strongly-driven, and thus also tend to make the most noise and vibration. I prefer the motors with the lowest holding torques in order to minimize vibrations. I also add dynamic dampers, but this requires dual-shaft motors, which are harder to come by.

I can't help with the Stonemaster, ServoStepper, or similar questions as I have no experience with them.
Last edited by ray_parkhurst on Tue Sep 14, 2021 11:33 am, edited 1 time in total.

Doppler9000
Posts: 111
Joined: Thu Oct 13, 2011 3:56 pm

Re: Questions about linear rail drive gear

Post by Doppler9000 »

For clarity, it is “stonemaster”.

https://www.stonemaster-onlineshop.de/f ... er-studio/

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Re: Questions about linear rail drive gear

Post by elf »

Why do you say toothed (timing) belts have disadvantages? I just added glass linear scales with 5 micron resolution to my rose engine. 1 micron and .5 micron scales are available. The sliderest (stage) is driven with a 82oz-in Nema 24 stepper, TR8 2mm pitch leadscrew, and 5:1 gear reduction using GT2-3mm pulleys and belts. Pulleys were 3d printed. Repeatability over 10 inches of travel is less than .001". There is no measurable backlash. Note: The glass linear scales aren't necessary, they just confirm the accuracy of the system. I would not hesitate to use GT2-3mm or GT2-2mm belts and pulleys on a macro focus stacking/stitching system although I would use USA or European manufactured belts.

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Re: Questions about linear rail drive gear

Post by mjkzz »

Belt driven system works very well and construction is simple (at least in my case), no noticeable backlash, low cost.

Some disadvantages are 1. long pitch, for example, 60mm, meaning that one turn of motor, the rail travels 60mm for a 30 teeth GT-2 gear. 2. lack of holding power when power off or when you have a power outage, this is a bit dangerous for vertical setup, not a problem for horizontal setup. 3. elasticity of the belt, those with steel thread inside could be better, but still an issue. So far, I have tried 5um step or for a 10x objective with large NA and it works, I do have doubts however going beyond this.

lothman
Posts: 957
Joined: Sat Feb 14, 2009 7:00 am
Location: Stuttgart/Germany

Re: Questions about linear rail drive gear

Post by lothman »

elf wrote:
Tue Sep 14, 2021 2:12 pm
Why do you say toothed (timing) belts have disadvantages? I just added glass linear scales with 5 micron resolution to my rose engine. 1 micron and .5 micron scales are available. The sliderest (stage) is driven with a 82oz-in Nema 24 stepper, TR8 2mm pitch leadscrew, and 5:1 gear reduction using GT2-3mm pulleys and belts. Pulleys were 3d printed. Repeatability over 10 inches of travel is less than .001". There is no measurable backlash. Note: The glass linear scales aren't necessary, they just confirm the accuracy of the system. I would not hesitate to use GT2-3mm or GT2-2mm belts and pulleys on a macro focus stacking/stitching system although I would use USA or European manufactured belts.
I do not doubt that your rig meets all requirements for your stacking application, but you are describing a very basic and low cost solution and state "no measurable backlash". That is hard to believe. Why on earth things like aerostatic bearings and piezo actuators are needed :shock:

P.S.: on focus stacking we usually drive in one direction so backlash is no matter you only need steps with the same size, that would be different if we have a milling table and want to do a circular motion.

ray_parkhurst
Posts: 3413
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Re: Questions about linear rail drive gear

Post by ray_parkhurst »

Belt drive also introduces additional variables into the system. Backlash is already present in the linear motion components themselves, and adding a belt drive won't make it better. However, belt drive does give an additional degree of freedom to choose a gear ratio, allowing smaller step sizes at very low additional cost. It's quite easy to achieve a 4x reduction in step size, so a 400-step motor becomes 1600 steps at the rail. Unfortunately this amplifies the backlash and other inaccuracies of the rail (from motor's perspective) but reduces the backlash due to the gears (from rail's perspective).

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Re: Questions about linear rail drive gear

Post by elf »

Everything depends on the desired magnification. My desktop stacking/stitching system can do 1 micron steps (+-.2 microns). It's also belt driven. I did spend several hours lapping the leadscrew. Zerene Stacker makes it easy to measure the accuracy of your system.

Generally speaking Acme or Trapezoidal leadscrew can't be back driven, which means a belt driven system should work fine on a vertical setup. If the vertical setup is heavy enough to back drive the leadscrew, then a counterweight or gas shock should be added.

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Re: Questions about linear rail drive gear

Post by mjkzz »

elf wrote:
Wed Sep 15, 2021 11:07 am
Everything depends on the desired magnification. My desktop stacking/stitching system can do 1 micron steps (+-.2 microns). It's also belt driven. I did spend several hours lapping the leadscrew. Zerene Stacker makes it easy to measure the accuracy of your system.

Generally speaking Acme or Trapezoidal leadscrew can't be back driven, which means a belt driven system should work fine on a vertical setup. If the vertical setup is heavy enough to back drive the leadscrew, then a counterweight or gas shock should be added.
Oh, that is a good idea, adding a counter weight for vertical setup, but I think that solves the problem of a weak motor, not when you lose power, at least with my setup, the weight of camera, lens will definitely crashing down (fast and furious :D ).

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Re: Questions about linear rail drive gear

Post by mjkzz »

OK, here is a video (part of another) on my personal account, showing how dangerous it is with VERTICAL belt driven setup. Though the motor and driver are still powered after powering off the Raspberry Pi, the idle power was set to 0.15A (I think) which is not enough and when RPi is powered off, it sets the drivers to idle state and booom, it crashes down fast. Of course, one can set idle power to be much higher, but that can be more problematic than beneficial (motor over heat when idling), etc. Plus, when there is a power outage, it won't help.

https://youtu.be/fD8NirPtYGI

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Re: Questions about linear rail drive gear

Post by elf »

I can see why that would happen :shock: Fix is pretty simple, just add a mechanical brake which is activated whenever the stepper is disabled. Wire it on the Enable pin the same as you would an NC limit switch.

chris_ma
Posts: 570
Joined: Fri Mar 22, 2019 2:23 pm
Location: Germany

Re: Questions about linear rail drive gear

Post by chris_ma »

this is all very interesting.

I would have thought that both, belts and gears, make the total system backslash worse unless going for the expensive high precision bits, but it appears that even low cost options work very well, specially when we add preload and only go in one direction.

I've been thinking on how to use a long rail to avoid repositioning the camera for different lenses and still have high accuracy on a short travel, the first idea was to use a long rail with 6mm ball screw and mount another rail with 1mm ball screw on top of that, but I'm a bit worried about getting more instability through to the added weight and junction points.

using a belt or gear on the motor might be an easier and more solid solution. has anybody experimented with a using a high quality rail, like a THK KR33, with a 6mm ball screw and a gear or timing belt to increase accuracy?
chris

lothman
Posts: 957
Joined: Sat Feb 14, 2009 7:00 am
Location: Stuttgart/Germany

Re: Questions about linear rail drive gear

Post by lothman »

chris_ma wrote:
Thu Sep 16, 2021 2:01 am
using a belt or gear on the motor might be an easier and more solid solution. has anybody experimented with a using a high quality rail, like a THK KR33, with a 6mm ball screw and a gear or timing belt to increase accuracy?
on my setup you can very easy slide the rail on the Thorlabs rail by openening the Arca clamps. I would not do rail on rail. when then I would get the longest available KR2602 rail (02=2mm pitch) and use a stepper direct coupled to the spindle.

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Re: Questions about linear rail drive gear

Post by mjkzz »

elf wrote:
Wed Sep 15, 2021 11:02 pm
I can see why that would happen :shock: Fix is pretty simple, just add a mechanical brake which is activated whenever the stepper is disabled. Wire it on the Enable pin the same as you would an NC limit switch.
Sure, there are motors specifically for this kind of situation, I think they are called self-braking or self-locking (I do not know their English name), these motors have brake mechanism around their shaft. But that is going to increase cost and redesign of controller.

The rail is superb in terms of performance, smooth, backlash free (not noticeable), wobble free (even under 10x), and I can get meter long ones . . . and best of all, most of it are off the shelf from companies like ratrig in EU, OpenBuilds in the US.

chris_ma
Posts: 570
Joined: Fri Mar 22, 2019 2:23 pm
Location: Germany

Re: Questions about linear rail drive gear

Post by chris_ma »

lothman wrote:
Thu Sep 16, 2021 2:13 am
chris_ma wrote:
Thu Sep 16, 2021 2:01 am
using a belt or gear on the motor might be an easier and more solid solution. has anybody experimented with a using a high quality rail, like a THK KR33, with a 6mm ball screw and a gear or timing belt to increase accuracy?
on my setup you can very easy slide the rail on the Thorlabs rail by openening the Arca clamps. I would not do rail on rail. when then I would get the longest available KR2602 rail (02=2mm pitch) and use a stepper direct coupled to the spindle.
at the moment I also adjust large distances by sliding the vertical setup on a thorlabs rail, but since need to change things quite often and I need to return to specific points, having a fully motorised way would save me some fiddling.

the full range of distance adjustment I need is about 50cm, so a KR26 unfortunately wont be long enough, and the KR30H and KR33 have 6mm or 10mm ball screws, what's why i'm wondering if step size could be reduced with gears or belts (I aim for about 10um step size)
chris

Post Reply Previous topicNext topic