0

I am currently working on project that uses the spectrum created by a diffraction grating. I am using a stepper motor to rotate my grating. I want the stepper motor to move exactly 0.01 degrees and wait for detector to take the reading. After the detector is done taking the readings, again rotate 0.01 degrees and repeat the above procedure.

I have been searching through the internet to find a stepper motor that rotates only 0.01 degrees but so far I have no success. After searching for a while, I found that microstepping will help me achieving the required angle but I am not sure about it.

JRE
  • 67,678
  • 8
  • 104
  • 179
Yadav Karthik
  • 1
  • 1
  • 3
    This sounds like a serious *mechanical* challenge rather than electrical. Even if you make the motor control with such a resolution, I have doubts the mechanical parts will even move. – Eugene Sh. Sep 11 '18 at 17:21
  • 1
    You need an encoder. If you get one, replace the motor with bldc. Otherwise it's a serious challenge for no gain. –  Sep 11 '18 at 17:25
  • 4
    Try attacking it from the other side. Have your motor run *at constant velocity*, and synchronize the readings with it. And note, that you don't have to synchronize with each angle you actually need, but you want to perform, say 36000 readings per revolution. And afterwards you can calibrate the "shift". – Eugene Sh. Sep 11 '18 at 17:26
  • What is the angular range of interest? ±1°, ±30°? Could you use a linear movement? How many mm? What step resolution? – Transistor Sep 11 '18 at 17:38
  • 1
    You need a gear drive – Scott Seidman Sep 11 '18 at 18:06
  • or https://www.newscaletech.com/micro-motion-modules/m3-rs-rotary-smart-stages/ – Scott Seidman Sep 11 '18 at 18:29
  • You could also use a stepper to drive a worm that turns a gear, on which your grating is mounted. – Scott Seidman Sep 11 '18 at 18:32
  • 4
    @ScottSeidman Any extra mechanical steps will only add to the backlash. – Eugene Sh. Sep 11 '18 at 18:36
  • 2
    @EugeneSh. Absolutely, but adding an output encoder might be the cheaper, faster, get 'er done approach. Also, if the motor only moves in one direction, backlash might not be an issue. Bottom line is that this sounds like a real design problem, worthy of the time needed to put into it. – Scott Seidman Sep 11 '18 at 18:48
  • How 'bout a linear drive, pushing/pulling on one end of a longish beam with a pivot at the other end? –  Sep 11 '18 at 18:55
  • A linear stepped :D –  Sep 11 '18 at 20:06

5 Answers5

4

Do you want the good news first, or the bad?

The bad news is that what you are asking for is not available.

The good news is that what you need (rather than what you asked for) is available. On eBay. It's called a "motorized rotary stage". An example would be this unit, which has a data sheet here, although the unit for sale is apparently an older unit which does not exactly correspond to the data sheet. At a guess, though, it will do much better than you need.

And eBay has a lot of this sort of unit. Of course, as in this case they are often old, unsupported, and you'll have to learn how to work with proprietary interfaces.

If you're more adventurous (that is, wealthy), various companies such as Thorlabs and Newport will gladly sell you new units much better than you need. Well, and almost certainly more expensive than you can afford, but you didn't set price limits in your question.

WhatRoughBeast
  • 59,978
  • 2
  • 37
  • 97
3

Stepper motors used in a microstep configuration are no longer working as step devices, they are operating more like a speaker where the current sets the position between steps. The problem is the pullout torque for very small microsteps. You have two ways to go to get this sort of resolution:

  1. Use a closed loop controller/servo such as the Clearpath CPM-MCVC-2310D-RLN. Drives such as this ensure that your get a step based on an encoder, and not on the BLDC motor. This would step at 0.45 degrees and you could belt drive or worm gear down to get your required 0.01 degree movement. I would not recommend this solution as it is a huge overkill for your task, but it is somewhat simple to engineer with less than 50:1 gearing. Providing you have decent home sensors and drive in only one direction it should be an able solution.

  2. You could use a lightweight stepper motor and a microstep controller in an open loop configuration. I'd suggest you should look at the Trinamic range of stepper controllers, they are cheap (and readily available, such as the TSMC2100 on Ebay) and can achieve 256:1 microstepping. You need the smallest stepper motor you can get, and here a Nema8 may well be suitable. Again you need to gear down the stepper to your grating drive.

One interesting way to achieve a very high gearing ratio (300+:1) is to do as was done on early 11-15" disc drives (I worked on these in the 70's) and 8" floppies. Use the stepper motor to drive an accurately ground pin with steel belt drive to a large lever. This works well if you only want a total of say 20 degrees of total rotation, but the surfaces need to be accurately ground. I've used this type of drive on linear positioning stages without the need to use high levels of microstepping. You can achieve micron movements with only x4 or x8 microstepping. If you are hacking together a single grating stage you might be able to find an old 8" floppy with this type of positioning system you could be re-used. I have also built positioners like this by using 3 long feeler strips of 0.001" which works well for positioning stages with absolutely no backlash.

enter image description here

Jack Creasey
  • 21,428
  • 2
  • 15
  • 29
  • Good thought. The same can be done now with a cascade series of pulley ratios for steppers. https://ae01.alicdn.com/kf/HTB1v9XgMpXXXXXmaXXXq6xXFXXXb.jpg however the floppy and HDA's used steel bands for low backlash. If there is not much load, then the backlash is minimal. – Tony Stewart EE75 Sep 12 '18 at 02:01
  • @TonyEErocketscientist The backlash is almost non existant. I've built functioning units using 3 feeler gauge strips, the only problem is tensioning until you tighten the holders. – Jack Creasey Sep 12 '18 at 02:16
  • Yes I was also in the biz 600 tpi 5.25" steppers with slew rate acceleration control and viscous dampers with < 10% overshoot. – Tony Stewart EE75 Sep 12 '18 at 02:30
  • @TonyEErocketscientist Interesting, I found the viscous dampers to have too much inertia. I used magnetic damping with a thin aluminum disk. To each his own, I guess. – Jack Creasey Sep 12 '18 at 02:38
  • Nippon Peripherals Ltd. (NPL sub. of Hitachi) use a thin brass disc in a thin wheel with oil. Very fast seek times. – Tony Stewart EE75 Sep 12 '18 at 02:51
2

Even a small amount of friction will prevent you from achieving 1/256 step as will anything else that requires torque. 0.01° repeatibility or accuracy requires a seriously good mechanical setup. Stepper motors are not nearly that accurate typically for the complete steps, which could be corrected but that's another complexity.

The torque to move it one microstep will be only 0.6% of the holding torque for a full step. See this paper from a motor manufacturer.

If you have a budget of a few thousand dollars, there are manufacturers who produce this kind of component, for example with an ultrasonic motor and an encoder.

Or, if you could relax your requirements by an order of magnitude or more, I would be a lot more optimistic about your chances.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
0

I have a Laser light with 2 counter rotating diffraction gratings to create thousands of RGB dots moving slowly with the stepper-motor gear reduction drive. They are quite miniature and low cost with the lasers with a remote control for 2 low speeds and stop for long cyclic pattern creations. I imagine it has 1000:1 gear ratio from the 10s second (est) max revolution time. Slow is about 4x longer. There is some jitter evident from the gears but a laser dot with a jitter of 1cm at 10m distance that can move 100 cm/s continuous at this range over a 40 second cycle is ballpark perhaps what you can tolerate. All numbers are guestimates.

I suggest you use full steps for max torque as 1/16 step torque is very uneven. Use small to large nylon belt tensioned pulleys for minimal backlash to obtain 1000:1 gear ratio or as required.

Do multiple cascaded pulley pairs to keep small. Get the largest pulley size you can deal with. enter image description here

or Shop around for garden lights which use the geared steppers.

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182
0

Most of the steppers are 1.8 or (rarely) 0.9 degrees per full step. There is a halfstep mode which uses both one coil and two coils energuzed. Any over microstep mode will not guarantee you the exact angle of the motor shaft.

So you can easy achieve 0.45 degrees resolution.

That means that you will need any type of reduction: belt, gear, worm or planetary. If you will do so you will need a 1:45 reduction (this will give you exactly 0.01 degrees per half step of 0.9 degree motor).

Any other solutions will require a feedback which in my opinion will be much more expencive at accuracy that you need.

Roman Matveev
  • 2,942
  • 7
  • 32
  • 75