×

INDI Library v2.0.6 is Released (02 Feb 2024)

Bi-monthly release with minor bug fixes and improvements

Precise backlash measurement

  • Posts: 319
  • Thank you received: 25
Hello,

Is there is a precise way to measure the backlash for a focuser? Online and youtube methods depend either on visual observation of the start of movement or Digital Caliper. I tried both but every time I get zero backlash results!!!


Thank you
2 years 11 months ago #70076

Please Log in or Create an account to join the conversation.

  • Posts: 1309
  • Thank you received: 226

Replied by Andrew on topic Precise backlash measurement

I used an artificial star averaging 2 frames. Then adjusted the backlash value until the points focusing in matched well with the points focusing out.
2 years 11 months ago #70078

Please Log in or Create an account to join the conversation.

  • Posts: 319
  • Thank you received: 25

Thanks Ihoujin for the tip. Can you please advise more details on the steps?
2 years 11 months ago #70079

Please Log in or Create an account to join the conversation.

  • Posts: 1309
  • Thank you received: 226
It's pretty straightforward. Setup an artificial star at a workable distance for your optics. I have the Hubble 5-star Artificial Star with 5 star sizes to choose from. Use the smallest one you can. Set exposure and gain and ensure Auto Select star can pick out the artificial star.
Initial steps should be set high enough to make a noticeable change. Then run an autofocus routine. Observe the graphs. Once it reverses directions note if the reverse HFR values are above of below the earlier HFR at the same step position.
If you are above, you have uncorrected backlash. Below, the backlash is overcompensated for.
You can do this with real stars, but atmospheric seeing will reduce the accuracy. Using average of 2 or more frames helps.
The following user(s) said Thank You: Mohamed
2 years 11 months ago #70080

Please Log in or Create an account to join the conversation.

  • Posts: 276
  • Thank you received: 52
If the focus routine always moves to final position in the same direction then oversizing the backlash steps when reversing basically makes it a do not care because the backlash cancel's out.

Think about it this way, currently at 500 with actual backlash = 25
Move back 200 physically moves it to 325 (175 = 200 - 25 backlash), counter though shows 300
Move forward 200 (physically moves 175 = 200 -25 backlash) up to 500 physical and counter shows 500
2 years 11 months ago #70105

Please Log in or Create an account to join the conversation.

  • Posts: 319
  • Thank you received: 25

Thank you Ihoujin, will give this a try in the cloudy week...
2 years 11 months ago #70113

Please Log in or Create an account to join the conversation.

  • Posts: 1309
  • Thank you received: 226
The linear algorithm approaches the final position from one side in an effort to eliminate the consequences of backlash. Which is fine. But it's only true for that algorithm. Iterative and polynomial don't do that, so good backlash compensation is preferred and will more easily fit a better curve.

As a matter of opinion, I have come to prefer the speed and accuracy of polynomial with a reliable focuser.
2 years 11 months ago #70114

Please Log in or Create an account to join the conversation.

  • Posts: 319
  • Thank you received: 25


Hi Gene,

The problem is as Ihoujin mentioned if you are approaching the final focus point from one direction as in linear it won't matter, but the problem is that (1) IN and OUT backlashes could not be equal, and (2) This confuses the algorithm as at the final position, the algorithm is not equally going IN and OUT, and that is the cause of the shift of backlash that causes the problem.
2 years 11 months ago #70116

Please Log in or Create an account to join the conversation.

  • Posts: 276
  • Thank you received: 52
Hi Mohamed,

True backlash in opposite direction can vary but more than likely comes from drawtube tilting than gear train backlash, which is more repeatable -if- the gears are fixed to the shaft and do not ride on the shaft they are mounted to. Coming from drawtube tilt also implies amount may change depending on how far the drawtube is extended.

Here is a video I did of a stepper motor head used, gears only in the head. Repeatability is good considering any slip that may have occurred of the dial indicator on the moment arm and the 5 pound weight hanging from the 2 inch moment arm :-)

Stepper motor head video from 2012
2 years 11 months ago #70123

Please Log in or Create an account to join the conversation.

  • Posts: 1009
  • Thank you received: 133
One thing you can try:
Most (all?) focusers have an option to reverse the direction.  So do an auto-focus run with the linear algorithm (that works around a possible backlash) in the normal direction, and write down the value (also look at the curve to verify it is a reliable one).  Then activate the 'reverse' option in the focuser driver, and repeat the AF run.  This time, because of the reversal, it will always end moving outward, and you get the focus 'coming from the other side'.  The difference between the two values you get that way would be your backlash.
 
2 years 11 months ago #70210

Please Log in or Create an account to join the conversation.

  • Posts: 319
  • Thank you received: 25

Hi Peter,

Thank you for your suggestion, but when I tried it, I got variable readings. Please find the below table...

 
2 years 11 months ago #70400
Attachments:

Please Log in or Create an account to join the conversation.

  • Posts: 319
  • Thank you received: 25
Hi Gene,

Thank you for the video, but I couldn't grasp how it was related to the focuser backlash. From what I can tell your stepper motor worked well :)

 
2 years 11 months ago #70401

Please Log in or Create an account to join the conversation.

Time to create page: 0.689 seconds