4th Axis Rotary Accuracy - Problems
I have one of these rotary axis on my router. I have it tuned to make 10 full rotations clockwise and then back counterclockwise. It is coming to "zero" each way. BUT, when I use it in machining for indexing my operations sequence has the rotary indexing -90, 90, -90, 90 then back to zero. I get a stacking error that ends up causing problems with part dimensions.
https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1
Right now I am using it as a manual indexer and setting the position with measuring tools before starting that next operation. It works but kills using the full function of the axis via programmed moves.
Thoughts on ways to deal with this or what may be causing the errors?
Re: 4th Axis Rotary Accuracy - Problems
You probably need to fine tune the steps per to get a more accurate setting.
Also, test things with acceleration settings to make sure it isn't skipping at higher movement speeds.
Sent from my SM-G996U using Tapatalk
Re: 4th Axis Rotary Accuracy - Problems
I have fine tuned the steps. Started 0-360 and got it very close, then 0-360-0 and tuned more. The 0-3600-0 and have it dialed in by 5 decimal places on the calibration.
I will look at the acceleration numbers and slow them down. I don't have it moving real fast. Again just using it for indexing.
Thanks!
Loren
Re: 4th Axis Rotary Accuracy - Problems
Is the cable for the stepper shielded? Do you have an end stop on it? I use an inductive end stop on mine. Printed a bracket to home it and another bracket that attaches to the chuck back side. I put a 3mm bolt in the chuck side to trigger the sensor.
That way I have an accurate starting point each time. If indexing only, you could home on each rotation to get back to zero repeatably. I haven't had to use it like that myself. I just thought I would have to if forces were high enough to change position during a cut.
I know it's not a great option considering you definitely purchased it for the ability to set an angle correctly.
Sent from my SM-G996U using Tapatalk
Re: 4th Axis Rotary Accuracy - Problems
Another thing I didn't ask did you have the issue only when actually doing work with it or also just testing the motion in the pattern you originally mentioned?
Sent from my SM-G996U using Tapatalk
Re: 4th Axis Rotary Accuracy - Problems
Using high quality shielded cables and grounded at panel.
No end / zero stop on the rotary. I set a g54 zero when setting up work. Have the problem when running a program or when testing using MDI control. If I give an MDI of G0 A90 it will not travel to a measured 90 deg., using a level. If I give it a MDI G0 A360 I get a complete revolution and back to zero, using a level to measure zero.
Adding a home sensor and driving back to home with each move would take the stacking error out of the equation.
Re: 4th Axis Rotary Accuracy - Problems
It's pretty crazy that you can repeatably go to 360 but 90 is off. That would have me obsessed with figuring it out.
There must be an answer for this specific issue. I am sorry I can't add more right now. I would love to see the fix though. Someone has to add to this sometime today.
Sent from my SM-G996U using Tapatalk
Re: 4th Axis Rotary Accuracy - Problems
I appreciate your thoughts on it, you gave me a few things to think about.
Re: 4th Axis Rotary Accuracy - Problems
Not too knowledgeable about 4th axis indexers, but; I presume a 360 rotation of your indexer equals a whole number of of steps (I also presume your motor drive has an encoder?)? Does a 90 degree rotation also equal a whole number of steps. If not, could that be where your discrepancy is? Just a thought.
Re: 4th Axis Rotary Accuracy - Problems
I thought similarly but with micro step settings on the controller. The 90 degree thing and compound issue is something I can't explain with that. Unless full steps work fine for 360 and not for lesser numbers.
Sent from my SM-G996U using Tapatalk
Re: 4th Axis Rotary Accuracy - Problems
No Encoder.
Stepper motor =200 steps/rev
Micro step 10X so 2000 steps/rev
6:1 belt drive reduction from motor to rotary head, so 12,000 steps per rev.
12,000 /360 deg = 33.333333... steps per degree
or 2999.99 steps for 90 degree move.
I can't imagine that small rounding error is giving me the problem. Drive is a Gecko G540. Using UCCNC for control.
Re: 4th Axis Rotary Accuracy - Problems
Most rotary tables have a little backlash; the worm gear turns a little before it engages the pinion gear when changing directions. Have you measured this and accounted for it in your commands? If you start with the gear engaged in one direction and don't start turning the other way, backlash isn't a problem, but as soon as you reverse, it is.
Re: 4th Axis Rotary Accuracy - Problems
It is a direct belt drive, with no gear drive involved. I have not been able to determine any backlash in the belt drive. It is that small if any.
Re: 4th Axis Rotary Accuracy - Problems
Quote:
Originally Posted by
awerby
Most rotary tables have a little backlash; the worm gear turns a little before it engages the pinion gear when changing directions. Have you measured this and accounted for it in your commands? If you start with the gear engaged in one direction and don't start turning the other way, backlash isn't a problem, but as soon as you reverse, it is.
I do not think it's a worm gear. It's belt driven from the stepper to a larger gear on the axle.
Still, backlash could be a small part and I suppose it can multiply at every change.
What happens when going 0 to 360 and 360 to 0 then going 0 to -360 a few times? Backlash would add more each reversal and then doing smaller movements while going back and forth would show that well. I just don't see the being enough room for much backlash with the belt driven setup it has.
Sent from my SM-G996U using Tapatalk
Re: 4th Axis Rotary Accuracy - Problems
You may also want to do the math (2999 or 3000 steps equals how many degrees of rotation) to see if there is a correlation to your error.
Re: 4th Axis Rotary Accuracy - Problems
You are missing/gaining steps, I do not see any other explanation. Rounding, backlash, calibration - none of that should cause a cumulative error, unless there is a bug in the controller math.
Re: 4th Axis Rotary Accuracy - Problems
I was able to resolve the problem. My Lenox level just did not have enough accuracy between the level bubble vial and the plumb bubble vial. I bought a digital angle finder with an accuracy of 0.1 degrees and that increased the accuracy of the calibration of the steps per unit in UCCNC. It is returning to Zero and indexing within 0.2 degrees which is good enough for the work I am doing.
Re: 4th Axis Rotary Accuracy - Problems
Oh ok. I didn't think of how you were measuring the differences previously.
It's great that you got it figured out.
Sent from my SM-G996U using Tapatalk
Re: 4th Axis Rotary Accuracy - Problems
Quote:
Originally Posted by
LorenBMS
I bought a digital angle finder with an accuracy of 0.1 degrees and that increased the accuracy of the calibration of the steps per unit in UCCNC.
If you are "measuring" or "calibrating" your steps per unit, you are doing something wrong. Steps per unit is a hard number determined by the mechanics of your axis. You don't measure it with an angle finder, you don't calibrate it, you don't tweak it - you calculate it once and leave it alone. Unless you make an error in your math, your 90-degree angle will be exactly 90 degrees within one microstep.
Re: 4th Axis Rotary Accuracy - Problems
To expand on what CitizenOfDreams said, you just need to know the tooth count on the belt gears for a fixed and true ratio, then the steps per rev on your motor which is also a physical element. That will give you your deg/step as well as you can get unless you calibrate with extreme accuracy. At your step/rev count, 0.2 deg is massive.
The ratio is likely not exactly 6:1, do it by actual tooth ratio.