Hello Guest it is May 25, 2019, 09:47:31 PM

Author Topic: Precise calibration but inconsistent incremental movements  (Read 467 times)

0 Members and 1 Guest are viewing this topic.

Offline Pmw

*
  •  2 2
    • View Profile
Precise calibration but inconsistent incremental movements
« on: July 14, 2018, 08:18:56 PM »
Hi all,

I'm new to the forum but have spent quite a bit of time looking through past topics finding a lot of helpful information along the way. However I am at a total loss for a solution to the following problem I am encountering. Some previous posts have eluded to what sounds like the same or similar problem but no definite solutions seem to have been found.

I have a self built 4x8 rack and pinion gantry router, running a closed loop system with 2hss86h drivers and nema 34 steppers with encoders to suit on both the x axis and the y/slave axis (one stepper either side of the machine). Each stepper then runs through a 6:1 planetary gear box.

I have set the steps per unit for both axis using an accurate 1000mm steel rule and microscope over the full length of the rule.

The y axis works exactly as it should, I can move from 0 to 1000mm and it hits the mark perfectly. I can move along the axis incrementaly using MDI Y100,Y200,Y300 etc and it hits each mark. It works in 10s or single units.

With the x axis I again set steps per unit over the full length of the rule. As with the y axis I can move from 0 to 1000mm and it with hit the mark perfectly. However if I move along the axis incrementaly X100, X200, X300 the movements are inconsistent, with some running too long and some coming up short, there seems to be no pattern.

X0 = 0
X100 = 100.05
X200 = 200.10
X300 = 300.20
X400 = 400
X500 = 499.95
X600 = 599.95
X700 = 699.9
X800 = 800
X900 = 900.05
X1000 = 1000

I have checked the machine over mechanically and everything is sound, no loose grub screw holding the pinion to the planetary shaft, no lose grub screw holding the planetary shaft to the stepper motor. Steppers are bolted down tightly. Large springs engage the pinion with rack at all times. The rack is fix tight, the spindle is trammed and fixed tight and there is no binding along the axis. I can even perform the above incremental movements using my fully body weight to resis the x axis movement and it just moves me out of the way and still hits the inconsistent measurements shown above time and time again.

I have also carried out the test further along the X axis, zeroing at what would have been my previous X500 position, and again it gives the same inconsistent results as shown above (I was only able to  test to X500 that time as I was only using the second half of the axis). Yet this showed me that the errors weren't localised to a section of the axis as it behaved the same when moved to another part of the axis, so ruling out possible damaged or deformed rack or pinion.

I have also swapped the x and y stepper motor with the thought that the encoder on the x axis stepper might be faulty, but the y axis stepper performed exactly the same when in its place

With regard to Mach3 both the x and y axis are set up exactly the same. I have set the step and direction pulse within the motor tuning page to 3 and 3 after reading some previous treads on this forum but with no luck.

Sorry for waffling on but does anyone have any ideas to a cause or a solution or this problem?
Has anyone encountered anything like this previously?
I have pretty much run out of ideas now

Thank you for reading this far and thank you for any help in advance
Cheers
Chris

Re: Precise calibration but inconsistent incremental movements
« Reply #1 on: July 14, 2018, 09:25:42 PM »
Hi,
sort of sounds like a rounding issue.

Can you post the 'steps per unit, 'max velocity' and 'acceleration' values for both axes?

Craig
My wife left with my best friend...
     and I miss him!
Re: Precise calibration but inconsistent incremental movements
« Reply #2 on: July 14, 2018, 09:45:32 PM »
Hi,
have you got the software on your PC and have you connected the serial cable to the drive to program it?

Do you have the ALM output of the drive hooked up to Mach? Do you have any provision for errors?
Can you confirm the position loop parameters are identical for each drive, the position error particularly.
Can you run the offending axis with the error panel displayed....what happens?

Craig
My wife left with my best friend...
     and I miss him!
Re: Precise calibration but inconsistent incremental movements
« Reply #3 on: July 15, 2018, 02:01:01 AM »
Hi,
one last thought....could it be the rack?

While the errors seem to vary depending where on the rack you are if the error is always 300.2 say then it might be the rack.

Craig
My wife left with my best friend...
     and I miss him!

Offline RICH

*
  • *
  •  7,334 7,334
    • View Profile
Re: Precise calibration but inconsistent incremental movements
« Reply #4 on: July 15, 2018, 06:42:26 AM »
Quote
an accurate 1000mm steel rule and microscope over the full length of the rule

Just a few thoughts on the matter.
The basis for any measurement is the reference you are using which provides comparison of length.
You should have complete confidence in the reference you use. The optical metal scales we used were
calibrated. A 1000mm calibrated scale would be VERY  expensive and any deviation from a start point
to over the entire length and between divisions are known. Can measure to 0.001" ( 0.0254 mm) @ 17 feet
for example.

The scale you have is surely not to the caliber above,but, is usable to a degree, and only if the calibration
reference used is acceptable. I doubt very much that your scale is accurate to 0.05mm (0.002") at all locations
along it's length. Remember any measurement is only as accurate as the reference used! Also note that a mesurment
is only relative to the length being measured. Yes, you can then compare different items measured, but any difference is
realy only relative to the reference used ( may i also add that setup for measuring is very important).  

So one can ask a seemingly stupid question "whats your reference to the linear length measured".

Just some things to think about ..........

Practicaly speaking, one must ask "What is being measured"?
It is length, but that lenght is made up of individual components. The gear box should be repetative in one direction
over it's complete reduction, but it probably will not be in the opposite direction and  would say the same for the rack & pinion.
Now if you are using an electronic device which is based on steps to relate to length then likewise you should have confidence
that the steps are correct. Now you need a different calibration device. We used a calibratated counter accurate to four  parts
in 29x10 6th.

So if one wants to be anal about calibrating their machine, I will just say that you have what you  have, each component is contributory
to the system accuracy, and what you measure is only as good as the basis.


BTW, I also believe it is your rack & pinion along with the gear box. Need to isolate each and compare to the other axis to
be meaningfull.

FWIW,
RICH

Offline Pmw

*
  •  2 2
    • View Profile
Re: Precise calibration but inconsistent incremental movements
« Reply #5 on: July 17, 2018, 12:45:50 PM »
Hi guys,

Thanks for your replys and input. After spending alot of time trouble shooting and testing different things which both of you guys touched on, I believe that the errors have been caused by an eccentric pinion. The circumference of my pinion is 137.4 mm, and when I MDI multiples of this distance form zero, the marks are hit spot on (as they should as they are full revolutions). However distances anywhere under this value can run long or short depending on where abouts they land on the rotation of the eccentric pinion. The  pinions are probably the cheapest thing on my machine and I must have fallen lucky with the y/slave axis. I removed the pinion and measured the hole to be off centre by 0.15mm at worst. Anyway lesson learned, I'll do the machining myself with the new one. Thanks again for you time and sharing your knowledge.

Cheers
Chris
« Last Edit: July 17, 2018, 12:49:47 PM by Pmw »

Offline RICH

*
  • *
  •  7,334 7,334
    • View Profile
Re: Precise calibration but inconsistent incremental movements
« Reply #6 on: July 17, 2018, 06:19:04 PM »
The trick is to be able to isolate each item that makes up the system and glad you found the culprit.

RICH