Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: UCCAPE on September 03, 2020, 06:07:27 PM
-
Z axis inconsistent movement
My Z axis is creeping upward during the program run.
I have run the following gcode with a repeatable error.
The original location of Z 0
Code Run
Z10
Z0
Will give me a Z offset of +0.002mm each time I run it. But Mach 3 display is Z0.
Original location of Z 0
Code Run
Z5
Z10
Z0
Will give me a Z0 each time I run it and Mach 3 display is also Z0.
Basically start at Z0 with Z10 Z0 will give me an error in Z but Z5 Z10 Z0 will be correct.
Does anybody know how to fix this?
-
Try reducing your Z axis Acceleration and Velocity (in motor tuning) by 50% and apply the changes. Does the problem still occur ?
Tweakie.
-
Try reducing your Z axis Acceleration and Velocity (in motor tuning) by 50% and apply the changes. Does the problem still occur ?
Tweakie.
Yes, have tried that. Even to 10% it still has that. The odd thing is if I do step up or step down it does not have any error. Z5 Z10 Z15 Z20 Z0 will give me no error. Z20 Z0 will give me a consistent error each time I run it.
-
What are the Acceleration and Velocity figures you are currently using ?
Tweakie.
-
Hi
My acceleration is 150mm's and velocity is 2800mm's
thanks
-
Will give me a Z offset of +0.002mm each time I run it.
0.002mm = 0.000079"
What is your theoretical and actual resolution for the Z axis ?
How are you measuring that error?
How is that axis driven and are the axis components of dimensional quality ?
Just a few practical considerations.
RICH
-
the 0.002 is an averaged number. as I did repeated z5 z0 10 times it becomes 0.02. I have also done z5 z10 z0 10 times it has no error.
thanks
-
If you repeat your "problematic" sequence 100 times, what is the result?
How do you measure micrometer difference? I guess that the resolution as well as repeatability of your machine is much worse than 2um.
-
If you repeat your "problematic" sequence 100 times, what is the result?
How do you measure micrometer difference? I guess that the resolution as well as repeatability of your machine is much worse than 2um.
my machine have about 0.01 resolution.
I'm measuring using a dial indicator. I fixed a dial indicator to the bed with the indicator pin to the bottom of the spindle shell. marked that as Z0.
I can see a noticeable change every time I run Z10 Z0, which raise Z to 10 them back to 0. it has a consistent change of the Z hight each time I run the code. It averages out about +0.002mm. I have set up the code to run 100 times it had about +0.19mm in Z hight.
With the same setup, the code Z5 Z10 Z0, which raise Z to 5 then raise to 10 them back to 0. it has no noticeable change of the Z hight regardless now many time I Run the code. I have run the code up to 200 times it had no change to the Z hight.
I have also run the code Z5 Z10 Z5 Z0 , which raise Z to 5 then raise to 10 them back to 5 then back to 0. it also has no noticeable change of the Z hight hight regardless now many time I Run the code.
Thanks
-
Perhaps one more test you could do is zero your Z axis to a gauge block (any solid block would do) – carefully slide out the block – run your test 100 times – slide in the block – measure the gap with feeler.
This would verify the integrity of your dial indicator.
Tweakie.
-
Perhaps one more test you could do is zero your Z axis to a gauge block (any solid block would do) – carefully slide out the block – run your test 100 times – slide in the block – measure the gap with feeler.
This would verify the integrity of your dial indicator.
Tweakie.
I have set up as you suggested and I only run the Z10 Z0 code 20 times without moving the block as my problem is consistently creeping up. By using a feeler gauge the result had about 0.05 error so average out it had about 0.0025mm error each time it run.
with the same setup, Z5 Z10 Z0 code made almost no error as I do not have a feeler gauge small enough to fit the gap if there is any.
-
This may not be useful information but some time back we had a similar problem to this and it turned out to be this:-
The machine was setup with 10 microsteps and low power standby, when the axis stopped moving it instantly went into low power mode and the motor snapped to the nearest full step. on restarting the motor continued from the full step point, we killed the power down mode and everything went back to working properly.
-
This may not be useful information but some time back we had a similar problem to this and it turned out to be this:-
The machine was setup with 10 microsteps and low power standby, when the axis stopped moving it instantly went into low power mode and the motor snapped to the nearest full step. on restarting the motor continued from the full step point, we killed the power down mode and everything went back to working properly.
power down mode? where can I find that option?
thanks
-
I have set up as you suggested and I only run the Z10 Z0 code 20 times without moving the block as my problem is consistently creeping up. By using a feeler gauge the result had about 0.05 error so average out it had about 0.0025mm error each time it run.
with the same setup, Z5 Z10 Z0 code made almost no error as I do not have a feeler gauge small enough to fit the gap if there is any.
I think Graham may have solved the mystery so lets hope the cure is a simple fix.
Which motion controller are you using ?
What make and model is your stepper driver for the Z axis ?
Tweakie.
-
I have set up as you suggested and I only run the Z10 Z0 code 20 times without moving the block as my problem is consistently creeping up. By using a feeler gauge the result had about 0.05 error so average out it had about 0.0025mm error each time it run.
with the same setup, Z5 Z10 Z0 code made almost no error as I do not have a feeler gauge small enough to fit the gap if there is any.
I think Graham may have solved the mystery so lets hope the cure is a simple fix.
Which motion controller are you using ?
What make and model is your stepper driver for the Z axis ?
Tweakie.
Hi
I can not find the options for low power standby or power-down option. can you please give me some direction on where that setting is?
I bought it from china-cncrouter dot com have no clue what is the motion controller and the stepper is just what came with the machine as well
thanks
-
Add to the topic I have also tried swap the Z stepper on to X-axis and Y-axis controller and it gives me no error running the code.
so that gives me
no error in Z stepper
- it has tried with Z5 Z10 Z0 code with no error
it has tried with the x-axis andy-axis controller and with X10 X0 code give me no error
that leave's with me only when I do Z10 Z0 code with error. must be a setting in MACH3 or a software issue.
-
that leave's with me only when I do Z10 Z0 code with error. must be a setting in MACH3 or a software issue.
or a bad z-axis controller
-
that leave's with me only when I do Z10 Z0 code with error. must be a setting in MACH3 or a software issue.
or a bad z-axis controller
I do not think so as this controller give me an accurate Z5 Z10 Z0 motion.
-
Hypothetically…
When a stepper assumes the nearest whole step position (from it’s current microstep position) it generally rotates the shortest distance.
If your Z move form 0 to 5 created a positive error and the move from 5 to 10 created a negative then one error would cancel the other.
Without knowing the actual equipment you are using I can only speculate. Perhaps post some detailed pictures of you controller and the associated electronics.
Tweakie.
-
I do not think so as this controller give me an accurate Z5 Z10 Z0 motion.
if you do Z5 Z10 Z0 the axis will not come onto Speed upwards, so if the Controller is not able to provide the
required current, it may be enought for doing the small step's but not for higher Speed.
give us some Pictures of your Controllers. pls.
-
Hi
My acceleration is 150mm's and velocity is 2800mm's
thanks
Velocity is too high for Z axis which unlike z or y has to carry a heavy spindle against gravity.
Although I use speed of 1800/s for x and y (anything about 2000 is less precise), I only use speed of 1000 for z.
Also because you are using a made-in-China controller which most likely uses the Toshiba TB6???? IC as motor driver you need to increase the step and dir PULSE-WIDTH to 10 and 10 or even 15 and 15.
The Toshiba TB6???? driver IC unlike Gecko driver (for example) are pulse-width triggered and not pulse edge-triggered.
Short pulse-width pulse are assumed to be noise and so the Toshiba ICs specifies a min pulse-width.
I bought my router from a mechanical Engineer engraving jewellery who was unhappy with the z axis. He even got a FREE complete ball-screw part under warranty from the manufacturer but was not able to fix the problem.
I just fixed the problem by increasing the pulse-width.