Machsupport Forum
Mach Discussion => Mach3 under Vista => Topic started by: grsabourin on August 04, 2019, 11:20:41 PM
-
Hi
we just installed brand new servo motor kit on a old router we would like to renew. All axis are turning freely by hand all along the axis displacement. The motor is directly driving a screw by the way of a servor motor coupling. These coupling are tight.
We have configured alll Mach3 parameters to reach good accuracy for all axis. The issue we have currently is drifting of the origin along the program. It is like the axis are loosing some steps. Example. We have a reference for the zero position on the table of the machine. We are doing the zero of the machine at this position before starting the program. We run a small program doing a facing operation. (100 g-code lines). We run this program in the air (without milling material). Returning the machine at origin position (G00 X0Y0) and we have an offset of the origin around 3 mm in x and y, It is mostly the same in both both a little bit larger in X. The X and Y screws are the same.
Is someone can have any idea what can cause this drifting of the origin?
Thank you for your help.
-
Hi,
you are gaining/losing steps. Maybe noise is getting to your servo encoders and making the drive think the servo
is at a different position.
Tells us more about the servo and servo drive.
Craig
-
Thank you for the answer. I'm not sure it is noise. My drive and all cables are bran new. It is a upgrade kit sold by DMM Motion. (http://www.dmm-tech.com/multiaxissolutions2.html). On this kit I just added and ESS board (https://warp9td.com/index.php/products) to be able to drive the machine from RJ-45 port instead of parallel port. The addition of this board never change the behavior of the system compare to initial situation directly driven by the parallel port.
I did some test with basic code with the machine. Simple code like
G21
G00 X0
G00 X200
G00 X0
Repeating this sequence many time to see if the zero location is drifting. After few repetition it is possible to see the origin drifting in the direction of x positive. It is really reproducible. I have same behavior if I do the same experimentation along Y.
The next step will be to try to jog the axis directly on the servo drive to detect if the issue is coming from the mechanic. I will be really surprised but I need to eliminate this possibility. After that I will need a way to investigate the controller. Is someone has any suggestion to test that?
Thank you
-
Hi,
you are gaining or losing steps. If it were the servo/servo drive it would fault 'following' error in short order.
Ergo the extra/missing steps ocurr before the servo drive.
The ESS is likely to be good. I assume that you have a breakout board between the ESS and the servo drive.
What make and model? Is it fast enough for the signals it is expected to carry?
What is your motor tuning (steps/unit and max velocity)?
What are the electronic gearing parameters in the servo drive?
All these questions determine the required signalling pulse speed to the servo drive. Personal experience suggests that
any thing faster than 50-100 kHz is going to require differential signalling.
Craig
-
Hi
Thank you for your reply
This is some information
1 - My breakout board : http://www.dmm-tech.com/DMB4250-8B_main.html
2 - Is it fast enough : Good question
3 : Motor tuning : Step per : 10007.2, Velocity : 115.56, Acceleration : 16.720908, Step Pulse 5, Dir Pulse 5
4 - I did some test with signal pulse speed at 100Khz with the same result
-
I tried to change motor parameter and something interesting happen. I changed the velocity from 115 to 215 and the drifting is reducing (around 0.2 mm). I pushed on the other direction from 215 to 15. The machine is moving very slow but for the same number of line I obtain 13.6 mm drifting that is huge