Need Help!- Stepper gains steps

[CNCDevil]

Hi all.

A couple of days ago when tried to cut a relief in Artcam for the first time, I experienced some problems with the Z-axis. It's not returning to zero.

I pulled the stepper off the Z-axis assembly to do some tests. I wrote some code to make the stepper shift direction around 50 times and noticed that the coupling on the stepper had increased its position in the turning direction with 1/8 of a full turn.

G21
G00 Z0
G00 Z1
...and the code continous like 50 times...
G00 Z0
M30

I can conclude that the stepper is gaining som extra steps everytime it shifts direction. I have tried to change a lot of settings in Mach3 to locate the problem, but nothing helps (Velocity, acceleration, kernel speed...)

Maybe it's a hardware issue, but please guide me in the right direction to whatever the problem could be.


WinXP SP3
Mach 3, 45,000 Hz
Centent CN0162 drive, 10 microsteps
Hobby CNC 425oz/in stepper, bi-polar, 400 steps/rev.


Best regards,

Brian

Happy New Year

[awerby]

It could be a mechanical problem, like your coupler slipping. If it's the type that depends on a set-screw to hold it in position, that's the first thing I'd suspect (the clamping type is more reliable).

It could be a computer problem - are you trying to run Mach3 off a laptop? (You don't say much about your computer, but that won't work). If nothing else seems to work, you can try optimizing your desktop computer using the instructions on the Machsupport site.

It could be a bad driver, a bad motor, or a faulty cable. Try switching leads around (but make sure everything has shut down before unplugging anything, allowing time for the capacitors to drain), so you are trying the same thing with a different motor, or a different driver. (By the way, I doubt that your motor really takes 400 steps to revolve 360 degrees - that's pretty odd, since just about all of them, including all the ones on the Hobby CNC site, take 200 steps/rev.)

It can also be a motor-tuning issue. You say you've tried everything, but start again and try it some more. Make sure you've got the right steps/unit setting entered: (Screw pitch times motor steps/rev times microstepping factor) Drop back to 25k hz, go for a fairly low velocity like 20 ipm, and a low accelleration like 2. Press the up/down arrow keys and listen to the motor. If it sounds healthy and happy, try a little more speed. Notice when it starts to growl or hesitate, then back off 25% or so and try again. You want to find a sweet spot, where it goes fast, sounds zippy, and doesn't choke. When you get there, derate it about 10% as a safety margin, "Save axis settings" and try your Z-axis test again. It's best to tune your axis when it's assembled, not just on the bench, since motors act differently when they're loaded (don't we all?)...

Let us know if you get this licked, and if so, what the problem turned out to be, okay?

Good Luck;

Andrew Werby
ComputerSculpture.com — Home Page for Discount Hardware & Software

Originally Posted by CNCDevil Hi all. A couple of days ago when tried to cut a relief in Artcam for the first time, I experienced some problems with the Z-axis. It's not returning to zero. I pulled the stepper off the Z-axis assembly to do some tests. I wrote some code to make the stepper shift direction around 50 times and noticed that the coupling on the stepper had increased its position in the turning direction with 1/8 of a full turn. G21 G00 Z0 G00 Z1 ...and the code continous like 50 times... G00 Z0 M30 I can conclude that the stepper is gaining som extra steps everytime it shifts direction. I have tried to change a lot of settings in Mach3 to locate the problem, but nothing helps (Velocity, acceleration, kernel speed...) Maybe it's a hardware issue, but please guide me in the right direction to whatever the problem could be. WinXP SP3 Mach 3, 45,000 Hz Centent CN0162 drive, 10 microsteps Hobby CNC 425oz/in stepper, bi-polar, 400 steps/rev. Best regards, Brian Happy New Year

[ger21]

Cut your accel in half and try it.

[M250cnc]

Originally Posted by CNCDevil I pulled the stepper off the Z-axis assembly to do some tests. I wrote some code to make the stepper shift direction around 50 times and noticed that the coupling on the stepper had increased its position in the turning direction with 1/8 of a full turn. G21 G00 Z0 G00 Z1 ...and the code continous like 50 times... G00 Z0 M30 I can conclude that the stepper is gaining som extra steps everytime it shifts direction. I have tried to change a lot of settings in Mach3 to locate the problem, but nothing helps (Velocity, acceleration, kernel speed...) Best regards, Brian Happy New Year

Brian under ports & pins motors have you set the axis to be reversed ?

If so reverse the axis under homing/limits & try running your test code again

Phil

Sent from my HTC Desire using Tapatalk

[CNCDevil]

Currently I'm not at the place where my router is, but I will try to answer all your questions from what I have tried to do so far.

I have made sure the coupler is not slipping. If it was slipping, the z-axis should lower itself each time, but it actually rises. That's the way I realized there was a problem. I made the test with the stepper not mounted to the Z-axis.

I'm using an old laptop and I have used it before with a HobbyCNC board and there was no problem at all, but I will try to hook up a desktop to see if it helps.

I will try with another motor and driver. It is 400 steps/rev.
Stepper is 200 steps. 10 microsteps drive. 5 mm pitch/rev. => 200*10/5 = 400 steps/unit
Even if it wasn't correct, I would say that the drive shaft should return to where it was before I ran the Gcode (starting at Z0, ending at Z0), but the shafts position is shifting a fairly constant angle in the same direction everytime I run the code again.

I haven't tried to set it back to 25 kHz, but I have tried to lower the velocity and acceration.
V = 5000mm/min to 1000mm/min
A = 500mm/s/s to 50mm/s/s

Unfortunately it doesn't solve the problem. The motor sounds fine.

I will check out the Ports & Pins configuration and homing/limits.


I have read some other threads about noise causing this to happen, but all my cables are shielded. My power supply (20A) is located just above the drive, so maybe that's the problem.


Thanks for all the answers so far.


Brian

[Torchhead]

1. Noise tends to be random in nature so will not show up as consistent position shifts.
2. If it only occurs during DIR changes and appears to be the same amount it may have to do with the pulse timing. The Step and Dir have to fall at the right times. We had a problem creep up like this in one of our designs. It was because we were not turning the DIR signals ON/OFF fast enough (opto had too much 'storage' time. It was resolved by adding pullups on the opto transistor sides pulled up to +5 point in reference to the drive forcing the transistor to snap on/off faster. The main symptom was you could see the progression of the position shift by a small amount each direction change. To test we took the motors off, marked the coupler and ran a series of hundreds of back and forth tests. It was predictable in that after 100 moves it was off the same amount each time. That showed it was not noise.

[CNCDevil]

Originally Posted by Torchhead 1. Noise tends to be random in nature so will not show up as consistent position shifts. 2. If it only occurs during DIR changes and appears to be the same amount it may have to do with the pulse timing. The Step and Dir have to fall at the right times. We had a problem creep up like this in one of our designs. It was because we were not turning the DIR signals ON/OFF fast enough (opto had too much 'storage' time. It was resolved by adding pullups on the opto transistor sides pulled up to +5 point in reference to the drive forcing the transistor to snap on/off faster. The main symptom was you could see the progression of the position shift by a small amount each direction change. To test we took the motors off, marked the coupler and ran a series of hundreds of back and forth tests. It was predictable in that after 100 moves it was off the same amount each time. That showed it was not noise.

It sounds much like my problem. I'm not sure that I understand what you are saying. I'm using a BOB with optocouplers. What should I do?


Thanks,

Brian

[M250cnc]

Originally Posted by CNCDevil It sounds much like my problem. I'm not sure that I understand what you are saying. I'm using a BOB with optocouplers. What should I do? Use a bob without optos if your drivers also have optos Phil Thanks, Brian

Sent from my HTC Desire using Tapatalk