Z axis losing steps

[nlancaster]

Hello, I have recently tried a very long 3d job on my joes cnc 2006. I have not had nay problems with z-axis losing steps since I got the machine tuned. But I have also never done a 3d toolpath before. My machine is losing steps randomly only on the z-axis. I seem to have read somewhere that mach3 has issues with acceleration rates being different between axis, while it is doing a 3d move. Is this correct and what should I change.

I right now my settings for motor tuning are.

x-axis 180ipm acel 20
y-axis 180ipm acel 20
z-axis 70ipm acel 2

pulse length 15 15

This has worked quite well for all the work I have done on the machine so far.

I checked all the setscrews on the z-axis couplers and they are tight. reseated the set screw in my z-axis rattler-damper. and that helped alot. but it will gradually lose steps in either direciton, sometimes it keeps droping other times it climbs. And it appears to always been a small change that just builds over time. And then will change direction and start losing the otherway.

Also just double checked and the coupler for z-axis set screws are on the stepper flat and the lead screw flat.

Similar Threads:

• Z axis losing steps
• Losing steps in the Z axis!
• Problem- 770 losing steps on the z axis
• Need Help!- Losing steps with the Y axis
• Y Axis keeps losing Steps

[nlancaster]

So with further testing, I am back to losing steps on z-axis like I was once in the past. Messed with pulse length, sherline mode, acceleration, ipm and kernal speeds. Nothing seems to help. This loss of steps only appears to happen on the z-axis, x-y do not appear to be showing the same behavior.

I have been testing the z-axis with 1inch long moves and .1 inch moves. They all appear to have the problem. no matter what speed I run them at.

[CarveOne]

You do not give enough specifics about your machine for anyone to make specific suggestions. The trouble you mention can be related to mechanical binding of backlash, electrical issues related to drive electronics and motors, cross coupling of signals due to the wiring type or layout, incorrect settings in Mach3, and/or the computer itself. Getting a new machine running with complete reliability can sometimes be a real challenge.

The first thing I would suggest trying is to reassign the problem axis to a working axis and see if the errors follow the reassigned axis or stay with the original problem axis. This will confirm that the problem is related to just the Z axis assembly - or not. Start with swapping the motor wiring at the motor driver output terminals and interchange the Z axis motor wires with whatever you named the gantry axis. Make one change at a time and record the results as 'errors' or 'no errors' in a simple chart format.

Is your motor wing shielded cables or individual wires? Are the cables/wires tightly bundled together? Are all motors the same, or is the Z axis motor rated at lower torque? What is the voltage feeding your driver(s)? Is it a Gecko G540 or a Chinese driver? There can be a hundred other questions to come if the solution isn't obvious in a short time.

I had a Mach3 setting problem on my first build that that used slaving of two motors and it drove me nuts for three months. In my case it was causing racking of the gantry.

[nlancaster]

Sorry here is the more information,

Keling KL4030 drivers
Keling 425oz/in steppers all axis
C10 BOB
Shielded wiring on steppers
36v 3amp per stepper
36v 10amp PSU
z-axis single start acme with dumpster nut
x y axis 5 start acme with dumpster nuts
Windows XP, mach3

It seemed to be working fine, no idea when the issue started.

Problem happens no matter what ipm or accel I run at.

I will try moving the motor control cable, and I was also going to try linux cnc see if its a windows/mach3 issue.

[awerby]

It sounds like a mismatch between the motor and the axis. In a 3D toolpath, the Z axis often has to do a lot of little moves very quickly, but that motor may have too much inductance for that to happen without mishaps. A motor with less holding torque but less inductance would probably have an easier time keeping up.

Also, all your acceleration numbers seem on the low side for Z; your machine might have more pep if you boosted them a bit. I've had good luck with acceleration figures about 1/10 those of the velocity numbers, so X and Y seem about right, but Z is lagging.

[nlancaster]

aweby, I have heard that about these motors. purchased them before I knew exactly what I was getting into.

As far as the accel I have tried itas high as 10 and had the same issues.

And I may have to replace the z-axis motor, but which one should I get. the disadvantage is my coupler is setup for 3/8 shaft on the stepper and 1/4 shaft on the motor, so I would have to replace that as well.

[ger21]

Have you tried a damper on the motor?
Get a hockey puck, drill a hole in the center, and press it onto the rear shaft.
Cheap and easy way to see if a damper will help.

[nlancaster]

I have a rattler damper on it now gerry.

[ger21]

Try taking it off?

[nlancaster]

I have not tried without the damper, I will tonight.

[nlancaster]

Exact same issues with out the damper gerry.

I was able to get linuxcnc at least speaking to the drivers and moving the motors. but was getting too cold to do anything out in the garage. I will try again tomarrow, or wednesday.

[Devastator]

Hello, I have recently tried a very long 3d job on my joes cnc 2006. I have not had nay problems with z-axis losing steps since I got the machine tuned. But I have also never done a 3d toolpath before. My machine is losing steps randomly only on the z-axis.

Hey, I had the same exact scenario. I ran mine for a year, no problems until running a long 3D file. In both directions, sometimes high, sometimes low. Completely random. And I'm talking just a few steps either, several inches. Drove me nuts. I added a damper, slowed down accel, changed microstep settings, triple checked everything I could think of with no luck.....You know what fixed it? Opening up all of my cable chains and organizing all of the wire runs so that none of them crossed. Not one single problem since.

Let us know if that helps you.

[nlancaster]

NEMA23 282oz/in 3A Stepper Motor ¼? Dual shaft (KL23H276-30-8B) | Automation Technology Inc

If I have to replace the z-stepper, would that be a good replacement?

[ger21]

I'd get this one: NEMA23 270oz/in 2.8A 1/4? Dual Shaft Stepper Motor (KL23H276-28-4B) | Automation Technology Inc

The one you linked to is 4.2 amps when wired bipolar parallel. You don't want to wire it bipolar series, as the inductance is too high.

[nlancaster]

thanks gerry, going to try out a few things see if it does any good this weekend. then if still having the issue try the different stepper.

[daniellyall]

you may have backlash, you may have a lose bolt or nut some where on the Z, you may have a computer that needs Mach3 started twice to get it to work properly, or wrong size stepper.
disable all axis except z run you file if the start and end position are not the same something is wrong with it then.
check for one thing at a time starting with all nuts and bolts, if you have still got problems do a double start on Mach3 if its still there check all wiring making sure that it done correctly i.e.. shielding one end only that sort of thing. also it could be noise doing it make sure earthing and wiring is in good condition.
once all that's run through if its still out then it could be you stepper being wrong size or your acme screw and dumpster nuts are get worn out.
I had the same problem I farted around for 2 years trying to get it to work properly I read some where on this forum or mach3 forum cant rember what one to do it this way and I found what the problem was within hours. all it needed was double start on mach3 every time i use it and new correctly labled steppers

[nlancaster]

Never could get linuxcnc to run reliably. It probably has issues with the hardware I am running. But I seem to have fixed the issue.

1. Reinstalled windows and mach3 from scratch
2. Reseated all pins in all connectors.
3. Reseated grounding wires.

Seems to have solved it. 2 Test ran 500 repatitions of -.5 to +1 Z axis and it came back to zero. Tomarrow I will try the 3d carving again.

[nlancaster]

Yay!

After 2 hours of machining the first 25% of the machining is done. I will get some pictures up after dinner.