[Arkwhizzard]

I have a small router that I would like to run on EMC2 all axis work well on the free Mach 3 down load. On EMC2 the X and Y axis work fine 1/2 step but the Z axis stalls and sometimes will start to move then stall then start in the oppisite direction even does this when moved manually (jog) I switched it to full step and play with the settings but I can't get a a reliable setting on the Z. It is a small router homemade with 65 oz Nema 17's 3 axis china controller and all axis move easily by hand (the Z only moves a flex tool holder). Any Ideas?

[vlmarshall]

It sounds like two of the windings on the Z stepper are reversed.

[Arkwhizzard]

I checked but no I don't think that is it. All Three motors are identical 4 wire bipolar and I checked the Ohms across all three befor using them. The wiring to the controler bus is correct. Every thing runs fine on Mach 3 but I would really like to get it running and use EMC2 thank you for your response. :idea:

[Big John T]

Did you use the Stepconf Wizard to set up your config? Did you test the Z axis in Stepconf Wizard? The Z will most likely not be able to accelerate as fast as the X and Y due to it is lifting the Z mass. If you just tested the X and then used the same settings for the Z that might be your problem.

John

[Arkwhizzard]

Thanks John for your reply. When I set it up I used the wizard and did each axis separately. That is where I noticed that I was going to have trouble with the Z. When clicking on the test button the motor might hum or move a little. Letting off the button and clicking it again the motor would start and move like the other motors but you didn't know which direction it would go. This afternoon I am planning on looking into the Toshiba TA8435HQ5 data sheets to see if I can improve on the setup. I hope that helps.

[Big John T]

The first thing I would do is to reduce the maximum acceleration for the Z then test it out.

John

[CaptainVee]

Check the table travel for the z axis using stepconf wizzard. It needs to be like -60 to 15.

I tried 15 to -60 and that froze the z axis.

Edit, placed this in the wrong tread.... Too enthousiastic I found a solution.

Have you run the latency test with the suggested "abuse"? I also had strange stuff going on with the z, just as you describe. I found that my latency numbers were too low and entered higher ones. This slows your machines speed but may get rid of those symptoms.

[dilbert0815]

The first thing I would do is to reduce the maximum acceleration for the Z then test it out.

I have a similar problem as a newbe with my Z-axis loosing steps.

However in stefconf the test wents well and all my axis have different max speed limits.
Now when I get to AXIS display I get the feeling that those speed limits in the hardware conf are ignored by AXIS. With being adventurous I can get all axes to start the buzz of death (stall) by manuall control and moving the speed slider to higher values that the axis can realy do.
The axis speeds I can manually do in AXIS in fact are faster that during the tests in stepconf.

I tested in stepconf larger travel for all axes too so each axis has reached the top speeds during testing. So I'm pretty confident that it's AXIS (or the G-code interpreter) ignoring the hardware limits set in stepconf.
How is this supposed to be configured propperly ?

Peter

EDIT
Did a little more try'n'error today.
It seems that AXIS honors the speed setting in the configuration somehow.
But actually only one of the axes is honored and it's not the slowest one.
It seems to use X or Y as reference in my case and ignores the lower speed limit of Z altogether ...

[Zig]

DO NOT HAVE DIFFERENT ACCELERATION SETTING ON VARIOUS AXES.

Set all three axes to same acceleration.
Set all three axes to same maximum speed.

[chester88]

Y do you say that?

[Zig]

Chester,

The idea behind CNC control is to have multiple axes moving in a coordinated fashion in a multi dimensional space.

If the axes have differing accelerations then the motion can not be coordinated.

[chester88]

During coordinated moves EMC only moves as fast as the slowest axis that is required to move,
but during rapid moves it will accelerator / cruise at each axis's maximum.
so having difference between axis is no problem and can speed up rapids.

I believe that stepconf may have not set the max velocity / acceleration settings properly.
As far as I can read in the manual in the INI file under the [TRAJ] heading
should have a MAX_VELOCITY and MAX_ACCELERATION entries.
This sets the coordinated move values I was talking about.
I am emailing the developers to ask about that as the manual may be out of date a bit.

Chris M

[Zig]

Chester,

I seem to recall reading in one of the manuals that the acceleration ought to be set to be same in all axes.


I mayhave to go back to the docos and check for chapter and verse.

[dilbert0815]

I seem to recall reading in one of the manuals that the acceleration ought to be set to be same in all axes.

That's a workaround.
However it would be sensible if the system would pick the slowest axis when it's required to use te same speed for coordinated move.
The speeds in the ini files seems to be ok. All axes have their dedicated speeds and accel. values filled in.

Peter

[Zig]

Dilbert,


So you too have seen it?

What is it a workaround for?

[chester88]

Peter try adding:

MAX_VELOCITY= 5.0
MAX_ACCELERATION= 20.0

under the heading [TRAJ] in your INI file.
change the 5 and 20 the velocity and acc. from your lowest performance axis.

See if that helps.

Chris M

[chester88]

Hey did you try it?

[dilbert0815]

I didn't modify the ini file by hand but used stepconf again to limit X/Y. So I don't get excessive speeds now in Axis.
You can also simply limit your fingers and just jog Z rather slowly.

However this does not solve this realy.
It's a crippling fix only.
Why limiting all axes when only one needs to be limited is beyound my understanding.
Afaik most CAM tools set Z to lower speed defaults as X/Y.
The typical 2.5D router needs high X/Y speeds but only modest Z value.

I think this SHOULD be fixed as it realy misleading and a serious limitation of an otherwise nice CAM tool.

Peter