Trouble adding an A axis

More
25 Sep 2013 05:24 #39170 by andypugh
Replied by andypugh on topic Trouble adding an A axis

But this problem is not a hardware issue. Since there is no feedback in
the system other than homing switches and the spindle encoder Linuxcnc
is unaware of what the X Y Z & A axis are actually doing.


There are two ways that a stepper can generate a following error.
1) If the step generator can't actually make enough steps in the allocated time. This can often be a problem with software stepping, when there simply aren't enough base periods to make the steps, but you could also see the same thing with a Mesa card if the step space and step length add up to more time than the requested frequency.

2) If a single step pulse is actually bigger than the following error number.

I suspect that the second one was the problem. .0005 degrees means 720,000 f-errors-per-step. For 4x microstepping, if there is less than a 900:1 gear ratio between the motor and the axis, then you are likely to see the problem you saw.

Please Log in or Create an account to join the conversation.

More
29 Sep 2013 05:27 #39336 by garymcrobertpdx
Yes I agree that explanation #2 is the probable answer How ever I do not have a good understanding
of this.

The current conditions are compumtotor S6 drive configured with 20,000 micro steps per rev (the motor
I have works best with this) connected to a Mesa 5i25 board with a Step pulse 1us and step space 1us.
The gear ratio motor 20 tooth spin indexer 72 tooth = ratio 3.6

The INI file max speed for the A axis is 360 degrees / second. This should equate to a step rate of 72,000
pulses / sec.

Please Log in or Create an account to join the conversation.

More
29 Sep 2013 06:15 - 29 Sep 2013 23:32 #39337 by PCW
Replied by PCW on topic Trouble adding an A axis
What I think is happening is you are asking for more precision that the step generator/LinuxCNC system is capable of:

Your MIN_FERROR of .0005 is 1/720000 of a rotation or 1/36 of a uStep.
Your linuxcnc system does not have low enough jitter to keep the hardware position
within 1/36 of a uStep at low speeds. This is not at all important as .018 (one uStep)
is the smallest ferror that makes any sense to use (at 20000 uSteps per turn)

Depending on your jitter specs, I would think you could do better than .1 degree
(1/3600 of a turn) at the high speeds where FERROR applies however.
Last edit: 29 Sep 2013 23:32 by PCW. Reason: sp

Please Log in or Create an account to join the conversation.

Time to create page: 0.074 seconds
Powered by Kunena Forum