X axis zero point drifting to the right

Its set in the hal file
it needs to be fixed on all axis

PCW wrote:

Its set in the hal file
it needs to be fixed on all axis

Then I guess MAX_ERROR is referring to this?
setp pid.x.maxerror .0005
should be changed to
setp pid.x.maxerror .01

No, it should be set . 0.01

Now I have spent most of a day testing some modifications to my system that is losing steps in one direction on all axis. I travel 30 km from home to my shop so testing is done batchwise.

1. modified setp pid.x.maxerror .01 to be compatible with mm machine units. This had no appearent effect, but I will keep it as it seems logical.
2. Tested with ridicously low speed and acceleration, like 10 mm/min. No difference in amount of lost steps.
3. Bypassed the PID block by changing stepgen to pos mode and connecting axis.0.motor-pos-cmd to hm2_5i25.0.stepgen.00.position-cmd. This had no effect either better nor worse. So the problem is not in the PID block anyway.
4. Rewired all step and direction signals with twisted pairs with one separate ground wire each directly from the 6i25 to the Gecko drivers. Worried that this is an electrical disturbance problem after all. Did not help.
5. Inverted the step signal from 6i25 to G213 if it is some timing issue between step and dir. No better, no worse.
6. Changed microstepping setting in the G213 from 10 micro steps to full step mode. Still losing steps at the same rate.

My next move will be to put a scope on step, dir and actual step motor winding and running a program with a few steps each way and see what it gives.

Any other ideas to test are highly appreciated.

Other possibilities:

Not enough drive -- add a break out board or a 470 Ohm 5V pullup to the step/dir outputs

Missing or wrong common ground (not returned to 5I25 with step/dir wires so there's cross
talk between DIR and STEP

andypugh wrote:

I wonder if it is the dir-setup or dir-hold settings?

I tried to increase those values 10x, it had no effect. G213V spec is 200ns min for both.

Also tried to invert the step output to see if there was a difference at what level step is when dir changes. Did not seem so.

PCW wrote:

Other possibilities:

Not enough drive -- add a break out board or a 470 Ohm 5V pullup to the step/dir outputs

Missing or wrong common ground (not returned to 5I25 with step/dir wires so there's cross
talk between DIR and STEP

I already rewired all step and dir signals as twisted pairs with one ground each directly to the 6i25 to avoid a possible common ground interference.

It really seems like one transition but not the other of the DIR signal gives one extra STEP pulse. Tomorrow I will look with a scope and I will also try with some pullups. The G213V input spec is
"They will operate with 2.5V, 3.3V or 5V logic outputs with 2.5mA minimum source drive current."