I believe I have done that as well. I’ve gone down to the point just above where it won’t move and close to the max to see if that caused a change.
I’m stumped
Hmmm… ok, spitballing crazy ideas. Does your g code contain any instructions to change the steps/mm in the middle of it? Or are you just jogging it? It could also be faulty electronics, unfortunately. You’re swapped the motors out with no change, so that’s not it. Mechanical wouldn’t make sense because you’re gaining steps from after 100 mm.
Please list the exact starting and ending positions for each movement. Was it 0 - 100, 100 - 200, 200 - 300, etc.? Or did you try 0 - 100, 0 - 200, 0 -300, etc.?
It’s when I’m jogging the machine. It happen on 100mm increments and is also off if I tell it to go from 0-400 or 0-600. But always returns to zero
This afternoon I will write down where it travels on each command.
1 Like
Ok here are the numbers I found
calibrated on 0-200. spot on.
0-100= 99.5
100-200=200
200-300=300.5
300-400=401
400-500=501.5
Then from 500 back down to 0
500-400=402
400-300=301.5
300-200=201
200-100=100.5
100-0= 0
checking larger movements
0-300=300.5
0-400= 401
0-500=501.5
I also calibrated it to 500 mm and the result was about the same. it was off on all the hash marks except 0 and 500.
during this calibration I went in sets of 10 and it was under consecutively until 500.
Just so I understand. You moved the wiring at the gShield from one driver to another and the problem stayed with the X axis. Correct?
If so, try swapping the actual X motor with a Y motor and see if the X axis still fails.
So just to make sure I understand. This is on the X axis movement? No mater if you are using the X or Y axis controller input to drive it?
If that is so it sound mechanical in the X axis. Weird.
It almost sounds like a backlash issue… In screw drives backlash is caused by the gap between the thread and the nut when the screw reverses direction.
Double check the set screws on the pulley. Make sure it is tight against the flat spot and (no wobble).
Yes. That is correct Larry. With 2 new motors Same exact problem.
It may also be a manufacturing tolerance issue in the belt itself.
@StephenMonroe are you saying you have physically swapped out the X axis motor with a Y axis? Like, mechanically?
I’ll double check it and check tension on belt again. Would backlash be consistent in its deviation? And allow to a return to 0 every time?
It might, but it seems very odd to me that you get different results from 0 - 100 than 100 - 200. That’s the same distance, and the same number of steps. If backlash were the issue I’d expect it to be identical for those two movements.
No just swapped them in the gshield. The wiring. I have gone through new motors on the X-axis to try and figure out the problem. Inventables has been great in this. they shipped me 3 new motors to test out.
Also the problem was the same with the 2gt2 belt and the 3gt2 belt I currently have on it
I agree. I’m grabbing at air to figure this out and appreciate all yalls help.
If it was consistently under or over I could figure this out. But it’s under then over.
Brain Storm Check
Ok so you have driven the X axis with the Y output from the shield.
- So it isn’t a controller issue
You have driven the X axis with a different stepper motor.
- so it isn’t the motor
- and in changing the motor out the odds of a loose set screw is out.
So that leaves the pulley or the belt?
- Maybe one of the teeth on the pulley is off somehow?
- Or the teeth on the belt?
Maybe something is wiggling, bent V wheel screw or loose… something?
Ideas to try:
Try swapping pulleys?
Maybe work out how many mm = 1 revolution of the motor and try a few distances that are increments of that. (100 turns, 200 turns…)
See if there is a pattern?
Try a few distanced that are a fraction of a revolution (100.0 turns, 200.5 turns, …)
See if that changes the pattern?
Try different directions, see if that affect the pattern
Setting the home in the middle of the axis
See if 0 to 100 to 200 to 300 to 400 has the same pattern of error as
0 to -100 to -200 to -300 to -400
Does 0 to 100 to -100 to 0 return to the same spot?
0-200 =200
200-0=0
0- -200 =200.5
-200-0=-.5
Continued to 200 and it landed spot on
Back to 0 and that was spot on
Get a new ruler 
3 Likes
Haha!i needed a laugh. I should put down some aluminum and have it carve it out. Then it should be perfect.
If the motors work right (new motors), the gShield seems to work (you swapped x and y drivers), you’ve beat the mechanicals to death, could it be the Arduino?
Which version of grbl are you running and have you re-flashed the Arduino?
I haven’t reflashed it. Not sure what version. I bought it this year in February.
But would it still be the arduino if it shows the same behavior on X axis regardless of being plugged into X or y control on gshield?
Y axis is spot on whichever it’s plugged into.
I will grab another bit of cable and try that. That’s the only thing I haven’t tried.
Also Paul with inventables is sending a new g shield! Great company
Not likely, but we’ve hit just about everything. The new gShield is a good idea, I’ve seen position errors before with a bad driver chip, maybe yours is marginal.