# Unable to calibrate

I am trying to calibrate the machine so I have fixed a metal ruler to the waste board and with a 1/16 inch ball nose I start at zero and tell the machine to travel 20inches in Y, it travels 19 7/8 so I press the 20inches again to send it back to zero and it travels 1/8 inch past zero, so it travels 20 inches on the way back, I have been having trouble with the machine and think I am losing steps, here are my GRBL settings, any help much appreciated
\$0=10

\$1=255

\$2=0

\$3=5

\$4=0

\$5=1

\$6=0

\$10=0

\$11=0.020

\$12=0.002

\$13=1

\$20=0

\$21=0

\$22=1

\$23=3

\$24=25.000

\$25=750.000

\$26=200

\$27=1.000

\$30=12000

\$31=0

\$32=0

\$100=160.400

\$101=159.700

\$102=188.976

\$110=1100.000

\$111=1100.000

\$112=500.000

\$120=75.000

\$121=75.000

\$122=100.000

\$130=750.000

\$131=750.000

\$132=100.000

I first would make sure that the belts are tensioned, the wheels are tight, but not too tight, and that the pulleys are not loose on the motors. Then I’d put a straight bit in the router so you have a sharp edge to gauge the measurement. You need to keep doing the math until the movements are as accurate as you can get it. It’s a ratio of what you want to what you get. (20/19.875) x \$159.70 = 1.006 x 159.70= \$160.70 as the new number for the Y axis. These numbers could change after you tighten things up though…so it’s just an example.

Thank’s Martin
I will give that a try, I have 1605 ball-screw and linear rails, so I did check the couplings, I don’t understand where the \$159.70 comes from

But…

What controller?
You shouldn’t really need to calibrate with ballscrews.
Assuming 4x microsteps?

360°/1.8°=200 steps per rev.
200 * 4 (microstep) = 800 steps per revolution.
800/5mm per revolution = 160 steps/mm

That should be the same for any axis with the ballscrews. They don’t stretch like belts.

If you’re losing steps, it should be easy to notice if you watch the shaft coupler on the motor. If you lose about 3mm, that’s about 480 steps or about 2.5 revolutions. Watch that coupler and see if the motor stalls at any point.

What machine is this on?

Also, what steppers/drivers and PSU voltage do you run again?

Thank’s Neil,
I have the X Controller from Inventables, as the machine was an X Carve before I upgraded to ball-screw and linear rails, I will try again this morning, the reason I think it is losing steps is it does not return to work zero after a carve, it seems to be off by 1/8 inch 3mm, I will watch the coupler closley today to see if it stalls, I have also tried doing an air carve and get the same results

Haldor
I am using the 24v Xcontroller from Inventables, I had to abandon the controller I was working on due to me not having enough experience, I was trying to run before I could walk, I will get back to it when I have more experience

Not returning to same position is the symptom, why it looses steps will be the cause. A stalled stepper will/should make a quick/noticable grinding noise

I got it calibrated this morning and I then squared it using the 3-4-5 method and it went from the home position to work zero perfectly, I then measured from a fixed point on the gantry to the aluminium rail support at the front of the machine on both sides and the were the same, I then cut a shallow (0.02) 8 inch circle and when it returned to work zero it is off again by the same amount 1/8 inch, when I measure from the fixed points on the gantry again it has gone off square, it just wont stay square I have not noticed any grinding noise, and have not noticed the motors stall, the I got the 1605 ball-screws on EBay from china and wonder if they could be the problem, I have swapped the motors over but it made no difference

I have been doing some more testing and found that by reducing the acceleration it is getting better, not quite there yet but a lot better

Have you checked the alignment of your linear rails?
Do you have two Y motors?

Yes, I have two Y motors, I have not checked the alignment of the linear rails, I am not sure how to do that, do I need a dial calliper, when I put them in I made sure they were parallel, but I will check again, a few days ago when I started to home the machine the left-hand side did not move for some reason but the right-side did move and it travelled about 5-6 inches before noticed and hit the emergency button,

That suggest there are some sort of issue with the Y´s.
I´d disconnect the mounting pillow so each Y move freely and troubleshoot them individually.

1 Like

Thank’s Haldor, yes it is a problem with the Y axis only, I will try what you said, I have to leave it for a few days now but will let you know how it goes

I got back to testing again this morning and have found the gantry was not square to the Y axis, probably caused when one side did not move and the other did, I tried to square it as best I could, and did a carve, which did not go well , it was crooked and one stepper was making a grinding noise, I have ordered some aluminium to make new gantry plates and I am going to renew the whole gantry, I have ordered a new C beam, I suppose the stepper that is making the grinding noise is faulty, so I will have to order a new one, or should I replace both steppers

Steppers very rarely are at fault.
Grinding noise = stalled stepper due to insufficient torque to overcome the friction. A faulty/misconfigured stepper driver or improper GRBL parameters can also cause this.

This is with the Xcontroller correct?
Do Y1 and Y2 actually move the gantry the same way or is one mirrored relative to the other? Poor wiring can cause all sorts of troubles.

Haldor
Yes it is an Xcontroller, and I am pretty sure the wiring is correct, Y1 and Y2 move the gantry the same way, the stepper drivers are built in the controller and I have no way of knowing if one is faulty, I have been playing around with the GRBL settings to try and fix it, mostly acceleration, I have it stripped down now and I will rebuild the gantry and new Y plates, that should make it rigid and square again, I have ordered 2 new nema 23 motors as they are not too expensive, so after I do this rebuild and it is still not right it must be the controller drivers or a GRBL setting, is there a default GRBL setting

What are the current \$100-102, \$110-112 and \$120-122 values?

To trouble shoot the Y1/Y2 driver, swap the Y-steppers over - do the problem persist?

I have already tried swapping the stepper motors over and the problem was still there
\$0=10

\$1=255

\$2=0

\$3=5

\$4=0

\$5=1

\$6=0

\$10=0

\$11=0.020

\$12=0.002

\$13=1

\$20=0

\$21=0

\$22=1

\$23=3

\$24=25.000

\$25=750.000

\$26=200

\$27=1.000

\$30=12000

\$31=0

\$32=0

\$100=160.400

\$101=160.075

\$102=188.976

\$110=1000.000

\$111=1000.000

\$112=500.000

\$120=40.000

\$121=40.000

\$122=80.000

\$130=750.000

\$131=750.000

\$132=100.000