[Balbz]

This is more for the experts of the group. I designed/built a 5x10 gantry that runs on gear rack with steppers motors, running off a mx4660 4 axis driver and a UC400eth pulsing engine. My issue is that I am gaining steps and having erratic movement because of this. Right now if I move to 6" its exactly 6" 10" is exactly 10" however thats as long as I can measure with my micrometer. When I go out to 36" using a tape measure its off by roughly .25" I've used both UCcnc and mach 3 with somewhat similar results. This only occurs in the y axis from what I can tell, I have been focusing on that axis of movement. Its a slaved axis with individual drivers and wiring. I have even replaced the gear rack and drive gears on both sides and have similar if not identical results ( I am not super accurate with the tape measure). I have squared up the gantry sides with a angle gauge and some 1 2 3 blocks. Ive also noticed that when I cut this specific bracket the outlet and nearly all the bolt holes line up but one. This one bolt hole is cut halfway through the program and the machine returns back to zero exactly so I know I am not losing steps or anything during the cut. Any ideas? I am going to swap to UC100 USB and a different (TB6600HG 4 axis board) to test motion here tomorrow just to confirm, might even try and swap out the laptop. I am not sure what else I can try, just doesn't make sense. Never had an issue like this on any of the other machines I've built however this is the biggest on thus far. Speed is set to 500IPM and 60 accel. I've run the machine at over 2000IPM and 150 accel with smooth motion

[robertspark]

What gear rack is it (details / specification)

Is it direct drive or geared ( what ratio)

What micro setting are your drives?

How many steps per inch / mm / unit have you set it up to

[OlfCNC]

Your steps per unit value is likely to be wrong. Make a movement and measure the distance. Do the same with different distances movements and write down the measurements.

[ger21]

Make sure you don't have a loose pinion or set screw somewhere.

Put a V bit in the spindle, and use the point of the V over your tape measure, and calibrate the axis over it's full length, rather than just 6"

[Balbz]

Used mcmaster Carr gear rack first them replaced it with some from CNC router parts IIRC.

I calibrated using a dial gauge originally that was max 1" After that is when I replaced the gear rack. From there I build an end stop for squaring up the gantry wuickly and from there I was able to use my 12" micrometer to set the steps per unit over 12" of movement using the auto calibrate feature of both programs.

everything is very tight, no discerenable backlash or loose gears. Cant put a bit in it as its a waterjet, going to mess with it tonight or tomorrow with a tape measure over a longer distance to see what happens.

Motiontek | Rack Pinion Spear Gear drives CNC Plasma Router Canada USA

Correction that is the rail I am using, part number RSHC102000 with a 16 tooth pulley.

can rapid from end to end and it will still end up in the same spot.

[awerby]

What happens if you move repeatedly back and forth - does it return to zero afterwards? If it's returning to zero correctly, then you're right - you aren't losing (or gaining) steps. If one hole out of many is being cut inaccurately, check the model and g-code.

[Balbz]

returns to origin perfectly everytime

[he1957]

Incremental versus absolute move introducing some error?