Wrong measurements after plasma cutting

[wiltshire_chris]

I've only cut a few small things in the table since building it, bottle openers simple flower, the boys homework project you get the idea, last week end I set about cutting multiple parts the same, the total width was about 400mm but when doing a dry run the torch kept running off the sheet. Much head scratching and measuring was showing the if I move a axis over a set distance it measured different to what the dro said, the only way I found to ajust and compensate is adjusting the pitch of the belt in stepconf. I'm using gt2 belt and as of last night I have the x axis at 1.82mm and the y at 1.9825. To try and get it as close as I can over 600mm and 1000mm on each axis. The only thing that has changed between last week end and a few week before is the desktop shat itself so I got a bother computer pretty much the same and swapped the hard drive over. Any ideas?

[Rodw]

Chris, we need more info. The accuracy is set by the steps per device unit (mm or inch) in the ini file so the distance setting in mm make no sense to me. Its best to calculate this rather to measure it.

GT2 belt has a pitch of 2mm. How many teeth on your drive pulley?
What reduction do you have?
What microstepping do you have?

Example:
20 teeth on pulley so distance travelled per rev = 20 x 2 = 40mm
3:1 reduction on motor (eg might be a 60:20 belt reduction)
10 x microstepping
so it will take 3 revs of the motor to turn the drive pulley 1 rev
3 motor revs = 40mm
1 motor rev = 40/3 = 13.3333 mm
200 steps per rev * 10 x microstepping = 2000 steps
so 2000 steps per 13.333mm = 2000/13.3333 =
150 steps per mm
so STEP_SCALE = 150 in ini file

[wiltshire_chris]

Gt2 belt, 20 tooth drive gear, direct drive, 1/8 micro stepping. Nema 23 180 steps per rev.

[wiltshire_chris]

I think I just answered my own problem, 180 steps per rev!

[acourtjester]

the driver sets the steps per rev, the calculations for distance is need after the axis is assembled. The easy way to do that is a formula that uses the existing steps per distance, how far you tell the axis to move and how far it did move. This is shown in the attached image. You do the calculations and then enter the number in the motor tune and do it again using a longer travel to fine tune the steps per number, do this about 3 times.

steps per calulations.JPG (16.05 KiB) Viewed 13735 times

steps per calulations.JPG (16.05 KiB) Viewed 13735 times

[wiltshire_chris]

Something is wrong with the software then, I've done as you said, multiple times and each time the distance run changes sometimes longer sometimes shorter, it's random. I'm ready to get the 9 inch grinder to it

[Rodw]

Gt2 belt, 20 tooth drive gear, direct drive, 1/8 micro stepping. Nema 23 180 steps per rev.

200 base steps/rev x 8 microstepping steps = 1600 steps/rev
20 teeth gear x 2m = 40mm per rev
Metric: 1600/40 = 40 steps per mm
Inch: 40 x 25.4 steps/inch = 1016 steps/in

Edit your ini file directly
If your dimensions are changing its either your gcode or you are missing steps. (bad step timings in ini, accelleration and/or velocity to fast)
Share you hal and ini files

[acourtjester]

Do a simple test move the axis for a distance in both directions and measure each time to see if it move the distance you tell it. Enter in G1 X 250 F50 If you are using MM or G1 X10 F50 if you are using inch. Remember to put the -X10 or -X250 when moving the other direction. Use Y if it that axis. The F50 will give you a controlled speed when moving. If not you have slippage in that axis drive assembly.

[wiltshire_chris]

I love and hate this thing, so it turns out that the jog speed was to fast, I guess it was missing steps as it was moving.

[acourtjester]

I use the keyboard keys to quickly move a axis back and forth to test the jog speed. If the motor stalls reduce the velocity until it does not.

[adbuch]

I use the keyboard keys to quickly move a axis back and forth to test the jog speed. If the motor stalls reduce the velocity until it does not.

Tom - I do the same thing with my Mach3 and Centroid controlled cnc wood routers. It makes it particularly handy to have a wireless keyboard for this so I can walk around the machine with the keyboard. I'm using a Logitech MK530 on several machines and they work great.

David

[acourtjester]

Ditto David they are very handy