Not good at math OR scale is all wrong

Printable View

05-28-2013, 05:02 AM
Michaelofburien

Not good at math OR scale is all wrong

Ok so I'm not good at math. Built a fixed gantry 3 axis MDF machine. Got everything together and running. Noticed a funny thing while testing travel versus coordinates. Everything in gcode is about 1/10 actual size.

My fix was to change the pulse per increment in Mach 3 from (calculated) 2000 to 22000 on each axis. That seemed to dial it right in. But obviously I've done something wrong along the way. Is this just another newb error? If so, what might I do to correct it?
05-28-2013, 11:42 AM
robe_uk

Maybe check your controller micro step setting
05-28-2013, 02:40 PM
Michaelofburien

Quote:

Originally Posted by robe_uk

Maybe check your controller micro step setting

Right. That's what I thought. I have the DIP switches set at 2000. These are the Longs DM542A controllers.
05-28-2013, 02:54 PM
LeeWay

Okay. There is some math involved. If you are using Mach 3 and have a dial indicator, you can eliminate most of the math. There is a button on one of the pages that lets you calibrate each axis. This is needed even when you do the math. There are screw errors sometimes and the math just isn't quite accurate enough.
I have used this method to calibrate each axis on every machine I have with great results.
Just a gloat, I have 6 working right now. That is if I include the 3D printer.

Let us know if you need more help with it. These cnc machines are only as accurate as your calibration.
05-28-2013, 03:05 PM
robe_uk

Quote:

Originally Posted by Michaelofburien

Right. That's what I thought. I have the DIP switches set at 2000. These are the Longs DM542A controllers.

Ok, what type of drive? Screw or rack and pinion? A screw might be 10 turns per inch (roughly) which would account for the steps you are seeing, again please post more info of your setup.
05-28-2013, 04:47 PM
Michaelofburien

Quote:

Originally Posted by LeeWay

Okay. There is some math involved. If you are using Mach 3 and have a dial indicator, you can eliminate most of the math. There is a button on one of the pages that lets you calibrate each axis. This is needed even when you do the math. There are screw errors sometimes and the math just isn't quite accurate enough.
I have used this method to calibrate each axis on every machine I have with great results.
Just a gloat, I have 6 working right now. That is if I include the 3D printer.

Let us know if you need more help with it. These cnc machines are only as accurate as your calibration.

This was going to be my next step if I didn't hear a good answer here. So I'm kind of glad to hear my assumptions are pointing me in the right direction.

Quote:

Originally Posted by robe_uk

Ok, what type of drive? Screw or rack and pinion? A screw might be 10 turns per inch (roughly) which would account for the steps you are seeing, again please post more info of your setup.

10 TPI ACME screws all 3 axis. It is highly likely that I have the DIP switches set wrong.
05-29-2013, 07:45 AM
Walky

Some chinese drivers' dipswitches are counter-intuitive and use 1 as OFF and 0 as ON, maybe that's the problem and you have the switches inverted?
05-29-2013, 01:44 PM
CarveOne

Quote:

Originally Posted by LeeWay

Okay. There is some math involved. If you are using Mach 3 and have a dial indicator, you can eliminate most of the math. There is a button on one of the pages that lets you calibrate each axis. This is needed even when you do the math. There are screw errors sometimes and the math just isn't quite accurate enough.
I have used this method to calibrate each axis on every machine I have with great results.
Just a gloat, I have 6 working right now. That is if I include the 3D printer.

Let us know if you need more help with it. These cnc machines are only as accurate as your calibration.

Leeway is correct.

It's on the Alt-6 Tab, just above the RESET button. You can get very close using this utility, then edit the last few digits of the number manually until the actual cut measurements are as close as you need them to be. Record the x/y/z steps per unit numbers somewhere for future reference. This corrects for any tolerance build-ups in the parts, but don't rule out that there may also be some backlash involved that need to be corrected or accounted for.