Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: foamlvr on August 31, 2009, 09:56:14 AM
-
I am very new to cnc, and have recently built a Phlatprinter MKII from phlatboyz.com I am using a Mechatronics bi polar 3 axis micro step motor driver, with 3 nema-23 stepper motors. The machine is all belt driven. The x axis is a large belt driven roller similar to that of a large format printer. Y and Z are on a gantry below the cutting surface, and the cutting bit protrudes from below. When I run a cut file X and Y do not finish where they start. Y drifts progressively farther and farther in the + direction, and Z does so as well. To help diagnose the problem I rewired coil 1 to coil 2, and coil 2 to coil 1 reversing the steppers, then in mach I reversed settings so everything ran in the right directions again. After reversing the steppers I still have the drift but in the opposite directions. The drift is very consistant. The same cut file ran over and over leaves the cutter the same distance off. I have tried lowering the feed rates and acceleration, with no changes in the drift.
Is it possible that one coil is working better than the other in the steppers? If I could set steps per unit in the + direction, and then in the - direction independantly couldn't I compensate for this drift?
Jeff
-
Sometime, board can miss step because of the optocoupler that are not fast enough. In motor tuning, try to increase the step and direction duration, if you have one in both , try with 5.
-
What breakout board are you using?
Regards,
Ray L.
-
Ray L. I am using a mechatronics bi polar 3 axis micro step motor driver, from www.easy-cnc.com ,
ostie01 I have already tried increasing the step and direction pulse
-
It could be mechanical? Have a VERY close look at your belts/pulleys when you are running your program. Are you slipping a tooth on your pulleys? Sometimes this will happen if your belts are not tight.
-
FAOMLVR,
Are you steps per units correct?
RICH
-
sshneider belts and pulleys are tight, been through this thing with a fine tooth comb, also checked set screws, all stepper shafts have flats ground for the set screws.
RICH Steps per unit are as close as I can get them, with a 6" digital vernier calipers. Regardless of whether they are right or not shouldn't zero be zero?
-
Have you tried changing the Active state of the Dir signal?
Hood
-
Hood If you mean "dir lowActive" Ports & Pins, Motor Outputs, then yes I have tried it.
I may have found something, I don't know what video card I have. The system is fast enough, and has enough memory, but I don't know what video card I have, or how to find out.
Another thing I found, today I was listening to the radio, when I powered up my machine I got a lot of static on my radio.
Jeff
-
Is it an onboard card or is it an AGP?
Right click on desktop, choose properties then settings tab then advanced tab then finally adapter and it will tell you the card make/model etc you have.
Hood
-
Intel(R) 82865G graphics controller
Dac type- internal
96MB
I think for Mach we need a non integrated card, right?
onboard or AGP? I'm not sure
-
It is certainly recommended that you use a non integrated video card. Some work fine, some dont.
Hood
-
Every thing works very well, except this drift problem. Could the vid card be the problem?
Thank you for sticking with the newb, and his dumb Qs
Jeff
-
Suppose its possible but not convinced myself especially if the motors sound good and I think you said the distance is very repeatable..
Hood
-
How much drift are you talking about?
RICH
-
about 3/8" on the Y, and about 1/4" on the Z, running at 200 ipm for 5 min, with both changing direction quite a few times. I am trying to cut parts for rc model airplanes.
-
Try this, decrease your accel and velocity on one of the troubled axis to 25% of what it is now. Run the same file before and after this adjusment and let us know if the error is the same amount. As far as your steps per being right, do the math to get to that if you can. That will eliminate a lot of guessing, If you measure it your assuming you have not skipped any steps and often backlash isn't accounted for (depending on how you have been measuring).
Brett