Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: smartarse101 on March 05, 2011, 05:28:33 AM
-
Hey,
I have just built my own CNC mill from scratch. but i have found a problem that i cant seem to solve.
to cut a long story short my x axis which is directly driven by a leadscrew will not consistently match up with the DRO readings in mach 3.
for a start, my motors are running at 200 steps/rev, driving a 1.5mm leadscrew, controller is set to 1/8 microstep. which works out to be 1066.6 steps per mm. but moving the axis 100mm on mach 3 will only physically move the gantry about 80mm. i have played around and found a number that will move it almost exactly 100mm (some fine tuning needed) but when i went back to it the next day to fine tune it, the steps per mm were still the same as what i had programmed it to the day before but it was moving 20mm more!
i cant think of any other way to describe this problem but ill answer any questions u have for more information.
please help. its getting very frustrating! lol
-
Sounds like it could be missing steps but it could be a lot of other things as well. :-\
When you jog your axis from one extreme to the other do the steppers make a nice smooth sound or do you sometimes get a little 'tick' or hesitation in the sound ?
Have you set your motor tuning velocity and acceleration to a comfortable figure well below the point at which the steppers stall ?
Tweakie.
-
Do as Tweakie posted and with the lowered vel and accel values click the Settings Tab and in the lower left corner of the screen you will see
Axis Calibration. Use that to check and set the steps per unit. Make sure you save settings under config. Go out of Mach and come back in, your settings should be what you saved.
Now do a move in one direction and then the opposite to check for backlash.
Let us know how things go for you.
RICH
-
Ive done all of the above, im confident that the motor is not missing steps. and it has almost no backlash on the axis at all. the velocity and acceleration are both set to a nice low value and when i try to stall the motor, i have a nice amount of torque there.
-
to cut a long story short my x axis which is directly driven by a leadscrew will not consistently match up with the DRO readings in mach 3.
For this statement to be true the step pulses that Mach sends out are not being translated correctly by the stepper motor.
Have you thoroughly checked all the obvious mechanicals such as slipping coupling joining leadscrew to motor, tight or binding spots on the slides or leadscrew, anything loose etc etc ?
Tweakie.
-
I have checked the coupling,turned the leadscrew by hand moving the axis from one extreme to the other a couple of times and cant find any tight spots, all fixings are nice and tight, the axis is moving nice and square, bearings are fine.
I have swapped the motor with another of exactly the same type and i still have the same problem.
-
I had some strange issues with my setup, when I would jog, the motors would move in one direction 27 in a min. but when jogged in the reverse direction they would only move 1 inch a min. and the problem would change directions. Out of curiosity, I loaded up the mach3 mill profile and plugged in my settings and wala! all my issues were gone. I went back to my custom profile and did a restore from the menu bar OPERATOR>Restore Settings... and rolled back one setup config and it got rid of my problem. Somehow the custom profile got corrupted.
I'm not saying this is your problem but you might want to try setting up a different profile using your settings and see if the problem carries over to the new profile. Just something to try.
-
What make and model stepper driver or drivers are you using ?
Perhaps there is a problem with the board configuration.
That is a good point about re-doing your Mach profile from scratch, can't do any harm
This is all becoming bit of a mystery but, from past experience, the solution will be simple once it has been identified.
Tweakie.
-
Hey, sorry this has taken so long, ive been away for work.
I have completely programmed mach 3 from scratch to the point of completely un-installing and re-installing a newly downloaded copy. Still the same problem is there.
My Driver board is a HY-TB3DV-M (imported from china, cant find the make).
Im starting to think i may need to buy another driver board and give it a go.
-
The basic setting up instructions for this board are here http://fozztexx.com/CNC/HY-TB3DV-M_3Axis_Driver.pdf may be worth checking the settings against against your setup of both the board and Mach.
My guess is that you have tight spots or binding (perhaps lead screw, perhaps linear slide) which are causing the motor to loose the steps and thus produce the inaccuracy you describe. Providing the motors are not getting hotter than you can keep your hand on comfortably then you could try increasing the drive current and or increasing the drive voltage to 36 Volts.
Tweakie.
-
I have mach 3 set up as per the instructions.
I have removed each lead screw from the coupling, and turned them all through their complete length multiple times. there are no spots where anything is binding.
The motors are getting a little warm,i can keep my hand on them, but i wouldn't say its comfortable if you know what i mean. they are a little hot.
-
Hmmm, sounds about right then for motor current. I will give this matter some more thought. :-\
The driver board is not the best but it is not bad either so should work OK. There is an issue with some opto isolators requiring the full 5 volt TTL levels and PC parallel ports only outputting the 3.3 volt TTL logic signal. If you can, it may be worth checking your LPT port to see if the direction signal pins swing approx 0 to 5 volts when jogging an axis or if they only make 3.3 volts.
Tweakie.
-
how would i go about that? just with a multimeter from the pin to ground?
-
Yes.
Pins 18 to 25 are ground (negative lead of multimeter) connect the positive lead to a step pin and then jog the axis one way it should read approx. 5 Volts and the other way it should read approx. 0 Volts.
Tweakie.
-
Id check the pulse width in Motor tuning. Bump it up to 5 and try it