need help calibrating my ox

how about this thread?
http://openbuilds.com/resources/howto-calibrate-your-ox.145/

 

steppers like high voltage, so use about 80% of what your drivers can do. most small drivers are rated to 30 volts, so use 24 volts.

however, the steppers don't like more current than they are designed for.
the openbuilds NEMA23's for example, are rated for 2.8 amps so you have to limit the current to no more than that. even so, at over 2 amps they will get very hot (which is why industrial drivers go to half power when standing still).

most small drivers don't really have enough heatsink for that much current , even if they can do it, so try for about 1.5 amps.
if you cannot put a finger on the chip for more than 5 seconds it is too hot, lower the current (not the voltage).

now, if your steppers are running fine and then go a bit wild the problem it probably interference. all the pulsing to make them step makes a lot of electrical noise which can get into the input circuits and can cause crazyness.
have a look at this
http://openbuilds.com/resources/cnc-electronics-training-videos.44/
and this
http://www.openbuilds.com/threads/electronics-101.693/

to minimize electrical interference:

• separate the signal wires from the power wires, never run them parallel to each other.
• shield the signal wires
• have only one common earth point

the other factor is speed, acceleration and binding.
any binding in the mechanism will cause missed steps, especially at higher speed/acceleration.
stepper motors get weaker the faster they go, and are weaker with more microsteps, so use the smallest microstep setting that still gives you the resolution you want.
reference http://www.openbuilds.com/threads/micro-stepping-myths-and-realities.2478/

so, disconnect the belts etc and move the carriage by hand to make sure there are no rough spots.
then, set your top speed fairly low, say 1000mm/min and a low acceleration too, and test end to end motion using G0 commands
increase acceleration (just double it each time) until the motor stalls during the startup.
go back to the previous setting that worked properly.

now, increase the top speed and test again until it stalls or even loses a few steps. reduce to 80% of that setting.

so say you have a 1 meter long travel on X you should be able to do: (starting at 0)
G0 X1000
G0 X0
and the carriage must return to exactly the same place. if it does not, reduce speed until it does, then try to increase acceleration and find that limit then reduce to 70 or 80% of the max you found.
THEN test while cutting something at a normal cutting speed, cutting forces can make you reduce the acceleration further to where it is reliable.

 

strange, I had no problem doing a number of test setups with GRBL. default settings ran all the motors I tried, and tuning for speed etc 'just worked'.
maybe it is because I read the manual? (-:

I did find that some motors do not like microstepping and will do odd things. set your microstepping to '2' to check that this is not the problem. this happened with some motors recovered from old printers. the NEMA23's from Openbuilds 'just work'.

 

I will have a look when I dig out the arduino, but I don't think i have a record of the settings that worked well since jumped around a bit testing various motors and cable drives.