More following area moving up than down (Z-axis)

I recently restored an old Dyna Myte 2400 (an old machine from the 1980s/90s) by rerouting its stepper motors to a Gecko G540. These steppers are somewhat odd in that they use a 3.6deg step angle. So far tuning my X/Y axes have been straight forward (I just needed to plug in some backlash compensation into my .ini file) but my Z-axis is having more complicated problems.

I've ran into two distinct (or at least I think they're distinct) problems with my Z-Axis. I'm measuring actual distance traveled by using a dial indicator.

1.) Directional Following Error: When I tune my following error such that commanding the Z to move down by 0.010" gives me 0.010" of measured travel (I originally had 0.001" of following error in this direction), I find that it still have following error in the upward direction. When I move up (after the system overcomes back lash), I consistently only get 0.009" of travel when I command for 0.010". Has anyone ever dealt with this before? Is there a way to tell the .ini file to use different FERROR values depending on if you're moving up or down?

2.) Random Travel Error: Sometimes when I tell the z-axis to move by 0.010" it only moves by 0.005". This will occur well after backlash has been overcome so I know it's not that. Also it typically only happens randomly (or seemingly randomly). Like I'll tell the Z to move down in 0.010" increments three times after overcoming backlash, and the first and third time it'll travel 0.010" but on the second time it may only travel 0.005". Any ideas on what could be the cause of this?

Thanks for the help!

First I would try checking the SCALE on the Z axis by making longer moves say 2" instead of .010
Second try reducing both the acceleration and the velocity to see if that changes anything.
Third try turning backlash compensation off and make several moves in the same direction to check the SCALE.
Do you have a adequate power supply and can the GECKO drives supply the correct amps?
And of course check for wear and mechanical problems, it is possible there is more wear in one area of travel than in others.

Rick G

Sounds like the extra torque required to lift the weight of your milling head is resulting in missed steps. On my manual mill, cranking up the X axis is hard work! I read somewhere that doubling the stepper voltage doubles the available torque. When building a rotary table indexer running on an Arduino, I found increasing the power supply from 19 volts to 48 volts avoided the need to upgrade a small stepper motor which proved what I read was on the money. I also experienced random missed steps prior to increasing the voltage.