I've spent a good bit of time wrestling with the issue of flat spots on what should be circular features. They occur at the topmost and bottommost (+y and -y) quadrants of the features. This of course led me to suspect backlash in the y-motor assembly, but even after correcting for it in Mach 4, using a dial indicator to check how much the machine actually moved compared to its nominal movement, the flat spots remained. Varying this correction value seemed to have no effect on spots, either, so I am beginning to wonder if there are any other factors that could contribute. Any and all suggestions would be welcome!

Machine info: Retrofitting a Shopbot PRSAlpha to work with Mach 4 via the PMDX-422

Backlash compensation must be implemented in the hardware device and its plugin.

The SmartBOB family does not yet support backlash compensation.

Yes, it is on our list of things to do. We are currently looking for the best way to
do this. There is no magic bullet because the compensation cannot be done in
zero time. This means that some compromise must be made. It may seem that
Mach3 or other systems have already solved this problem, but behind the scenes
there were compromises being made. Our task is to work out the least worst
way to do it.

The only good solution to backlash is to improve the mechanical system to get
rid of it.

Steve Stallings
www.PMDX.com

if you check your Y axis with a DTI over many move`s like 100 small and long moves in Y axis only moves starting and finishing in the same spot does it start and finish in the same spot you can get a axis to + or - 0.01 with one move, 100 could be = or - 1.0.

I check at 1000 moves on any axis doing one at a time it can answer many questions to what`s wrong