I have tried to calibrate the machine 20-30 times, and the best I’m getting is 4.3mm … We plan to do logo/detail work and need better than that….. I’m at a loss. I’ve measured everything multiple times; I’ve used a caliper, a laser measure…. I don’t know what else to do.
Please help
-A desperate calibrator.
adam wrote:
I have tried to calibrate the machine 20-30 times, and the best I’m getting is
4.3mm … We plan to do logo/detail work and need better than that….. I’m at a
loss. I’ve measured everything multiple times; I’ve used a caliper, a laser
measure…. I don’t know what else to do.
my first question is how repeatable is the machine? historically the maslow has
been very precise, but not as accurate
if the machine is repeatable, you can distort your cad to correct for the
accuracy problem
how do the automatic measurements compare with measuring manually?
what are your current limits? we’ve found that the calibraion is more accurate
with lower current levels (less force on the bits) but you need to have arms
that move freely to lower the limits.
check your Z offsets. The defaults assume that your anchors are at the same
level as the surfact that the sled is riding on, if that is not the case, it
will throw things off.
David Lang
@dlang this is a MakeMade machine so it won’t have current limits.
@aa99 MakerMade sells copies of our machines from about 7 years ago so the design had advanced a lot since then. There is still room to improve on the calibration process, but we have made a LOT of progress since the chain days 
Bar wrote:
@dlang this is a MakeMade machine so it won¢t have current limits.
ahh, missed that.
We never did figure out the right math for that machine type. It is hyper
sensitive to the distance between the motors being incorrect (even the error of
a normal tape measure is enough to cause grief). This can’t just be corrected
out easily, the errors are not linear, they are curves.
It is very precise (repeatable) so once you have a design that corrects for it’s
errors, it will repeat that very readily. This would be something like cutting a
grid into a piece of wood, then loading a picture of that into your cad along
with whatyou intended to cut and see if you can find a cad tool that will
correct for the error (probably will take a couple iterations, with each
iteration getting closer, as the errors are not linear, the calculated
corrections may not be right either)
The makermade machine doesn’t support webcontrol and it’s holey calibration
method that was the best we came up with for that style of machine (it drilled 6
holes and with the 11 measurements between those holes did a mathmatical model
to try and figure out where the motors are)
David Lang
