I programmed a 30” circle and cut a symetrical elipse 28x31. Has this happened to anyone else? When I rebuilt the Maslow, I might have changed the sequence of the four arms. It seems that my x dimensions are less than my y dimensions. I hope it is proportional. Calibration went relatively quickly compared to earlier calibrations.
1 Like
This is after updating to 1.12 which came out yesterday?
This means that your anchor locations were not accurately calculated.
What you need to do is to set your Z offsets properly (the distance from the
anchor to the arm. measure where the belt end is at the arm, and measure where
it is at the anchor, the difference is your z offset)
Then you can set your wasteboard thickness and run calibrate (now called ‘find
anchor locations’) again and the anchor locations found should be noticably
different than what you currently have in your maslow.yaml file.
There is a parameter to stretch X and Y, but if you have your arms in a
different order, that means that the stock Z offsets are going to be wrong, so
the calculations for the anchor locations is going to be wrong, so you should
fix that before you try to scale things.
David Lang
1 Like
I am at the previous update 1.10 or 1.11.l
1 Like
Thanks David. So I should measure the height from the top of the waste board to the arm retracted, then measure the height from the top of the waste board to the height of the anchor for each arm, and enter the difference into the z offsets section of the maslow.yaml file for each arm. Then enter the height of my waste board (.500”) into maslow.yaml, and run “find anchor locations.” I’ll try that today.
I would try updating to 1.12 and see if that fixes it.
I wouldn’t mess with the z-axis offsets unless you are sure that you swapped the order of the arms when assembling
1 Like
I’ll update and re-calibrate first, before doing the z-offsets. Thank you Bar.
1 Like
Stephen Roberts wrote:
Thanks David. So I should measure the height from the top of the waste board
to the arm retracted, then measure the height from the top of the waste board
to the height of the anchor for each arm, and enter the difference into the z
offsets section of the maslow.yaml file for each arm. Then enter the height
of my waste board (.500”) into maslow.yaml, and run “find anchor locations.”
I’ll try that today.
If you measure the height from the anchors to the arms with the sled on the
wasteboard, that distance will include the wasteboard. If you think that you
will change the wasteboard to a different height later, subtrack the wasteboard
thickness from the per-arm numbers and enter the wasteboard separately.
When the maslow tries to figure out right belt length it needs to find the
height of the triangle from each anchor. it takes the per-arm Z offset, adds the
wasteboard, then adds the workpiece, then adds the Z position of the router
The default Z offsets are the distance from the bottom of the sled to the bottom
of the belt ends when they are fully retracted.
David Lang
1 Like
I updated to 1.2 then re-calibrated. Jogged a square and used a caliper to measure. Still off. Measured my heights and changed z-offsets in maslow.yaml. I Realized that I had changed the order of the arms. Did a re-calibration, jogged a new square and measured again. All good! Re-calibration was very quick in version 1.2 and I was able to lower the retraction force to the defaults I feel better about this because the Belts were too tight, putting a strain on my frame. Thank you BAR and thank you David. I really appreciate the help.
1 Like
hi,
great post. I’m currently on firmware 1.7 and I didn’t set any offsets. The calibration was long, and the fitness came out around 0.5 and the result was from the X and Y axes was “disproportional”. The axis scale solved it. I recalculated the differences and set the scale.
Then it cuts me about 1mm / 2m accuracy. I’m happy with that for now.
So I’m wondering if I should start with the new firmware? Will it mean the need to change the Z-offset and then calibrate and then play with the scales again? Or will the scales then no longer be needed?
I’m currently on the ground on concrete with the dimensions of the frame rectangles 4.4 x 3.6m. thanks JURA
for the reminder .
You shouldn’t need to re-run the calibration process when you update the firmware.
By firmware 1.7 do you mean 0.7 (we’re only up to 1.12 
)?
If you are going to update from below 1.0 to above 1.0 that process will wipe your settings, but the only information that we really care about is the anchor point locations.
You can write those down from the settings before the update and type them back in and not need to re-find the anchor points.
JURA23 wrote:
great post. I’m currently on firmware 1.7 and I didn’t set any offsets. The
calibration was long, and the fitness came out around 0.5 and the result was
from the X and Y axes was “disproportional”. The axis scale solved it. I
recalculated the differences and set the scale.
check in different areas of your work area. the errors from inaccurate anchor
locations do not tend to be consistant across the work area.
Then it cuts me about 1mm / 2m accuracy. I’m happy with that for now.
So I’m wondering if I should start with the new firmware? Will it mean the need to change the Z-offset and then calibrate and then play with the scales again? Or will the scales then no longer be needed?
If you are happy with it, don’t change it.
when you upgrade past 1.09 the location in the maslow.yaml changes, and it
doesn’t copy the values, you will have to re-enter them in the new location
otherwise, just keep a copy of your maslow.yaml.
we’ve seen improvements in accuracy by setting the calibration current limit as
low as we can go and be accurate, and by making sure that Z offset, workpiece
thickness and wasteboard thickness (and that the router is all the way down) are
all correct for your setup
I’m currently on the ground on concrete with the dimensions of the frame rectangles 4.4 x 3.6m. thanks JURA
That frame size looks good.
David Lang
1 Like
Thank you all for the advice. I wrote it wrong. I’m on 1.07.
1 Like
I would recommend updating to 1.12, I think that we’ve found and fixed a whole bunch of bugs since 1.07 and I think it would be a bummer to track down what is going on only to discover that it’s something we already fixed.
1 Like