We were able to calibrate our Maslow in October and make our first cuts, but we ran into issues and decided to re-calibrate. (probably a bad idea.) To calibrate, we used the latest maslow.yawl, updated the firmware and used the latest index.html files.
Despite trying for 2 days we cannot get the calibration to complete. Attached is a photo of our setup from our first try and cut), as well as one of our latest logs.
Does anyone have any suggestions?
hi,
I’m probably not a complete specialist, but I see the anchors being raised to the height of the coils? How stiff is it? Doesn’t it flex?
Hi - the belts are attached to 3/8 threaded rod sunk in to the garage floor. They do flex a little bit, however the last time I was able to calibrate I used the same bolts. I also tried to calibrate with the belts attached lower on the bolts, but it didn’t seem to make a difference.
what are the z offsets set to? if you updated to a new starting maslow.yaml file
you would need to set the Z offsets for your frame.
David Lang
I didn’t update the maslow.yaml file (attached here). We set the Z axis as far down as it would go, following the calibration video Bar posted.
maslow (2).yaml (6.8 KB)
kinematics:
MaslowKinematics:
tlX: 139.199997
tlY: 3230.600098
tlZ: 100.000000
trX: 4308.899902
trY: 3230.699951
trZ: 56.000000
blX: 0.000000
blY: 0.000000
blZ: 34.000000
brX: 4340.399902
brY: 0.000000
brZ: 78.000000
these Z values are assuming that your anchors are at the height of the bottom of
the sled.
If you have the anchors raised to level with the arms, then all these Z values
would be 0
David Lang
My anchor points are definitely above the bottom of the sled, but not level with the arms. I’ll set the anchor points to be level with the bottom of the sled and try again. Is that your recommendation?
Jeff wrote:
My anchor points are definitely above the bottom of the sled, but not level
with the arms. I’ll set the anchor points to be level with the bottom of the
sled and try again. Is that your recommendation?
the important thing is that the Z values should be correct for your setup (much
more important than what they are)
flex is bad, the more level the belts the better. but it’s a trade-off for your
setup, and without knowing the flex you see, it’s impossible to know.
so you can try both, just make sure that when you change the anchor heights,
tell the machine the right values.
David Lang
Today I downloaded a new.yaml file and updated the file and set Z values to 38.1000000 (1.5 inches); anchors are 1.5 inches off the floor. I verified the anchors do not flex.
I attempted a number of calibrations, each with different grid sizes, but they all failed. The log files from the attempts are attached.
Any suggestions?
Maslow-serial_2000x1000x9x9.log (200.9 KB)
Maslow-serial_300x300x3x3.log (25.5 KB)
Maslow-serial_440x400x3x3.log (16.5 KB)
Jeff Wrote
Today I downloaded a new.yaml file and updated the file and set Z values to 38.1000000 (1.5 inches); anchors are 1.5 inches off the floor. I verified the anchors do not flex.
I attempted a number of calibrations, each with different grid sizes, but they all failed. The log files from the attempts are attached.
The Z offset isn’t the distance from the floor, it’s the distance from the
anchor to the arm
the defaults in the file are correct if the anchors are even with the bottom of
the sled.
If all your anchors are the same height, then leave the per-anchor Z values at
the stock values
use the wasteboard field to record the distance from the top of the wasteboard
to the anchors. If the anchors are above the wasteboard, this would be a
negative number
David Lang
From looking at the logs it seems to be consistently having a hard time with the fourth measurement point:
Are you seeing all four belts getting tight with each measurement? This looks to me like the machine might be having a hard time taking accurate measurements.
Following David’s direction, we updated the .yaml file to show the anchor points 1/2 inch above the spoil board.
We tried 2 more times, each failing (see attached files).
We did notice that the bottom left belt did not tighten when the machine started. It only seemed to tighten when the sled needed to move in its direction.
Thanks David and Bar - Jeff
Calibration 11112025.docx (71.9 KB)
Maslow-serial400x400x3x3).log (12.2 KB)
Maslow-serial_2000x1000x9x9.log (11.0 KB)
I think that this is the issue. You want it to look like it does in this video with all four belts getting tight when it takes a measurement:
I think that increasing this setting here is probably the key to making it pull tighter when it takes each measurement:
Thanks Bar - I was able to get through calibration this time! I used a smaller grid (400 x 400, 3x3); I noticed the fitness values weren’t as high as I’ve seen before. Will this be an issue?
Thank you for your help, Bar - Jeff
Maslow-serial (10).log (14.5 KB)
That is still a bit low and higher is better so might be worth trying to improve on those results, but I would make a backup of that Maslow.Yaml file so you can always reset to here
We tried calibration again yesterday
maslow CALIBRATED 11122025.yaml (6.8 KB)
and got a very good result. (over 1.0); we were able to make a successful cut yesterday. Today we tried a second cut, however we weren’t able to get the machine to move. We’ve retracted and extended the belts a number of times, and two of the belts aren’t always retracting. In all cases we get the message that the Maslow can’t move. Any suggestions? Latest log is attached.
Maslow-serial (12).log (12.4 KB)
I think that this is likely the issue. Can you try increasing the retraction force under Setup → Config?
From the log it looks to me like one of the belts might not have been fully retracted before extending which is leading to it’s position being off.
[MSG:INFO: Requesting state change from Extending Belts to Belts Extended]
[MSG:INFO: Succeeded]
Apply Tension
[MSG:INFO: Requesting state change from Belts Extended to Taking Slack]
[MSG:INFO: Succeeded]
Apply Tension
[MSG:INFO: Requesting state change from Taking Slack to Taking Slack]
[MSG:INFO: Cannot take slack until the belts have been extended]
[MSG:INFO: Measured waypoint 0]
[MSG:INFO: Center point deviation: TL: -0.000 TR: -0.000 BL: 52.914 BR: 6.241]
[MSG:ERR: Center point deviation over 12.000mm, your coordinate system is not accurate, maybe try running calibration again?]
[MSG:INFO: Requesting state change from Taking Slack to Belts Extended]
Basically the machine is detecting that something is off with it’s belt lengths and refusing to move to prevent it from damaging itself.