I run the find anchors function and it goes through everything fine and even completes it but then will just keep resending the same coordinates. Not sure why? I’ve refreshed the browser. Deleted the browsing data. Used a different internet browser. It just stays in this loop until I unplug it. Here’s the info from the first calibration and after a browser refresh.
Maslow-serial (Initial).log (32.0 KB)
Maslow-serial (Refresh).log (49.6 KB)
what firmware version is this? (check both the index.html.gz and firmware.bin
versions)
This is saying that at this point in the process, the UI is not responding, it
may be that it’s dropped out of the ‘find anchors’ mode (you clicked on
something else or something)
David Lang
Both are V1.21
After reading other forums I placed my results in this calculator ( Maslow Levenberg-Marquardt Calibrator ) and then input them manually into my yaml.
Now I am running a dry G-code and this is happening in the top left corner.
Any help would be appreciated. Here’s my yaml as well.
maslow.yaml (6.9 KB)
Nicholas Van Sipma wrote:
After reading other forums I placed my results in this calculator ( Maslow Levenberg-Marquardt Calibrator ) and then input them manually into my yaml.
thanks for testing this.
Now I am running a dry G-code and this is happening in the top left corner.
what angle do you have the machine at? tipping it more towards horizontal may
help
this could be caused by three things
how tight are the belts? tight or guitar string tight?
is it only in that corner (what about the top right corner, that has a similar
difference in belt heights, but the top belt is not at as steep an angle
tipping it closer to horizontal will make the weight of the router on the sled
do a better job of keeping it flat. The initial guidence was 15 degrees from
vertical (because that’s what worked well on the first gen maslow) but we are
finding that the maslow 4.* wants a little bit more of an angle, 20-25 degrees
David Lang
I have found opening a second browser to the Maslow, while leaving the existing window open, forces the computer to read the coordinates and continue. The next version (due for release soon) does not have this problem as all processing is done in Maslow.
For your tilting problem, I suggest you put a block under the TL Anchor Point and raise it 100mm then adjust the Z value to compensate.
The machine is at 20 degrees. The belts are just tight not overly excessive tight. Its only the top left side. It does it from about the middle to the top. The bottom left seems to be fine.
Only the top left anchor?
I have all belts parallel but try the top left first. I am on a horizontal table. I did try vertical but couldn’t get enough height.
I have seen recommendations a while back the angle for vertical should be 25 degrees at least.
Have you got room to run it horizontally?
I can easily change the angle as the top is on hinges. Space is an issue so I opted for vertical. I’ll try 25 degrees
Even with the frame at 30 degrees and the mount raised up to level it still tilts and doesn’t stay flat. Anyother suggestions? Or is it hopeless and just try to use it horizontal?
Is it any better? I haven’t had much experience with vertical as I was always worried that a belt break would send it crashing to the floor.
No it’s the same. There was no change at all.
What do you have the retraction force set to? If you can run the anchor point locating process with the force lower it will lead to the belts being less tight while moving which can help the machine stay flatter.
Does the position of the z axis have an impact?
I have it set to 500 and no it doesn’t. Tilts no matter how high or low the z axis is.
Bit desperate here, could you try putting the top two arms on BR BL and the lower two on TR TL. You would have to adjust the Z values in yaml file to suit.
@bar any ideas, you have run vertical more than most.
Maybe the belts are not quite tight enough, try setting to Retract force to 700-800 and running a Find Anchors.
I actually already tried this and it tips it the opposite way. Where the top lifts away instead of the bottom
Maybe try playing around with that, must be a balance point in between. Which was it you tried, moving the arms, or changing the Retract Tension?
Ok so I moved the CNC to Horizontal. Ran it manually to the left corner and it still lifted. Lowered it all the way down and measured the EXACT height to the center of all belts and put the amounts into the correct Z axis’s. Then ran a Find Anchors and now am on the endless loop again. I threw the log into ChatGPT to see if it could find why and it pointed out that this number: {2961.58} is placed randomly at the end of the log with no indication to where or what the number is for. I’ve tried opening dual browsers, 2 different browsers simultaneously, clearing browsing data on both. Still stuck in an endless loop of data not acknowledged by computer. Any ideas? @dlang @ian_ab @bar
Log:
Maslow-serial (2).log (9.6 KB)
