Aloha,
After finally getting a decent calibration, I used jog controls to move the sled to home. The first move or two was as directed (100mm diagonal down right), but after that any jog in any direction moved down right and significant slack developed, to the point where I had to stop and sort it out.
My calibration was with a 1400 current threshold for both retraction and calibration. Can I increase those thresholds without rerunning calibration?
Mahalo,
David Negaard wrote:
After finally getting a decent calibration, I used jog controls to move the
sled to home. The first move or two was as directed (100mm diagonal down
right), but after that any jog in any direction moved down right and
significant slack developed, to the point where I had to stop and sort it out.My calibration was with a 1400 current threshold for both retraction and
calibration. Can I increase those thresholds without rerunning calibration?
you can, but they only affect retraction and calibration, during normal
operation it will go up to 4000
David Lang
This sounds like a bug maybe unrelated to calibration? I’m not really sure what is happening there. I would start out by just doing power off and back on followed by Retract All → Extend All → Take Slack and see if that fixes it.
Mahalo for that clarification. I will continue to strive.
I did as you recommended (twice) and when I jog, some belts slacken to allow movement and do not take up that slack. That seems to be the relevant symptom. Top right and bottom left were the ones I noticed most because I was jogging down right, but trying multiple jogs I see the same symptoms on any belt.
David Negaard wrote:
I did as you recommended (twice) and when I jog, some belts slacken to allow
movement and do not take up that slack. That seems to be the relevant symptom.
Top right and bottom left were the ones I noticed most because I was jogging
down right, but trying multiple jogs I see the same symptoms on any belt.
retraction force would not help this.
what was your calibration fitness?
are you sure the Z was all the way down when you did the calibration?
David Lang
I’ve seen the same behavior but didn’t spend the time to find a repeatable pattern or document it. Anecdotally it seemed to happen to me as I did repeated diagonal jogs toward the bottom left of the workspace. I wonder if the logic to unspool some of the trailing belts is being over aggressive in its calculations as the machine is further away from center?
Bob Craig wrote:
I’ve seen the same behavior but didn’t spend the time to find a repeatable
pattern or document it. Anecdotally it seemed to happen to me as I did
repeated diagonal jogs toward the bottom left of the workspace. I wonder if
the logic to unspool some of the trailing belts is being over aggressive in
its calculations as the machine is further away from center?
it sets all the belts to the length that it thinks it needs for that location,
if you are getting a lot of extra belt, that indicates that the frame size does
not match what it thinks it does (bad calibration), or that the retract/extend
cycle didn’t fully retract the belts (which would lead it to thinking the belts
are shorter than they actually are)
David Lang
Also double-check you have the arms in the right order, and have zeroed the Z axis (drive it down until it clicks with no bit in it, then hit zero-z button on the setup dialog).
I was not hitting zero-z in setup! Otherwise, all good. Arms are in the right order.
I started over—flashed the latest firmware and installed fresh index.html.gz and maslow.yaml, and got a good calibration and everything seems to be working. No more slack belts, no more odd directions, none of that.
Part-way through a cut, it stopped due to an overcurrent condition, and said I could restart at line 91? I have no idea how to do that, nor do I know why it happened. Maybe feed is too fast?
I’ve been having similar issues with over spooling. I really gnawed a belt yesterday in the process.
Double checked the arm install order which was correct, and I have my z-height all the way down with no bit when calibrating.
It’s difficult to know what I should expect in “normal” operation. I was beginning to just assume the over spooling is part of operation. Then I had one go where the Maslow actually tensioned everything and it seemed to be moving correctly. That was followed by more errors.
I left it yesterday with
[MSG:WARN: Position error on Bottom Left axis exceeded 15mm while running. Error is 15.057mm Counter: 6]
[MSG:WARN: Previous error was 15.057mm]
[MSG:ERR: Emergency stop! Stopping all motors]
[MSG:WARN: The machine will not respond until turned off and back on again]
I’m assuming the frame measurements need to happen from the center of the bolts that hold the belt ends?
I’m also having an issue where when I update the firmware, the startup messages in the console show the previous version. Was saying 0.74 after I had update to 0.75, and after updating to 0.76 yesterday, it now says 0.75. Issue persists even follow restarts.
This is an issue that I’m working to track down right now, but it’s not anything that you are doing wrong. It seems like it might be related to EMF, and the question we’re working on right now is is there a source of EMF that we can reduce or can we make the code more robust to handle it without issue. 100% not your fault. Hopefully this week’s update will help
Hey bar, thanks so much for chiming in and letting me know. Really appreciate it. I’ll be gone a lot over the next two months, but if there’s any more information I can gather for you to help in troubleshooting let me know. I’d either have to tomorrow, or July 1-3.
Also, I didn’t realize the Maslow had caught a belt when I was most recently working with it and it snapped it completely. I’m working on a smaller table right now, but I’d like to replace the belt if possible so I can have the full length to work with when I’m ready for it. Can you point me where to purchase them?
Thanks for all the work you do making cnc available! It’s one of the coolest projects out there.
Absolutely, we’ve got all the spare parts up in the shop here: Shop — Maslow
Getting that bug fixed is the top priority right now because it seems like it’s the last thing holding a lot of folks back right now so we should have a solution ASAP!
I had this issue today as well. I got the 15mm position error. I didn’t restart and got an error saying that the yaml file was corrupt try a restart. After restarting the bottom right belt kept unspooling. Here is the serial messages when it was unspooling the belts. Side Note: this is the first time I’ve used dust collection a.k.a. vacuum attached.
unspooling serial messages.log (4.8 KB)
In my testing this seems to be a big factor, not the only factor but a big factor for sure. We’re investigating that exact setup right now:
Just posting to say I had the same issue yesterday, running firmware 0.77. I Got about 90% through a program and stopped with an over current. I raised Z and tried to jog and all of the belts came unspooled like this. I was able to get the belts back in and thought everything was ok, but now bottom right seems a little chewed up and it isn’t moving right when extending belts so I guess I’m going to take that arm apart and see what is going on.
Poking this issue… I had some spare time to get back in the studio with the Maslow today since posting in June.
I started by updating firmware.bin, maslow.yaml, and the index.html.gz files.
I recalibrated the Maslow, which completed successfully.
My maslow is still experiencing issues once it calibrates. It still shoots out a bunch of extra belt at times. Trying to run a gcode file usually results in a position error exceeding 15mm. Restart, retract, reattach, tension. It all happens again, sometimes reporting that maslow.yaml is corrupted but a restart might fix it.
Opened maslow.yaml on my computer. Diffed it with the maslow.yaml I downloaded this morning. Changes seem reasonable.
Cut all the belts to be shorter so they wouldn’t bunch up so much when over spooling. Console suggests I should try running calibration.
Recalibrate. Maslow shoots out a bunch of extra belt while moving. I tension belts. Try to run the Maslow sample gcode. Position error. Restart, retract, reattach.
Is this still an issue other people are having? I can be a bit daft sometimes, am I just missing something?
Although I have printed the dust collection attachment, I haven’t attached it yet. Maslow is on a table in the middle of the room, running horizontally.
Do I need to order a replacement pcb? I’d be excited to figure this out and be able to maybe cut something 
what are the calibration results? how do those measurements compare to manually
measuring the distances between anchors?
are you remembering to lower the router all the way to the bottom before you do
the calibration?
When you are doing the calibration, what is the Z distance from where the
anchors are to the bottom of the sled?
@bar, I believe that the stepper drivers you used support sensorless homing (aka
skipped steps), are they wired up in a way to support that? can we make it so
that when you do calibration, it runs the Z motors down until they skip steps?
if not, should we just run the Z down 50mm worth (pretty much enough to
guarantee they are all the way to the bottom?
David Lang
