Hello all…
First time using after assembly and running into an odd issue with the calibration process and now im stuck on where to go from here.
some details
4.1 - anchored to floor
running firmware 1.17
anchor point width/height - 4000 x 3300
Get to about waypoint number 14 and it then goes into low fitness retry
Have retracted and extended and repeated the find anchor point process and the same issue repeats
Maslow 1.17 calibration failed.log (15.4 KB)
any thoughts?
Austen wrote:
4.1 - anchored to floor
A common cause of this is if the Z offset for the belts is different from what
the machine thinks it is. The default values assume that the anchors are at the
level of the bottom of the sled.
are you running the maslow on the floor to calibrate? or are you running it on
top of some wasteboard?
you need to set the spoilboardthickness (go to the fluidnc tab, then the config
(round bubble on the right0 and then down about 15 or so items)
David Lang
A common cause of this is if the Z offset for the belts is different from what
the machine thinks it is. The default values assume that the anchors are at the
level of the bottom of the sled.
thank you for the speedy reply. i am running it on top of a wasteboard and have the anchors spaced to make them pull level with their appropriate belt spools. i will remove the spacers so that the anchors are on the floor and will set a value for the spoilboardthickness and try again. will report back
Austen wrote:
thank you for the speedy reply. i am running it on top of a wasteboard and
have the anchors spaced to make them pull level with their appropriate belt
spools. i will remove the spacers so that the anchors are on the floor and
will set a value for the spoilboardthickness and try again. will report back
No, in that case what you want to do is to change the Z offset for each arm.
enter the height difference between the arm with the router all the way down and
the anchor (bottom of the belt connector in each case)
David Lang
Looks like that seemed to solve the issue. got through calibration and saved to my .yaml and went to run my first test cut and got disconnected. rebooted and now the status is unknown so it looks like i have to go through the retract,extend, anchor point process again? is there something im missing to not have to calibrate every power up? sorry for getting off topic didnt want to start another thread
Austen wrote:
Looks like that seemed to solve the issue. got through calibration and saved
to my .yaml and went to run my first test cut and got disconnected. rebooted
and now the status is unknown so it looks like i have to go through the
retract,extend, anchor point process again? is there something im missing to
not have to calibrate every power up? sorry for getting off topic didnt want
to start another thread
you do not need to calibrate every power up, once you have valid anchor
locations, you have two workflows to remember
then you power on, extend the belts (if they were retracted), connect them to
the anchors, and apply tension.
in a crash, you may need to power on and release tension to be able to
disconnect the anchors.
now, the issue is going to be figuring out why you are getting disconnected.
there are other active threads on that topic right now, please read through them
and either ask questions there or start a new thread for your issue.
David Lang
Hello, I even have a lower fitness.
maslow is anchored to the floor fixed to 3D printed ‘Triangle Anchor’s. there’s no spoilboard so the machine is directly on the floor, little bit below the anchor points. Z-axis completely down. Can someone point me in a good direction to debug? Thank you!!
Maslow-serial.log (52.8 KB)
woody_s wrote:
Hello, I even have a lower fitness.
maslow is anchored to the floor fixed to 3D printed ¡Triangle Anchor¢s. there¢s no spoilboard so the machine is directly on the floor, little bit below the anchor points. Z-axis completely down. Can someone point me in a good direction to debug? Thank you!!
Maslow-serial.log (52.8 KB)
v1.18 just came out, which includes some calibration fixes, try that.
I would also set the wasteboardThickness setting in the maslow.yaml, if the
surface your sled is on is below where the anchors set (and all anchors are the
same height), then this value will be negative. I doubt this is the deciding
factor here, but it can’t hurt to set this correctly.
There is nothing obvious in the logs.
are the belts pulling tight at each measurement point? if not, you may need to
increase the retraction setting.
David Lang
Thanks, I just upgraded to 1.19 and tested the spoilboard thickness. I set it to -10 and also to 0, but the result remains the same.
What I find strange is that my frame is recognized as:
Frame dimensions: 4352.4mm x 4352.4mm (aspect ratio: 1.00:1 )
but in reality it is somewhere around 3800 x 4800. Could this be an issue?
Also, when assembling it, I had problems with the TL motor. It was slightly damaged, a wire had come loose and it wasn’t working.
Maslow-serial.log (30.8 KB)
After repairing the wire and tapping it a few times and I got the motor working again. But when I set a retraction force of 900 on this one, doesn’t engage, whereas it does at 1000. The other belts work perfectly with a retraction force of 900.
The belts are all tightly tensioned when searching for the anchor points. I could play the harp on them 
We’ll get there, I can feel it.
From looking at the log I don’t see anything wrong 
It seems like everything is right.
I took the numbers out of the log and plugged them into the math for finding the anchor points and I got:
Fitness: 0.3604593
Maslow_tlX: 82.2
Maslow_tlY: 1879.3
Maslow_trX: 5855.9
Maslow_trY: 2069.3
Maslow_blX: 0.0
Maslow_blY: 0.0
Maslow_brX: 5876.2
Maslow_brY: 0.0
The fitness isn’t quite good enough to move on to the next step, but it’s WAY better than the .02 that we were seeing in the log.
Every time that you run the anchor locating process it will improve so having a good starting point can help it to go smoother.
The next thing that I would try is to enter those values that I found in for the anchor point locations and then try again. That should give it a better starting point so hopefully the math will converge better.
To enter them you want to use this config items menu:
And then scroll down to these settings:
Thanks for the insight. I tried to adjust the values, but I get an error message:
[MSG:ERR: Frame dimensions out of bounds. Top=5226.5mm,
and the red led is blinking
maslow.yaml (6.4 KB)
Maslow-serial(1).log (3.2 KB)
Ah right, we have a limit on a maximum frame size programed in but, playing around with the simulator it seems like that is too restrictive of a maximum size.
I’m going to up it to 5500mm.
I believe that you are the first person to ever hit that size limit which is kind of a cool achievement 
I should have a new firmware version with the limit increased later today.
Edit: Give this one a try, that should have a large limit.
firmware.bin (1.9 MB)
index.html.gz (133.9 KB)
Oh no, silly me , I had to follow the recesses in the ground a little to place the anchor points so that people wouldn’t trip over them. Thanks for the quick update, I appreciate it.
I’ve already run the search for anchor points 3-4 times and I feel like it’s going in the right direction.
The second to last time was slightly better than the last time, but I’ll try again next week when I have some time.
My frame is about 4800x3800, so that’s almost reflected in the yaml file.
Maybe I’ll also play around with a negative spoilboard thickness, since the Maslow anchor points are the 3D-printed versions and the Maslow is on the ground. Greetings
Maslow-serial(5).log (52.5 KB)
maslow.yaml (6.8 KB)
woody_s wrote:
Oh no, silly me
little to place the anchor points so that people wouldn¢t trip over them.
Thanks for the quick update, I appreciate it.
when your anchors are not level with the bottom of the sled, you need to tell
the machine that.
you can set the individual arm Z values, or if it’s the same for all four
anchors, you can adjust this with the spoilboard thickeness value (settings →
config or in the fluidnc tab)
a postitive number is the height up from the anchors to the top of the
spoildboard
a negative number is if the anchors are above the spoilboard.
David Lang
