That was a temporary fix. I replaced the belt ends and they are now installed inside the printed frog mouths.
Can you take a photo, trying to visualise you setup
Does it look like I have one spool that is not measuring lengths properly?
I’ll have to take a photo when i get home, currently at work. Is there specific parts of the setup you would like photo’s of?
Max deviation on Bottom right seems to be weird
Try David’s suggestion.
Measure from Anchor Point bolt to bolt, not just the outside frame size. The more accurate you can get the better.
David he is running V.15 at the moment.
when the fitness starts below ~0.2 is seems to never recover, that means that
there is a problem with those first measurements.
IIRC you said you have your frame very square (worried about being off 1/4" or
something like that), measure the width and height of your frame and enter those
numbers in the maslow.yaml and then do the calibration. The 1.20 code seems to
not do well if the starting frame size is wildly off (as is being discussed
elsewhere)
If I’m remembering wrong and you aren’t the person who had the very square
frame, measure all 6 distances betwene the anchors (including the diagonals) and
post them, we cand run them through a manual calculator like the one I have in
onshape at
to get the anchor locations
David Lang
The anchors are within >1/4 of square. I’ll try manually entering the numbers tonight. Her are some photos I found on my phone.
Bones wrote:
The anchors are within >1/4 of square. I’ll try manually entering the numbers tonight. Her are some photos I found on my phone.
getting exactly the right numbers in the manual calculator and you can just
enter those values and cut (no need to do the find anchors routine at all) but
just getting close should make it so you can run the calibration successfully
looking at those, it looks like the anchors will raise the belt ends up a bit,
so you would want to enter a negative wasteboard thickness (since the bottom of
the sled is below where the anchors are) by the thickness of the base of the
printed parts.
I wouldn’t expect that small a difference to cause this sort of grief, but it’s
hard to be sure.
if you can post a video of it moving and measuring the belts, it may help
(sometimes it’s easier to see a slack belt or something catching on video than
in real life, especially if you use fast playback)
David Lang
I tried to load a video of the calibration cycle earlier in this thread. I can only load about 5 seconds. I have set my video quality to the lowest setting and it ends up being over 10 Mgbt if its longer than 5.
Easiest way of beating that is to create a utube video and just put the link here.
YouTube (https://www.youtube.com/)
Select create up the top. You may have to create an account.
Here are the photos of the anchor point in relation to the spoil board and the measurements for the frame measured from center to center in all directions. My son is uploading the video to youtube and I will post the link when he is done.
Maslow-serial.log (43.5 KB)
Ohhh aaaand. When did the retract all only 3 of the 4 cassettes retracted, did retraction a second time and the 4th retracted. This is the first time this happened.
Bones wrote:
Ohhh aaaand. When did the retract all only 3 of the 4 cassettes retracted, did retraction a second time and the 4th retracted. This is the first time this happened.
I’ve seen that happen, where one arm goes from ‘ok’ to “doesn’t work”, usually
it’s something inside the arm getting gunked up.
David Lang
So I geuss I’ll disassemble this thing for the millionth time and un-gunk the cassette.
I’m a little aggravated at the moment. Ive spent innumerable hours trying to get this thing to just complete a calibration and its just one thing after another. The thought of taking this thing behind the tool shed and tossing it in wood chipper has crossed my mind; at least it would finally make some chips.
It sort of looked like it jerked a bit when it was retracting TL, did it catch on the bottom of the frame momentarily? If so, not a problem, don’t pull it apart yet..
I just had a look at the maslow.yaml file you posted earlier. Is that still the same settings for the Z values? If so, I think they in the wrong order.
The way you have the arms the blZ should be 100, trZ 78, brZ 56 and tlZ 34
Turned up retraction to 725 and the hung belt did not duplicate the fail. Changed the numbers and ran it again, no fitness.
Maslow-serial.log (97.2 KB)
maslow.yaml (6.5 KB)
I changed the values, clicked “SET”, then save, then restart icon. Did the same process and did a power off reset. looks like it didnt update to the new values.
maslow.yaml (6.8 KB)
Should the frame dimensions generated in the serial log match my frame.
Serial frame size is 3120.800 x 2178.20
Actual frame size is 2936.875 x 2406.65 (115 5/8 x 94 3/4)
Converted dimension to mm and averaged anchor height to 94 3/4"
This matches.