4.1 Calibration Issues and Questions

[Dano]

Something like.

Failed to find reasonable calibration values.
Automatic retry has started.

Dano

[dlang]

boslaw wrote:

20+ mins of restarting fitness… no change

Ok, one thing that looks wrong to me is that I thought that if it re-started
calibration, it re-measured as well, not just running the same numbers through
the calculations again.

There are some folks here who have a spreadsheet that will visualize the points

CLBM:[{bl:2268.54, br:2108.93, tr:2285.73, tl:2284.08},{bl:2213.02,
br:2187.33, tr:2185.05, tl:2281.47},{bl:2318.70, br:2096.65, tr:2077.11,
tl:2298.02},{bl:2379.22, br:2210.13, tr:2115.00, tl:2184.81},{bl:2281.08,
br:2320.25, tr:2158.00, tl:2075.50},{bl:2167.51, br:2400.47, tr:2261.38,
tl:1980.24},{bl:2270.27, br:2277.70, tr:2290.86, tl:2108.17},{bl:2166.46,
br:2176.67, tr:2226.30, tl:2216.25},{bl:2057.97, br:2102.23, tr:2349.83,
tl:2307.92},{bl:2012.11, br:2213.42, tr:2397.78, tl:2195.54},{bl:2097.48,
br:2306.63, tr:2299.02, tl:2087.22},{bl:2214.86, br:2387.68, tr:2186.30,
tl:2070.21},{bl:2326.72, br:2288.49, tr:2077.40, tl:2118.51},{bl:2378.39,
br:2176.05, tr:2012.20, tl:2237.69},{bl:2279.66, br:2067.70, tr:2118.78,
tl:2324.23},]

the fitness is increasing, slowly (starts 0.01 ends 0.04)

again, please measure the anchor distances (inc diagonals)

David Lang

[boslaw]

173 inches diagonal (4394.2mm) (I double checked with tape measure)
8x12’

[boslaw]

@bar

That message has been in the firmware since before it could automatically try again if it failed to find a solution. You are absolutely right that it needs to be updated.

Does anyone have a suggestion for what it should say to communicate that it failed to find a solution, but it will automatically try again?

I think part of my issue is that I keep seeing in the forums that people are “stuck in fitness failed” loop so I immediately assumed that my machine would never calibrate once I saw that loop start. I think any documentation should explain that this is not necessarily a problem and that the. machine should eventually calculate calibration values (although this has not happened for me).

[boslaw]

@dlang

the caclulations are done in your browser, so the computer/browser performance
can make a big difference in how long it takes to compute

I have a top of the line MacBook Pro M2 - if calculations are based on computer processing power and speed mine should calculate quite quickly.

[dlang]

boslaw wrote:

173 inches diagonal (4394.2mm) (I double checked with tape measure)
8x12’

we need all 6 measurements with a tape measure to at least 1mm of accuracy (if
you think you can do better, that’s fantastic)

David Lang

[boslaw]

@dlang

we need all 6 measurements with a tape measure to at least 1mm of accuracy (if
you think you can do better, that’s fantastic)

@dlang
143 15/16" (bottom) - 3656.01mm
143 9/16" (top) - converted to 3646.49mm
96 1/16" (right) - converted to 2439.99mm
96 2/16" (left) - converted to 2441.57mm
172 12/16" (bottom left to top right) - converted to 4387.85mm
172 13/16" (bottom right to top left) - converted to 4389.44mm

Note that in one of my log files, the calculated measurements seem to be pretty close to my manual tape measure measurements (assuming I’m dividing the diagonals in half to get the numbers Maslow calculates below - orientation of the Maslow was different in this log than my manual measurements so the numbers won’t correspond perfectly):
[MSG:INFO: Bottom Right 2281.542]
[MSG:INFO: Bottom Right 2281.477]
[MSG:INFO: Bottom Right 2281.466]
[MSG:INFO: Bottom Right 2281.466]
[MSG:INFO: Top Right 2185.055]
[MSG:INFO: Top Right 2185.055]
[MSG:INFO: Top Right 2185.044]
[MSG:INFO: Top Right 2185.044]
[MSG:INFO: Top Left 2213.030]
[MSG:INFO: Top Left 2213.030]
[MSG:INFO: Top Left 2213.019]
[MSG:INFO: Top Left 2213.019]
[MSG:INFO: Bottom Left 2187.284]
[MSG:INFO: Bottom Left 2187.294]
[MSG:INFO: Bottom Left 2187.284]
[MSG:INFO: Bottom Left 2192.011]

[dlang]

ok, the diagonals don’t match by ~1/4" but here is what I come up with for the coordinates based on your measurements:

tlx=0.56
tly=2441.5
trx=3647
try=2440
brx=3656

try putting these in the maslow.yaml (either by editing or through the web menu) and see if the machine moves correctly (without doing an automated configuration)

[boslaw]

@dlang @bar

Failed again.

I realized that the top and bottom numbers I gave you were inverted.

1. I punched your numbers into my yaml (swapping the trx and brx numbers - all others were correct).
2. Delete Maslow.yaml from fluidnc
3. Upload new maslow.yaml via fluidnc
4. reboot Maslow from fluidnc
5. confirm numbers in config (correct)
6. retract all
7. extract all
8. apply tension. get error message about center point deviation. bottom left and upper right belt are loose. Upper left and bottom right are tight.

Can’t do anything else. Jog does nothing except turn on fans.

Tried all of the above without “Apply Tension” at the end - same result. Nothing happens.

Maslow-serial-11.log (3.2 KB)

After trying this a few times with the manual numbers, I tried auto-calibration again (what else do I have to lose at this point). Calibration failed, restarting fitness loop running for 25 mins so far - doesn’t seem like that will solve this either.

[dlang]

swapping the top and bottom numbers, the coordinates are

tlx=-9
tly=2441.5
trx=3647
try=2440
brx=3646.5

that 9mm shift to the left would make a couple of the belts slack.

David Lang

[boslaw]

Ok, I’ll try that but they are REALLY slack, not 9mm slack.

Auto calibration still running after 40 mins. Clearly not going to complete.

[boslaw]

Tried your new numbers. Failed again.

Removed spoil board to run on bare ground like @caddad did.

Retract All - fitness numbers all good
[MSG:INFO: Bottom Left pulled tight with offset -0.011]
[MSG:INFO: Top Right pulled tight with offset -0.032]
[MSG:INFO: Top Left pulled tight with offset -0.086]
[MSG:INFO: Bottom Right pulled tight with offset -0.054]

Extend All

Apply Tension - 3 belts tighten. Upper Right belt does not tighten.

[MSG:INFO: Measured waypoint 2]
[MSG:INFO: Center point deviation: TL: 0.000 TR: 0.000 BL: 96.271 BR: -15.049]
[MSG:ERR: Center point deviation over 12.000mm, your coordinate system is not accurate, maybe try running calibration again?]

Can’t do anything.
Maslow-serial (1).log (2.3 KB)

My frame is out by ~ 1/2 inch. I can’t believe everyone else using this thing has a perfectly square frame down to .1mm so the problem must lie somewhere else.

Anyone want to buy a beautiful unused assembled Maslow 4.0? I’ll sell it for the unassembled price. No guarantees you’ll ever be able to get it to work

[dlang]

with one belt not tensioning, increase the retraction force limit (the default
is 900, with concrete anchors, you should be safe up to 2000 or so, but will
probably work with just a little higher)

David Lang

[boslaw]

@dlang You really think that is the problem? It retracts nice and tight all other times. When I Retract All the belts all pull tight with good offset values. It just doesn’t retract one belt all the way when I’m trying to use it after extending all belts and attaching to my anchors or at the end of calibration. I’m at 1500 retract force already.

@bar any other ideas? I’ve tried everything possible with this thing. It simply does not work.

[bar]

@boslaw

• apply tension. get error message about center point deviation. bottom left and upper right belt are loose. Upper left and bottom right are tight.

I agree with @dlang that there is something funky going on with how the machine is taking measurements. After you apply tension the bottom left and right belts are still lose?

[boslaw]

@bar

When I retract all, all belts retract perfectly. Fully tight to the frame. My offset values are good. Even after I extend all the way and retract without doing anything else, my belts retract back perfectly.

However, after I extend all and connect the belts to the frame (all still loose), I press Apply Tension. Upper Right belt stays loose but the others pull tight against the frame.

Yesterday, I tried the same thing with auto-calibration but in that case the Lower Left and Upper Right stayed loose.

I have tried and retried and retried and measured and inserted manual measurements. I need some other suggestions because what has been suggested is not working.

[caddad]

So I went to recalibrate at the working height (on the foam) and with horizontal orientation selected this time, and one of the belts unwound and got tangled and mangled beyond repair in the process. I was watching it, but it happened too fast to rescue. So, I’m not able to use the machine.

[md8n]

@boslaw

I need some other suggestions because what has been suggested is not working.

One thing I would add is to be absolutely certain that the little magnets are properly glued in and that they are positioned immediately above the hall effect sensor chips. In other words, check for “dumb things”

@boslaw

Upper Right belt stays loose but the others pull tight against the frame.

Apply tension works from the dimensions that you currently have in the maslow.yaml - If they’re off (too ‘large’) then one or two arms may have loose belts

And if you haven’t tried this already - try swapping the arms around to see if the problem moves with the arm or not.

[md8n]

@caddad

with horizontal orientation selected this time

Considering your set up you should only be using this value.

@caddad

and one of the belts unwound and got tangled and mangled beyond repair in the process.

Yup - that can happen, make sure to pull the arm completely apart and check everything.

For the first part of the calibration cycle, the 3x3 grid, you need to help pull the belts out as they are unspooled.

[caddad]

@ian_ab suggested to use horizontal in an earlier reply, hence my restarting.

There were no instructions to assist the unwinding. @bar can I get a replacement belt? I already disassembled the arm.