Running Safari, or what? And yes, I know they all have to be webkit based for iOS/iPadOS
1 Like
can you rotate the control board so that things will reach?
The motor control and the ethernet should be next to each other, so if the
ethernet won’t reach, then the motor isn’t going to be plugged in the right
place either.
David Lang
I just checked mine and mine are wrong too 
It’s a bug for sure, but not the issue that we are looking for.
Funnily enough I remember fixing this not too long ago…I guess I didn’t fix it somehow
Ok, darn I was hoping i was onto something
1 Like
2 Likes
Didn’t get to it last night, flashed this morning and took measurements.
With the power cable facing me and measuring with a tape measure.
- Top Left 1,010mm
- Top Right 1,020mm
- Bottom Right 1,010mm
- Bottom Left 1,050mm
GRBL is reporting
[MSG:INFO: Top Left position is 1009.911]
[MSG:INFO: Top Right position is 1027.186]
[MSG:INFO: Bottom Left position is 1010.813]
[MSG:INFO: Bottom Right position is 1009.793]
This seems like it could be something important. That is quite a bit off.
Is that a repeatable error?
No.
Second time measured ~1,010mm on bottom left [MSG:INFO: Bottom Left position is 1010.781]
Third time measured ~1,010mm on bottom left [MSG:INFO: Bottom Left position is 1011.973]
1 Like
Another calibration 4(x) by 6(y) covering most of the spoil area. Horrid results. Only one waypoint went a little wonky (you will see an overcurrent in the log). at this point the lower right did not retract and the lower left did, moving the maslow to the left significantly.
excerpt:
[MSG:INFO: All belts extended to center position]
[MSG:INFO: Measured waypoint 0]
[MSG:INFO: Center point deviation: TL: 0.196 TR: 0.120 BL: -71.092 BR: -96.221]
[MSG:ERR: Center point deviation over 15.000mmm, your coordinate system is not accurate, maybe try running calibration again?]
[MSG:INFO: Point: 0 (0, 0)]
[MSG:INFO: Point: 1 (0, -1129.000)]
[MSG:INFO: Point: 2 (-519.000, -1129.000)]
[MSG:INFO: Point: 3 (-519.000, -677.400)]
[MSG:INFO: Point: 4 (-519.000, -225.800)]
[MSG:INFO: Point: 5 (-519.000, 225.800)]
[MSG:INFO: Point: 6 (-519.000, 677.400)]
[MSG:INFO: Point: 7 (-519.000, 1129.000)]
[MSG:INFO: Point: 8 (-173.000, 1129.000)]
[MSG:INFO: Point: 9 (-173.000, 677.400)]
[MSG:INFO: Point: 10 (-173.000, 225.800)]
[MSG:INFO: Point: 11 (-173.000, -225.800)]
[MSG:INFO: Point: 12 (-173.000, -677.400)]
[MSG:INFO: Point: 13 (-173.000, -1129.000)]
[MSG:INFO: Point: 14 (173.000, -1129.000)]
[MSG:INFO: Point: 15 (173.000, -677.400)]
[MSG:INFO: Point: 16 (173.000, -225.800)]
[MSG:INFO: Point: 17 (173.000, 225.800)]
[MSG:INFO: Point: 18 (173.000, 677.400)]
[MSG:INFO: Point: 19 (173.000, 1129.000)]
[MSG:INFO: Point: 20 (519.000, 1129.000)]
[MSG:INFO: Point: 21 (519.000, 677.400)]
[MSG:INFO: Point: 22 (519.000, 225.800)]
[MSG:INFO: Point: 23 (519.000, -225.800)]
[MSG:INFO: Point: 24 (519.000, -677.400)]
[MSG:INFO: Point: 25 (519.000, -1129.000)]
[MSG:INFO: Measured waypoint 0]
[MSG:INFO: Measured waypoint 1]
...
Fitness: 0.06669144107777232 in 15300 cycles
WARNING FITNESS TOO LOW. DO NOT USE THESE CALIBRATION VALUES!
Calibration complete
Calibration values:
Maslow_tlX: 92.9
Maslow_tlY: 3638.7
Maslow_trX: 2450.6
Maslow_trY: 3574.0
Maslow_blX: 0.0
Maslow_blY: 0.0
Maslow_brX: 2425.9
Maslow_brY: 0.0
maslow.txt (11.3 KB)
1 Like
I’m wondering if I ought to try making lower top anchors (~4ft down) off the the sides of my frame such that I have more of a “square” area and see what my results are with that to see if my hunch is right, which is that I really suspect that portrait orientation for a vertical frame is the biggest issue here, although I really am just guessing 
1 Like
I think that is an excellent idea for what to try next. I’m positive that you and @Frank_Wiebenga are seeing the same issue…I have truly no idea what could be causing it, but that was next on my list of things to think to try. I am wondering if maybe the calibration process has a bug in the math and certain ratios of length to width are breaking it somehow?
You both have the exact same symptoms. It is 100% a bug. I just have no clue where.
I’ve been over his machine pretty carefully and I’m pretty certain that there is nothing physically wrong with it. It’s either a configuration setting somewhere that is somehow messing things up, or the frame so I think trying to change the frame is a great idea.
I’m about to glue a 2nd set of anchors to my floor as an alternate location for Maslow while my composites table is in use. Is there any particular dimensions / orientation I could use to assist the troubleshooting?
1 Like
I love and appreciate the willingness to help us experiment!
At this point I have truly no idea so I think that any additional data points would help.
what exactly is the frame size you are using?
David Lang
Tried again with roughly 8’x8’ with latest firmware. (I moved my top bolts down 4’). Still crappy results. attached is serial console. 6x6 calibration of 1m square.
I noticed 3 points during the calibration had issues. #3 and #15 moved significantly off where I’d think the grid should be during the belt retraction step. #23 the lower right spool never retracted at all and it just skipped onto the next one.
Settings:
board: Maslow
name: Maslow S3 Board
meta:
Maslow_vertical: true
maslow_calibration_grid_width_mm_X: 1000.000000
maslow_calibration_grid_width_mm_Y: 1000.000000
Maslow_calibration_size_X: 6
Maslow_calibration_size_Y: 6
Maslow_brX: 2438.000000
Maslow_brY: 0.000000
Maslow_brZ: 117.000000
Maslow_tlX: 0.000000
Maslow_tlY: 2438.000000
Maslow_tlZ: 144.000000
Maslow_trX: 2438.000000
Maslow_trY: 2438.000000
Maslow_trZ: 97.000000
Maslow_blX: 0.000000
Maslow_blY: 0.000000
Maslow_blZ: 75.000000
Maslow_Retract_Current_Threshold: 1700
Maslow_Calibration_Current_Threshold: 1500
maslow.txt (8.2 KB)
1 Like
I have a 12’x8’ frame in portrait, vertical orientation (just downgraded to 8’x8’)
If you want to play around with removing those trouble points from the data and running it again check out: GitHub - BarbourSmith/Calibration-Simulation
You can copy and paste the measurements array into the index.html file and see how it looks. Some of the information is printed to the console.
You would just want to paste in [{bl:1708.84, br:1754.52, tr:1726.76, tl:1729.99},{bl:1386.86, br:1444.74, tr:2109.69, tl:2112.35},{bl:1002.05, br:1887.20, tr:2433.01, tl:1869.03},{bl:1216.34, br:1924.77, tr:2296.00, tl:1688.23},{bl:1326.83, br:2072.22, tr:2168.90, tl:1509.46},{bl:1503.77, br:2188.31, tr:2053.41, tl:1338.15},{bl:1684.55, br:2316.07, tr:1951.73, tl:1176.18},{bl:1872.00, br:2455.70, tr:1865.99, tl:1027.49},{bl:1954.94, br:2320.08, tr:1683.60, tl:1176.38},{bl:1775.17, br:2172.95, tr:1778.21, tl:1308.12},{bl:1601.79, br:2035.69, tr:1889.35, tl:1455.69},{bl:1437.70, br:1910.67, tr:2014.17, tl:1614.37},{bl:1284.08, br:1801.46, tr:2150.46, tl:1781.47},{bl:1145.31, br:1708.23, tr:2296.12, tl:1954.78},{bl:1302.66, br:1533.18, tr:2169.05, tl:2056.28},{bl:1426.66, br:1638.87, tr:2014.22, tl:1892.26},{bl:1567.39, br:1757.68, tr:1868.02, tl:1735.86},{bl:1719.89, br:1893.44, tr:1732.60, tl:1589.33},{bl:1883.73, br:2040.74, tr:1610.68, tl:1455.47},{bl:2052.71, br:2195.50, tr:1505.68, tl:1338.41},{bl:2162.07, br:2082.95, tr:1333.99, tl:1509.56},{bl:1999.51, br:1915.27, tr:1451.66, tl:1614.31},{bl:1848.45, br:1757.68, tr:1585.78, tl:1735.94},{bl:1706.04, br:1795.85, tr:1732.74, tl:1870.92},{bl:1581.97, br:1474.57, tr:1889.40, tl:2017.21},{bl:1473.11, br:1358.10, tr:2053.49, tl:2171.73},{bl:1644.96, br:1197.72, tr:1951.74, tl:2298.68},{bl:1742.92, br:1329.37, tr:1778.18, tl:2153.11},{bl:1858.57, br:1476.18, tr:1610.73, tl:2016.97},{bl:1990.39, br:1636.91, tr:1451.51, tl:1892.26},{bl:2131.54, br:1805.32, tr:1303.58, tl:1781.34},{bl:2282.78, br:1980.37, tr:1171.51, tl:1687.00},{bl:2413.05, br:1892.01, tr:1021.89, tl:1869.02},{bl:2269.13, br:1705.32, tr:1171.40, tl:1954.70},{bl:2136.05, br:1526.98, tr:1334.08, tl:2056.20},{bl:2015.58, br:1355.66, tr:1505.82, tl:2171.54},{bl:1908.83, br:1194.92, tr:1683.68, tl:2298.51},{bl:1819.98, br:1047.48, tr:1866.07, tl:2435.32},] instead of the data that is in there already. The number of measurements isn’t important so you can remove some.
1 Like
I like this! I think it would be super useful to have calibration results saved on the device (maybe SD?) and have a way for people to use that file or have this plotting be part of the maslow tab. I may spend a little time with this code and see what I can come up with; I feel like I’m better at contributing to software than hardware
looks like I have a lot of troubled points…
That’s exactly what I had hoped to do. I think it would be a neat way to visualize things.
You are a hero.
Your results look completely the same as what @Frank_Wiebenga is seeing. There isn’t anything so egregiously wrong that it’s clear where the issue is, but also the points just aren’t quite lining up.
I was really hoping that there was something funky going on in the math that would mean using a different set of anchor points would be different, but it seems like that isn’t the case.
We’ve got to figure out what you two have in common 
I followed your basic frame plans (although I got lazy and used squares instead of your fancy angles on the plywood parts, because I could just batch them out quick on the tablesaw. I used 12’ for the sides (vs 10`) and mounted a 3/4 OSB sheet as my spoilboard (what I used to use for my original maslow). OSB does have a pretty rough texture; I wonder if that is an aspect. I am going to get a sheet of 1/4 luan this weekend to put over top and can try again after that (or maybe not, I think my wife has plans for me this weekend, lol).
I will see what I can figure out on the ui front either way though.
1 Like