If you want to attach your maslow.yaml file I can edit and send it back to do the smaller grid
Then lets try:
power off → power on → retract all → extend all → run calibration
That should give us an answer in a couple of minutes rather than waiting for the full process
I’ll be right back online, I’m going to head to the office to see if I can replicate the issue that you are seeing.
It sounds like the only difference between what you are doing and what I am doing is that I’m turning it off and back on before starting so I’m going to try not doing that and see if I can get it to not work.
1 Like
So I made it through the 2x2 grid. So the console lists the proposed points. When it travels to them it then says “waypoint”
[MSG:INFO: Point: 0 (-754.500, -405.500)]
[MSG:INFO: Point: 1 (-754.500, 405.500)]
[MSG:INFO: Point: 2 (754.500, 405.500)]
[MSG:INFO: Point: 3 (754.500, -405.500)]
[MSG:INFO: Measured waypoint 0]
[MSG:INFO: Measured waypoint 1]
[MSG:INFO: Measured waypoint 2]
[MSG:INFO: Measured waypoint 3]
[MSG:INFO: Calibration complete]
The major difference in the console with the 90 points it moved though the 90 but never said “Measured Waypoint” for any
So I think it lost contact or had a bug and ran without reporting positions for the 90 points and actually nothing was being recorded. The error or loss of connection would have been right at the beginning of calibration.
I’ll get you that yaml file that was having errors yesterday too.
How big a grid should I calibrate with to have a useful setup?
thanks
1 Like
Did you get an output of the anchor point locations at the end?
Yes!
Maslow_tlX: -6.362284637490082
Maslow_tlY: 2186.914666584596
Maslow_trX: 3113.7865153243297
Maslow_trY: 2186.137286983873
Maslow_blX: 0
Maslow_blY: 0
Maslow_brX: 3125.5037190104667
Maslow_brY: 0
I put them in the yaml file and uploaded it and hit restart.
The machine is sitting at the last calibration point with belts tight. I’m not sure what is next? If I try too jog is it says
MSG:ERR: Position error on Bottom Left axis exceeded 1mm, error is -1926.500mm]
1 Like
This is excellent progress, great work!
I think that just four points close together on the sheet really isn’t going to give us enough information to get good calibration results…but now we know that the process works.
I think that the next step is to do the same exact thing again with more points.
That is a FANTASTIC question. Nobody knows! That’s something that we’re going to have to do a lot of experimentation to figure out.
I would go through the whole power off → power on → retract all → extend all → take slack process now to reset everything.
That being said I’m pretty sure that 4 points is going to be too few to have accurate results.
Any idea what the difference was (so that I can fix it and nobody else has to deal with this)? Was it the turning it off and on?
This is excellent progress, great work!
I think that just four points close together on the sheet really isn’t going to give us enough information to get good calibration results…but now we know that the process works.
the end result of calibration reports where it thinks the 4 anchors are, please
record that for each attempt here.
I think that the next step is to do the same exact thing again with more points.
I would also like to see a 4 point calibration where the points are out near the
edge of the workpiece.
That is a FANTASTIC question. Nobody knows! That’s something that we’re going to have to do a lot of experimentation to figure out.
by recording where it thinks the ancors are each test, you can tinker with the
number of points and the spacing and see at what count/spacing the anchor points
stop changing. That should be a ‘good enough’ level of calibration.
I suspect that getting the testing points to cover as large an area as possible
will be more important than the number of points tested (having the points
further from each other allows measurement errors to stack up and makes it
easier to detect them)
David Lang
1 Like
So with the 90 points I’m pretty sure communication was lost. The wifi here hangs out around 2 bars on a phone and sometimes croaks. I guess from your perspective you would want to look at what happens at the start of calibration. Those 90 points were computed, the console printed the 90 points, they must have been sent to the machine since it started moving to point 1 and tightened the belts, paused for 30secs or so and moved to point 2 and so on. BUT it never printed “Measure Waypoint…” in the console. It did carry on through the 90 points tightening the belts and all but never said “Measure Waypoint …” at the end it said “ok”. Console never updated.
I’ll run it with a 3 x 6 with less offset and see what happens
1 Like
3 x 6 grid failed. Here is the whole console. Same behavior.
Steps:
Load yaml with 3x6 and dims from 2x2 calibration > reset FLuidNC > power off>power on>new browser tab> connect to maslow.local > retract > extend >calibrate
then this is the whole console from restart, and it never changes while the machine goes through all the points loosening and tightening the belts correctly
Serial Messages
[GC:G0 G54 G17 G21 G90 G94 M5 M9 T0 F0 S0]
$/axes/x/max_travel_mm=2438.400
$/axes/x/homing/mpos_mm=0.000
$/axes/x/homing/positive_direction=true
$/axes/y/max_travel_mm=1219.200
$/axes/y/homing/mpos_mm=0.000
$/axes/y/homing/positive_direction=true
[MSG:INFO: Channel auto report interval set to 50 ms]
[GC:G0 G54 G17 G21 G90 G94 M5 M9 T0 F0 S0]
[MSG:INFO: Retracting all belts]
[MSG:INFO: 0 pulled tight with offset -918.708]
[MSG:INFO: 1 pulled tight with offset -1579.364]
[MSG:INFO: 3 pulled tight with offset -2080.556]
[MSG:INFO: 2 pulled tight with offset -2594.987]
[MSG:INFO: Retracting all belts]
[MSG:INFO: 1 pulled tight with offset -0.011]
[MSG:INFO: 0 pulled tight with offset -0.054]
[MSG:INFO: 2 pulled tight with offset -0.064]
[MSG:INFO: 3 pulled tight with offset -128.844]
[MSG:INFO: Retracting all belts]
[MSG:INFO: 0 pulled tight with offset 0.000]
[MSG:INFO: 1 pulled tight with offset 0.000]
[MSG:INFO: 2 pulled tight with offset 0.000]
[MSG:INFO: 3 pulled tight with offset -12.658]
[MSG:INFO: Extending all belts]
[MSG:INFO: All belts extended to center position]
[MSG:INFO: Point: 0 (-1206.893, -523.069)]
[MSG:INFO: Point: 1 (-1206.893, 0.000)]
[MSG:INFO: Point: 2 (-1206.893, 523.069)]
[MSG:INFO: Point: 3 (-724.136, 523.069)]
[MSG:INFO: Point: 4 (-724.136, 0.000)]
[MSG:INFO: Point: 5 (-724.136, -523.069)]
[MSG:INFO: Point: 6 (-241.379, -523.069)]
[MSG:INFO: Point: 7 (-241.379, 0.000)]
[MSG:INFO: Point: 8 (-241.379, 523.069)]
[MSG:INFO: Point: 9 (241.379, 523.069)]
[MSG:INFO: Point: 10 (241.379, 0.000)]
[MSG:INFO: Point: 11 (241.379, -523.069)]
[MSG:INFO: Point: 12 (724.136, -523.069)]
[MSG:INFO: Point: 13 (724.136, 0.000)]
[MSG:INFO: Point: 14 (724.136, 523.069)]
[MSG:INFO: Point: 15 (1206.893, 523.069)]
[MSG:INFO: Point: 16 (1206.893, 0.000)]
[MSG:INFO: Point: 17 (1206.893, -523.069)]
last thing is a “ok” at the end which was there before it started moving
1 Like
This is 100% a bug. I think that your observation that we get
When things are working right is excellent. That seems to be what we are missing. I am going to dig into it, but I have COVID at the moment and I have the brain power of a potato so it might take me a couple days to track it down
Can you attach your maslow.yaml file which is not working and I’ll try to replicate it?
I hope you feel better soon!
I ran it with 3x5 and it record the first point correctly!
[MSG:INFO: Point: 13 (1056.893, 0.000)]
[MSG:INFO: Point: 14 (1056.893, 523.069)]
[MSG:INFO: Measured waypoint 0]
But then the long belt to the other lower corner got stuck on the frame and in yanking on it to free it I must have over done the amps and it shut down. so need to start again.
This all reminds me of early 3d printing days - that ended up with a display on the printer and the self zeroing tech. This is similar in that we’re trying to get a calibration down the end of a one dimensional work flow, and maintaining that calibration with an endlessly perfect sensor read and stepper motor feed.
For this setup any ability to get in there and restart, rezero one leg, use the last known length plus a safety factor, anything to not have to go back to taking it off the frame and doing the belt retract would be a big help.
I wonder if you could have a spoil board sensor, either a reflection led type or a micro switch. Basically when that known point of the sled is no longer on the spoil board you know you’re at that edge. Probably need 4 sensors, one for each board edge so the sled doesn’t have fall off the edge to find it. User measures the spoil board and I’m guessing the math can be done on locating the anchor points
just a a thought!
1 Like
here is the yaml that gave the error 152. I hope this is the correct one! it must have a wierd character in it
maslow-trash-2.yaml (2.0 KB)
1 Like
The problem is that the zero for a belt is fully retracted. it measures relative
movement from that point, so to re-do a leg you would need to retract that leg
fully and feed it out again to get a known length on that belt.
David Lang
100% I agree. Let’s just hope it only takes us a few weeks not years to get it figured out 
Again I 100% agree.
I will give this a run tomorrow and see how it looks.
The more that I think about it the more that I think that you are right and there is some sort of connection issue.
This message appears right at the beginning of the calibration process to let us know that all of the points that it intends to go to have been calculated
This message lets us know that we have actually successfully gone to that point and taken a measurement
How 200% are you that you have version 0.65? It came out on Wednesday so Monday/Tuesday would be 0.64. It might be worth updating again just to be 200% sure since it’s free and 0.65 has a fix for exactly this issue (not to say that there couldn’t still be bugs). Don’t forget to grab all three files since there are changes to all of them.
Here is what I am going to do tomorrow:
I am going to add a heartbeat so that if the connection to the machine is dropped we get a warning message about that.
Test to better understand what happens if my WiFi network is turned off during the calibration process
And I am going to try running the .yaml file which is giving you trouble to see if we can figure out why it is doing that.
2 Likes
Add something to the calibration output that shows the firmware version so there
isn’t any question about that.
David Lang
3 Likes