15mm error on Bottom Right Axis

[dlang]

Paul Modiano wrote:

@dlang

did you try grounding the encoder pin?

No. I don’t know how to do that. Are there any instructions you can point me to?

Soldering Floating Encoder Board Pin

There was a request for a more clear and easy to find set of instructions for how to solder the encoder boards (the previous answer was buried in a thread and not easy to search for or find) so this is duplicate information but in a more searchable way. The issue is that the encoder chips have a pin which can be used to set the direction that the encoder reads (IE is rotating counter clockwise positive or clockwise positive). We left this pin unconnected (our mistake) which seems to under some …

@dlang

As the Z axis moves away from the workpiece, the distance from the anchors to
the arms increases (the triangle of the x/y and Z is getting bigger), so the
motors spool out some belt to allow for this.

The xy movement that I witnessed when I manually told it to move z by 1mm was
far too great to have been compensation for a change of 1mm to the XYZ
triangle. Is there a setting somewhere that would allow me to adjust the
level of compensation that xy makes to a z movement?

it just calculates the desired belt length based on the x/y/z triangle from the
anchor point.

If the Z was not all the way down when you did the calibration, or the Z stop
value is not corred (or was not correct when you did the calibration), then the
anchor locations that were calculated are not correct and you will have errors

@dlang

If the belts remain tight all the time, this should not result in any movement
of the sled, but if you have a bad calibration and your belts are at all lose,
this could result in sled movement.

I have not been able to complete a calibration in a long time. It takes a long while but inevitably ends with WARNING FITNESS TOO LOW. DO NOT USE THESE CALIBRATION VALUES!

I think you may need to troubleshoot this more. what is the fitness that it
calculates? (the default cutoff was 0.5, I think bar lowered it to 0.45)

If your anchors move, if the Z offset values in the config don’t match the
reality of your machine, if the belts aren’t tight as you calibrate, those could
all result in bad anchor coordinates (i.e. a bad calibration)

you may also want to do a manual calibration check and see how close the result
is to what the auto calculated one gives you.

Maslow 4 manual configuration (onshape doc)

note that if the maslow thinks distances are X and your tape measure says they are Y, the maslow is correct (at least as far as getting good cuts are concerned) since the current calibration gives different coordinates on different runs of up to a couple of mm, that seems in the same ballpark as the errors of manually reading a tape measure, so I threw together this onshape doc that lets you enter measurements and it shows the frame and calculates the coordinates. I have you measure both diago…

some people have had better results with the manual version, but it depends
HEAVILY on how accurate your measurements are.

David Lang

[pmodiano]

@bar

The way to check for magnet slipping is to do Retract All → Extend All ->Retract All again and look at the values that are printed out. Ideally all the numbers will be close to zero

[MSG:INFO: Top Left pulled tight with offset -0.032]
[MSG:INFO: Bottom Right pulled tight with offset -0.021]
[MSG:INFO: Bottom Left pulled tight with offset -0.043]
[MSG:INFO: Top Right pulled tight with offset -0.032]
[MSG:INFO: Extending all belts]
[MSG:INFO: All belts extended to center position]
[MSG:INFO: Retracting all belts]
[MSG:INFO: Top Right pulled tight with offset 0.064]
[MSG:INFO: Top Left pulled tight with offset 0.032]
[MSG:INFO: Bottom Left pulled tight with offset 0.043]
[MSG:INFO: Bottom Right pulled tight with offset 0.000]

What does that tell you?

[bar]

Unfortunately that tells me that the magnets all look good.

How does the frame rigidity look? Is there any flex?

When the calibration process is running do each of the belts get tight at each measurement? Could they be getting caught on something?

[dlang]

Paul Modiano wrote:

@bar

The way to check for magnet slipping is to do Retract All → Extend All ->Retract All again and look at the values that are printed out. Ideally all the numbers will be close to zero

[MSG:INFO: Top Left pulled tight with offset -0.032]
[MSG:INFO: Bottom Right pulled tight with offset -0.021]
[MSG:INFO: Bottom Left pulled tight with offset -0.043]
[MSG:INFO: Top Right pulled tight with offset -0.032]
[MSG:INFO: Extending all belts]
[MSG:INFO: All belts extended to center position]
[MSG:INFO: Retracting all belts]
[MSG:INFO: Top Right pulled tight with offset 0.064]
[MSG:INFO: Top Left pulled tight with offset 0.032]
[MSG:INFO: Bottom Left pulled tight with offset 0.043]
[MSG:INFO: Bottom Right pulled tight with offset 0.000]

What does that tell you?

that looks good. If the numbers were larger, that would indicate a problem. It
doesn’t hurt to do it a few more times.

David Lang

[pmodiano]

@bar

When the calibration process is running do each of the belts get tight at each measurement? Could they be getting caught on something?

They each get tight and they are not getting caught on anything.

@bar

How does the frame rigidity look? Is there any flex?

It is not a standard frame, but I don’t detect flexing. Two of its corners are on hinges connected to my shed. The plywood is screwed to the two center beams to minimize diagonal flexing.

When not in use, I take out one of the center beams and fold it upward, over the rshed’s roof and out of the way.

Is there any possibility that a recent firmware update is causing z movement to unduly move x and y, causing this issue?

[bar]

I’m almost certain that the issue is going to be frame flex there. It needs to be really really rigid to get good results

[dlang]

@bar

How does the frame rigidity look? Is there any flex?

It is not a standard frame, but I don’t detect flexing. Two of its corners are on hinges connected to my shed. The plywood is screwed to the two center beams to minimize diagonal flexing.

setup a tripod on each corner and take a video of the corner anchor at close
range, then play back on fast forward to see if it moves.

Is there any possibility that a recent firmware update is causing z movement to unduly move x and y, causing this issue?

no, nothing has changed in that calculation.

David Lang

[pmodiano]

I understand that you believe that frame flex is likely the cause of my failure to complete a calibration. I will be able to test that this weekend, when a regular Maslow frame will become available for me to test it on.

What is the most likely cause of my now-consistent 15mm error problem? I send it the job, it starts pocketing and consistently fails on the third pocket with this:

[MSG:INFO: Using spindle NoSpindle]
[MSG:WARN: Position error on Top Left axis exceeded 15mm while running. Error is 15.012mm Counter: 1]
[MSG:WARN: Previous error was 15.012mm]
[MSG:WARN: Position error on Top Left axis exceeded 15mm while running. Error is 15.048mm Counter: 2]
[MSG:WARN: Previous error was 15.048mm]
[MSG:WARN: Position error on Top Left axis exceeded 15mm while running. Error is 15.084mm Counter: 3]
[MSG:WARN: Previous error was 15.084mm]
[MSG:WARN: Position error on Top Left axis exceeded 15mm while running. Error is 15.111mm Counter: 4]
[MSG:WARN: Previous error was 15.111mm]
[MSG:WARN: Position error on Top Left axis exceeded 15mm while running. Error is 15.138mm Counter: 5]
[MSG:WARN: Previous error was 15.138mm]
[MSG:WARN: Position error on Top Left axis exceeded 15mm while running. Error is 15.166mm Counter: 6]
[MSG:WARN: Previous error was 15.166mm]
[MSG:ERR: Emergency stop! Stopping all motors]

[bar]

The 15mm error is caused by a connection issue between the encoder and the main controller board. The hot glue can help, but the real solution is the new encoder cables that we are working on.

[Direkt]

Just curious as to what conductors are being used in the Cat 5 cable from board to encoder? Would it be possible to hardwire (solder) one end ? Most likely for me would be the encoder board side. Just a thought.

[Beagle]

I am doing the exact same thing,there are 4 wires (5 if you add a sheild) 3.3v,Gnd,SDA and SCL.
for more details take a look at the Source here GitHub - MaslowCNC/Boards: The PCB design files for the Maslow4
Schematic_Encoder-Board.pdf

[Dano]

@bar

This issue is usually caused by one of the magnets rotating in the roller or less commonly by frame flex.

The way to check for magnet slipping is to do Retract All → Extend All ->Retract All again and look at the values that are printed out. Ideally all the numbers will be close to zero

This might be an opportunity for improvement. Im thinking sort of like calibration utility but instead for magnet check / frame flex utility tool to help determine this ?

[Andith]

@Dano

Im thinking sort of like calibration utility but instead for magnet check / frame flex utility tool to help determine this ?

Under “Config” there is the “Test” command. That won’t check it the magnet is slipping, but it will check whether or not the magnet is being detected in general.

[dlang]

Dano wrote:

This might be an opportunity for improvement. Im thinking sort of like
calibration utility but instead for magnet check / frame flex utility tool to
help determine this ?

@bar did make a version at one point that would test for fame flex.

to test for frame flex, you pull lightly to tighten the belt, then you apply
more force and see how much the belt lengths change.

for magnet slippage, you need to extend the belts and then retract them to a
known physical point. so the normal extend/retract cycle will show this.

David Lang