Slack belts after calibration

Troubleshooting
1

Have had 5 or 6 successful calibrations now, all with varying fitness levels and done with varying grid sizes and calibration forces to try and see if anything changes. They’ve all resulted in different coordinates for the anchors too, some more than others.

The issue is that after the calibration when i jog it around or run a cut file, the BL and BR belts go slack whenever the maslow is near the centre of the X axis and high up on the Y axis. I assume its just a poor calibration and the anchor coordinates are bad but surely this isn’t how its supposed to be, it will be no good to cut with two slack belts. my setup details are as follows
-Horizontal
-2400 x 1200 cut area
-3300 x 2400 anchor spacing
-Latest firmware
-Have tried manual calibration and it was even worse
-Have retracted, extended and then retracted and the offsets were zero or close to zero
-Smallest calibration grid was 1500 x 1000, largest was 2000 x 1200, smallest force was 900, largest was 1300

I already have a new control board on the way, any help or guidance is appreciated

2

Do you have a set of anchor coordinates you trust (e.g. the manual measurement)? If yes, you could try multiplying the coordinates with 0.9995 and see if that improves the belt tension. If that helps and if the belts are still too loose, you can repeat that step.

Attention: This increases belt tension everywhere. So if the bottom belts are tight when you are in the lower part of the frame, it might not be the solution to your problem and might cause a belt to snap.

Explanation: If Maslow believes the frame is larger than it actually is, it releases more belt than necessary. This might also happen if pitch of the belts is not exactly 2mm. By telling Maslow that the frame is smaller, you force it to release less belt, causing the tension to go up.

3

what version of firmware are you on, and did you have the router all the way
down when calibrating?

as noted, check if the calibrated coordinates are plausable with manual
measurements.

David Lang

4

Yeah i saw your thread about manual calibration, Ill try reducing my manual numbers and see if that helps

Latest firmware and z axis was all the way down and z offsets were correct. What is considered plausible though, within 5mm, 1mm?

5

This also sounds to me like there could be some frame flex going on. That’s the point at which the frame is under the most force

6

Deadset wrote:

What is considered plausible though, within 5mm, 1mm?

I’m not sure, but if it’s more than low single digit mm I would be concerned.

David Lang

7

I thought the same thing, i have wooden corner boards bolted to my floorand the anchors bolted to the boards and was using the 3d printed anchors but have since swapped them out for new custom anchors, it doesnt seem likely to be flexing but ill investigate further. The flex values at the start of calibration change every time, from around 2.5mm to 0.2mm

8

I cranked the calibration force up to 2000 and it was one of the smoothest calibrations ive had. Calibration area was 1800 x 900 and 9x9 grid. Calculations went pretty quick and now theres no belt slack anywhere. Preliminary tests of accuracy are good too, approx 1mm error across the x axis from what ive checked so far.

The wild part is that some of the anchor coordinates are heaps different to my manual measured values, for example tlY is 18mm lower, and some are exactly the same within 1mm.

Ive checked my encoders by doing the retract - extend - retract process and they were bang on too

1 Like