Fully Calibrated but Oblong Circles

Troubleshooting
61

Right, I agree. But for the wooden linkage kit that has to be a fixed ideal value. Because the arms are 9" long, the center of the router is at 4.5" from the mouting points, etc.

1 Like
62

Plus the distance from the arm to the chain (i.e., donā€™t forget the clevis)

64

Without recalibrating, you can change that value in the Settings.
Would that change make the difference?

65

Right, so if I draw this up and I assume the clevis adds 1.25" to the overall radius, then the 260mm makes sense. 10.263"

Meaning, my calculated rotation radius is incorrect. Likely a result of my adjustment to my chain pitch. This makes sense.

So my plan of attack tonight will be to;

A. Verify my linkage lengths, if 9" then proceed to

  1. Backup my ini file again,
  2. Set my known measured motor spacing to 3010mm,
  3. Set my ideal rotation radius to 260mm,
  4. Test the up-down accuracy of 6" movements before doing a calibration.
  5. Run a calibration and determine if the accuracy improved or worsened.

What still worries me is that my motor spacing value is not coming out correctly when I run the automatic calibration.

2 Likes
66

Perhaps take a tape measure and measure from the center of the bit to the end of the chain in the clevis and compare to the 260 mm.

67

I have done that, but on my old sled. I always assumed the machine was more accurate than I was, it was closer to around 263mm.

68

I trust your measurements over Maslowā€™s measurements.

2 Likes
69

I agree with that statement now!

70

@bar, is it possible to manually enter values into settings for distanceBetweenMotors, vertical motor offset, rotational radius and then go to calibration and skip all the way to the test cut? Iā€™ve actually never tried doing thatā€¦

2 Likes
71

I have actually been doing that. I still select the chain over the top, the kinematics type and etc. But when it comes to input values, I simply skip those.

I have to be honest but I am surprised that I am the only person experiencing this issue but I really appreciate you guys sticking through this with me. I have had to just walk away a few times due to frustration and total confusion as to how it is so consistently wrong, lol.

I really want to have this sorted out by the end of this Easter weekend. My sonā€™s birthday is coming up and he would like to cut some favors for his friends on the Maslow.

3 Likes
72

Weā€™ll figure it outā€¦ something, somwhere, must be amiss. Alsoā€¦ I havenā€™t tried to cut something so precise, so I donā€™t even know if my Maslow would do it correctly. Most everything Iā€™ve done hasnā€™t been so demanding. Iā€™ve been spending most of my time recently with trying to figure out how to do optical/camera based calibration.

2 Likes
73

You are not alone, I have the same issue you are having - been watching this thread closely to see an outcome. I have the top mounted pantograph from dlang and due to it binding (and not being able to fully calibrate) at extreme far corners, I have been waiting to try the ring system. My circles are ellipses and I have not needed precise parts for the first few things I have made.

4 Likes
74

@Gordo, I am probably sending you on a wild goose chase here but, have taken the time to check if your automatically measured motor distance vs. a tape measure? I am seeing a difference and I questioned my chain pitch.

It would simply be interesting to me to see if you are seeing the same thing.

75

Mine comes in within a mm or so - thereā€™s a little fumbling around on my end with the tape and some clamps, but it comes out close enough - I can get numbers if you like but not till a while later

76

Nah, thatā€™s fine. Mine is off significantly more. On the order of 5mm.

77

@arnoldcp
Canā€™t thank you enough for pulling this through.

11 months ago I tried to tweak numbers +25% -25% to discover directions and percentages of errors. (with brackets chain mount, heavy weight and steep angle)

The numbers then were:

  • Motor distance
  • Vertical height to the sheet
  • Distance of the 2 brackets the chains can pivot
  • Distance of brackets to bit to determinant the tilt on x/y
  • Centre of gravity on the sled to determinant the tilt on x/y

With triangular kits and now we have:

  • Motor distance
  • Vertical height to the sheet
  • Rotation radius
  • Chain sag

Missing or not:

  • Angle of the frame
  • Weight of the sled

Thank you!

P.S. With your persistence, perhaps it could happen that with 1.11 a freeze and global comparison on the same firm/soft and test .nc file gives us global data.

78

Grasping at straws, but a while back, I ran into a problem that led us to discover that the encoderSteps was wrongā€¦ I did a test where I issued a manual command in GC (used the macro) to spool out a distance that should guarantee (assuming everything was correct) that the sprocket would end up at 12 oā€™clock if it started at 12 oā€™clock. You may want to consider doing the same thing as a test.

I started with the first link on a top tooth of the sprocket set to 12 oā€™clock (like you do during calibration to measure distance between motors). I then used the macro on GC to issue a ā€œB09 L635ā€ to spool out 635 mm of chain. Since chain pitch is 6.35 mm per link, this should spool out precisely 100 links and with 10 teeth per sprocket, it should return the key tooth back to 12 oā€™clock.

2 Likes
79

I will do that right now.

1 Like
80

I hope you donā€™t find a problemā€¦

2 Likes
81

Looks good. Ended pointed straight up.

2 Likes