Delta Arm Calibration for Positional Tolerance

Post a reply

This question is a means of preventing automated form submissions by spambots.
:D :) ;) :( :o :shock: :? 8-) :lol: :x :P :oops: :cry: :evil: :twisted: :roll: :!: :?: :idea: :arrow: :| :mrgreen: :geek: :ugeek:

BBCode is ON
[img] is ON
[flash] is OFF
[url] is ON
Smilies are ON

Topic review

Expand view Topic review: Delta Arm Calibration for Positional Tolerance

Re: Delta Arm Calibration for Positional Tolerance

by Lucidwolf » Wed Aug 29, 2018 12:15 pm

On Repitier/Marlin you can set each towers diagonal rod length individually. So you can print a delta tower calibration measure each tower's distance in the X-Y plane use that to calculate a dimensional error to the tower and calculate the corrections for each individual arm. I'm starting to realize if google doesn't find it then it probably isn't an option.
(see for a print that lets you calibrate distance to each tower angle.

Its probably not nearly as important as having the average arm height correct and your tower angle adjustments which the Reprap firmware does a good job getting close on the auto bed probe. I moved on to torturing my extruder (you already know) after I knew i could get good dimensional tolerance.

This is separate from the tower height adjustments in M666.
Now the M500 scares me because the calibration also includes the change to tower angles where I was used to keeping those static based on my calibration print. You are probably right I am just inexperienced with such a smart delta controller. It looks like they are pretty stable each time I do an auto level.

Also the M665 and M666 was the initial setup in the guide. My dyslexia almost cause a head crash. So i made the config match the output of M665. :lol:

I'm used to a stone axe and just bought a light saber so i need to relearn my swing...

Thanks for your response,

Re: Delta Arm Calibration for Positional Tolerance

by mhackney » Wed Aug 29, 2018 9:30 am

Lucidwolf, I don't follow your question.

You can get all of the details at

M665 configures most of the delta parameters. This is mine. Diagonal rod length is what you are calling "arm length".

Code: Select all

M665 L360.310 R210.650 H440.00 B140.0 X0.387 Y0.235 Z0.000 ; Set delta radius, diagonal rod length, printable radius and homed height
Now, what do you mean by "My issue is tower X and Z are close to the same required distance, but Y needs to be slightly longer to balance the kinematics."?

RepRapFirmware does auto calibration so it can determine the "tower lengths" automatically. These are set un M666 (look up details on the wiki above) but you should not have to set these manually if you do a 6 factor auto calibration and then persist with M500.

bullet 1) While true, in practice the errors are so small that simple linear math makes it much simpler for the average person.

bullet 2) You don't need to copy and paste if you do M500 - which persists the calibration results in config-override.g. Make sure you have an M501 at the very end of your config.g file to load the override file.

Delta Arm Calibration for Positional Tolerance

by Lucidwolf » Wed Aug 29, 2018 9:02 am


Used to Repetier firmware and first time owning a board with the juice to run RepRap Firmware. I can't seem to find the location to modify the arm lengths directly. Michael Hackney post on the subject ( ... parts.html) only talks about the global rod dimension and an X-Y scale factor. My issue is tower X and Z are close to the same required distance, but Y needs to be slightly longer to balance the kinematics. M665 just has global arm length.

Extra info for anyone who cares from my positional tolerance dial in: :ugeek:
  • There is a slight change to Mike's scaling factor for X-Y dimensions in relation to the arm length. Based on the calculus the corrected L = original L * sqrt(measured X / true X) is more correct, but if you bought this kit the error is close enough that using a slightly worse convergence criteria doesn't really matter. If you want the exact converged values for you rods based on your delta radius use the excel at ( uses newton raphson method to converge on the answer for the non-linear equations. Necessary? Not at all but that never stopped me. Based off the work from PurpleSensation.
  • If you set up your M665 and M666 command in your config to match the output of the M665 and M666 after G32 it makes it easy to copy and paste without messing something up. Ex) M665 L367.05 R204.042 H457.872 B140 X0.052 Y0.356 Z0.0 ; Set diagonal rod length, delta radius, homed height, printable radius, X twr deg adj, Y twr deg adj, Z twr deg adj