Kossel / Delta Dimensional Calibration with Instructions

by Coffea Mar 24, 2017
Download All Files

Thing Apps Enabled

Please Login to Comment

In the new Marlin 1.1.9 can I change the DELTA_DIAGONAL_ROD variable in the PREPARE/DELTA CALIBRATION/DELTA SETTINGS/ Diag Rod section and obtain the same result as uploading firmware again? The variable there now is the same one in the configuration.h file.

Just an update to my post above. It seems changing the value in the delta settings does accomplish the same thing without having to reupload the firmware now. Just remember to save settings.

Thanks for your work on this and hope you don't mind as I linked this to my Thingiverse giving you full credit.
On a question I'm getting and finding it hard to explain why with this setting in Da Hai firmware version of Marlin for the Anycubic pully. With my adjustments for the Anycubic Linear plus.

And again thanks for the good work you have done with this.

Given shrinkage should this be calibrated to ~97mm for PLA or 92mm for ABS? I guess what I mean, not knowing how the rod lengths actually come into the equations here... if the model is 100mm and it prints 100mm, then actually it printed about 103mm and the print just shrunk 3% and landed where I wanted it to. This is great if I'm printing 100mm parts all the time, however if I print a 200mm part, does the new rod length mean it's actually printing a 206mm part that shrinks to 200mm? I haven't tested this yet but I will...

The reason I ask is because I thought if I calibrated the system to print a 100mm part at exactly 100mm compensating for shrinkage automatically then I'd simply continue to ignore compensating for shrinkage as long as I stuck to PLA or any material with similar shrinkage properties, however, I'm not finding that to be the case. I'm printing parts and still having to adjust by about 3% when precision is required.

Went from 99.2mm to 100.01mm. Very nice.

I am so glad I found this! Brilliant!!!

Would it not be easier to just measure your rods from centre pivot to centre pivot?
Slicer settings/width of extruded material could throw off your measurements no?

The length of the physical rods doesn't take into account how the design of linear guides, effector, joints, wear & tear, the human who assembled it, slicer, filament and ambient temperature.

The actual measured length of the rods is irrelevant, all that matters is how long your firmware thinks they are for it's internal IK calculations. What's important is producing accurate results and measuring a known print and using that to adjust the firmware rod length is the single most effective way to accomplish that.

It basically boils down to math vs human and mechanical slop :)

I see, like how accurate was the machine that made the machine that made the machine!!
Don't see how most of what you mention would effect diagonal_rod but ok, diagonal_rod really is another value with fudge factor, like smooth_rod_offest, for the numbers used final calculations.
Some human must not have bothered adding a value like: diagonal_rod_offset!
Personally I just can't get my head around the idea of adding slicer and filament extrusion variables to the mix, especially if designing parts to be printed on other machines!
I wonder if adding a pin under the nozzle to probe around a piece of paper (scale inkjet printed template?) could remove some of the variables.
A useful and interesting thing regardless, thanks for sharing :)

I think it's more of an issue with self builds and kits, my printer certainly has some deviations in the provided plastic parts. Easiest to think about it like a jointed robot arm with each joint amplifying any slop in the previous section.

you did a great job!
thanks for the detailed instruction.
also works on non-delta printers --> check out the description of my 'made' for further info

the only suggestion i have: my slicer wasn't able to handle the thin "walls" of the text. can you make it more massive?

Thanks! Glad it worked!! I've updated the instructions to mention your make for Cartesian printers. I will fix the text the next time I need to print this out.

hey coffea,

it turned out my previous comment was wrong.
the delta calculation does NOT work on Cartesian Printers.

so today i created a excel file, that -after gathering some numbers- does the calculations for me!!

i used your design and modified the text a bit. hope, you are okay about that?

here you are: Cartesian Dimensional Calibration with Instructions and Excel-Calculator

Cartesian Dimensional Calibration with Instructions and Excel-Calculator

Awesome, updated my listing :D Thanks !!