Welcome! Log In Create A New Profile

Advanced

Rich Cattell auto calibration, lots of iterations.

Posted by heine016 
Rich Cattell auto calibration, lots of iterations.
October 13, 2014 08:18PM
Hopefully a quick question: How many iterations does Rich Cattell's G30 A autocalibration require to finish?


Background:
- I have a 3DR I've built up, and have just incorporated the Trinket/FSR auto cal Z probe.
- The probe system works really well, the LED on the trinket shows hits exactly as it should, at good levels of pressure.
- Firmware seems to be working fine, I can run G30 at certain points and get repeatable results, and they adjust nicely with endstop tweaks (ie -0.40mm in front of tower X, increase offset 0.40mm and new probe is ~0.0).
- Printer is relatively calibrated already, ie +/- 0.1mm but only measured at centre, and in front of each tower.
- I haven't tried to correct tower offsets at all, but believe they are a little out.


I've run the G30 A and it has taken up to 50 iterations and it is still messing with the endstops, and seems to be going round in circles, doesn't get to a 'calibrated' state.
Understand I could just finish fine tuning the hard way, but having seen those amazing videos of Kossel's probing a field of points and then printing flawlessly, I'd really like to get some sort of auto-cal working on my 3DR!


Does anyone have any suggestions or info about using RC marlin? I'm open to changing firmware, anything really!


Appreciate your help.
Re: Rich Cattell auto calibration, lots of iterations.
October 14, 2014 06:59AM
It has a limit of 99, but still may not be calibrated.
Yes, it can easily hit 50 or more.

On my big printer I have to start with .5 accuracy, copy out the settings, then drop it a tenth and repeat until I hit .1mm
Even doing that, it took it something like 300 iterations total.
Re: Rich Cattell auto calibration, lots of iterations.
October 14, 2014 07:20PM
Ahh ok that could be doing it, I left it at default which was 0.03 I think! Will try doing it like you described.

I've found if my endstops are bang on it will start changing the radius, tower correction etc but as soon as it goes back to calibrating endstops it just goes in around in circles and I have to end it.

Cheers for the info
Re: Rich Cattell auto calibration, lots of iterations.
December 18, 2014 10:04AM
I would also add something: autocalibration works better if you have a small offset, a couple of milimeters.

If you have a big differente betwen the setup values (inside configuration file) and the real values... it will take forewer.

I noticed this...the hard way.

And yes, your endstoppers must be "perfect" :-)
Sorry, only registered users may post in this forum.

Click here to login