My printer has gotten an attitude.

[geneb]

What matter control does isn't calibration. It's a bed-leveling gimmick. Get yourself a JohnSL FSR kit, install it, send G32 and then sit back in awe.

g.

[dc42]

No need to use Matter Control, the Duet firmware has had least squares calibration built in since early 2015. Just make sure you have the Z probe type set to 0 (P0 in the M558 command in config.g), set up a bed.g file, and command auto calibration form the web interface. See https://duet3d.dozuki.com/Wiki/Calibrat ... ta_printer.

[rootboy]

@Gene: I like MC for manual control, laying out multiple parts, and scaling. What do you use? I'm certainly open to suggestions.

I had heard of the force resistors but I was thinking of installing it/them on the hotend. What do you recommend, bed or head? And do you have a link or a STL for the bed mounts if I go that way? Thanks!

@dc42: Thanks, I will try that out as soon as my print is finished.

[rootboy]

Well, I had to shop around a bit as Ultibots was out of the controller, and I got the FSR plate and heat spreader from Trick Laser. Hopefully it will all be in this weekend and I'll let the printer do its own cal.

[dc42]

What I meant was don't use MC for calibration. Use it for slicing if you are happy with it.

[rootboy]

For the most part it does what I need. On bigger slices, it chokes and dies a horrible death. And that's when I switch to Repetier & Cura.

I took a look at the calibration procedure that you suggested, and I created the bed.g g-code and ran that. I was in the process of using the calibration calculator when it occurred to me that I was just going to have to do it all over again after I add the FSRs. I might as well wait until then.

[geneb]

Keep in mind that when you run the G32 command to calibrate the printer, it's going to take around a minute. (My Artemis 300 takes 58 seconds to do a 16 point calibration.) With a process that fast, I just added a G32 after the G28 in the start g-code of the slicers I use. I don't even bother saving the cal results.

g.

[rootboy]

Keep in mind? Heck, I had to look up what "G32" did.

I doubt that I will get all of my parts for the FSR's by this weekend, so I broke down and calibrated it old school using my dial indicator. The towers and the bed are within .0005". It's printing *very* nicely at the moment.

Two things that I've noticed thus far (besides the whisper quiet operation of the motors) is that my temps on the head no longer run 5 degrees hotter than what I set it to. And it no longer over-extrudes (I would normally set my extruder multiplier to .91). Nice.

[cope413]

Just so you guys are aware, MatterControl can use FSRs or other "automatic" probes (including ones actuated by a servo) to perform leveling.

Currently, there are 4 options.

3 point plane
7 point disc
13 point disc
3 x 3 mesh

We'll soon have the option to probe any size mesh/grid you want.

MC 2.0 is just around the corner, and there are some new leveling features in that build, also. Specifically, there's a probe offset wizard (to calculate your Z offset). You can download the experimental version of 2.0 for Windows now. It should be in Alpha in the next couple weeks, with a release candidate following shortly thereafter.


Also, we're working on a leveling process that will be "smart". It will probe a spot on the bed, and if the first 2-3 samples are the same, or within some specified tolerance, then it will move on to the next point. If there are outside of the threshold, it will continue probing. This is similar to what RepRapFirmware does, but it will be much more granular - specify # of samples, tolerance, how to handle the values (average them, throw out high/low values, etc), and be much easier to tinker with than the M270 gcode.

[rootboy]

Excellent! I'm an unabashed fan of MC. It does 98% of what I need, and it's a snap to use. I'll try out the new version after this print is finished. Thanks!