Heuristic (AI) calibration for delta printers on Smoothie

[Aflac]

I'll check those things tomorrow, but visually it descended triggered, then barely went back up at all before going to the edge and quitting. I'm using fsrs, probe height is set to zero I think, I did not change that from what was I the sample config file, My config file is posted in another thread in the smoothie section. Don't have access to repost it here since I'm on my tablet now.

[Aflac]

Sorry I read the file wrong I was thinking z offset for some reason not probe height, that is 3 I think

[Aflac]

so i took probe height up to 7 and its working fine now...however when its near the edges its a lot closer to the bed than at the center...guess i have a number wrong somewhere in the geometry settings.

[626Pilot]

That's exactly what it sounds like.

[bubbasnow]

Hi, i am trying to run the calibration but i have run into a snag.
I get through the G29 ok, but when I do the G32 it stops very quickly. any ideas what i should change?

SENDING:G32

[IC] Iterative calibration in progress. Press Reset to abort.
[IC] /!\ PROBE CRASH DANGER /!\ Don't press buttons, send commands, or access the SD card.

[IC] Resetting kinematics.

[PK] Current kinematic settings:
[PK] Arm length: 330.770
[PK] Delta radius: 128.500
[PK] Endstop offsets: {0.000, 0.000, 0.000}
[PK] Radius offsets (ABC): {0.000, 0.000, 0.000}
[PK] Angle offsets (DEF): {0.000, 0.000, 0.000}
[PK] Virtual shimming: {0.000, 0.000, 0.000}, vector={0.000, 0.000, 1.000}, d=0.000, Disabled
[PK] Depth (Z) correction: Disabled


[IC] Iteration 1 (max 20)
[DP] do_probe_at(): steps=0 - this is much too small - is probe_height high enough?

Are your arms truly 330.77mm?

[Aflac]

Yes that is exactly what they are......a little long I know

Turns out my delta radius was off. Fixed that and it works much better.

And of course when I get to the next step out stood at can't read bed temperature. Yay

[Aflac]

so when i try G31 H O Q R i get this, the end where it says couldn't query bed temperature! seems to be the problem, but i have good bed temps i can see everything, and set temperatures, all the bed temperature control stuff seems to work fine.


Recv: [HC] Heuristic calibration in progress. Press Reset to abort.
Recv: [HC] /!\ PROBE CRASH DANGER /!\ Don't press buttons, send commands, or access the SD card.
Recv:
Recv: [HC] /!\ At start of method, kinematics are:
Recv:
Recv: [PK] Current kinematic settings:
Recv: [PK] Arm length: 330.770
Recv: [PK] Delta radius: 135.504
Recv: [PK] Endstop offsets: {0.000, -1.757, -3.242}
Recv: [PK] Radius offsets (ABC): {0.000, 0.000, 0.000}
Recv: [PK] Angle offsets (DEF): {0.000, 0.000, 0.000}
Recv: [PK] Virtual shimming: {0.000, 0.000, 0.000}, vector={0.000, 0.000, 1.000}, d=0.000, Disabled
Recv: [PK] Depth (Z) correction: Disabled
Recv:
Recv: [HC] /!\ And, cur_set is:
Recv:
Recv: [PK] Current kinematic settings:
Recv: [PK] Arm length: 330.770
Recv: [PK] Delta radius: 135.504
Recv: [PK] Endstop offsets: {0.000, -1.757, -3.242}
Recv: [PK] Radius offsets (ABC): {0.000, 0.000, 0.000}
Recv: [PK] Angle offsets (DEF): {0.000, 0.000, 0.000}
Recv: [PK] Virtual shimming: {0.000, 0.000, 0.000}, vector={0.000, 0.000, 1.000}, d=0.000, Disabled
Recv: [PK] Depth (Z) correction: Disabled
Recv:
Recv: [HC] Data source: Probe
Recv: [HC] Active tests: [Endstops (H), mul=1.00] [Delta Radius (O), mul=1.00] [Tower Angle Offset (Q), mul=1.00] [Virtual Shimming (R), mul=1.00]
Recv: [HC] Inactive tests: [Arm Length (P), mul=1.00]
Recv: [HC] Set geom during/after (J): During
Recv: [HC] Keep last settings (K): False
Recv: [HC] Annealing tries (E): 50
Recv: [HC] Max temp (U): 2.000
Recv: [HC] Binary search width (V): 0.050
Recv: [HC] Overrun divisor (W): 2.000
Recv: [HC] Eval metric (X): Mean
Recv: [HC] Zero all offsets (Y): False
Recv:
Recv: [HC] Baseline kinematics updated.
Recv: [HC] Keeping existing kinematics.
Recv: [HC] Depth-mapping the print surface...
Recv:
Recv: [PK] Current kinematic settings:
Recv: [PK] Arm length: 330.770
Recv: [PK] Delta radius: 135.504
Recv: [PK] Endstop offsets: {0.000, -1.757, -3.242}
Recv: [PK] Radius offsets (ABC): {0.000, 0.000, 0.000}
Recv: [PK] Angle offsets (DEF): {0.000, 0.000, 0.000}
Recv: [PK] Virtual shimming: {0.000, 0.000, 0.000}, vector={0.000, 0.000, 1.000}, d=0.000, Disabled
Recv: [PK] Depth (Z) correction: Disabled
Recv:
Recv: [DM] ERROR: Couldn't query bed temperature!
Recv: [DM] Couldn't depth-map the surface.
Recv: ok

[626Pilot]

Could be the bed temperature is oscillating by more than 1 degree. I have good luck with these PID settings (Onyx, 12V): P170 I0.5 D200

If that doesn't help, there is a config option that will force it to ignore the bed temp:

Code: Select all

comprehensive-delta.probe_ignore_bed_temp                 true             # Don't wait for bed temp to stabilize (ONLY use this if you DON'T have a heated bed!)

[Aflac]

thanks, ill have to try a few things, i have my oynx hooked up to a 24v power supply. the temperature seems very stable but ill take a look at that.

Edit: figured i would edit since no one wants an email alert for me to say yup that fixed it. so i disabled the temperature and it worked great, got my energy down to 0.010...then had a problem with my extruder and had to take the bed off to work wiring issues and not im at 0.095.

[Aflac]

So on the G31 i know it defaults to 50 tries, but is there anything to be lost, other than a couple minutes by setting E1000 or something like that?

[Aflac]

Ok I got all the calibrations done and I can print, however the first layer scrapes and doesn't put down any filament. Is there additional calibration I need to do after running everything it should I be good to go? If so what would cause the scraping?

[bubbasnow]

now that your bed is mapped thats fine, but i always drop my nozzle 10mm above the surface, and move it down incrementally then set the zheight using a piece of paper. then do M306 Z0 to ensure nozzle height is set correctly. This is done at temp.

[Aflac]

now that your bed is mapped thats fine, but i always drop my nozzle 10mm above the surface, and move it down incrementally then set the zheight using a piece of paper. then do M306 Z0 to ensure nozzle height is set correctly. This is done at temp.

And I am printing, Thanks that did it.

[EskimoRuler]

Just wanted to give another thanks to 626Pilot.

I recent cracked my glass plate, luckily the PEI wasn't damaged, so I used another 3m sheet and stuck it one of Brians Trick Laser Heat spreaders. It's not the flattest base to stick to, but with your Smoothie calibration, it prints great across the bed still.

[img]https://www.dropbox.com/s/lvjztqrmsdozr ... .jpg?raw=1[/img]
[img]https://www.dropbox.com/s/i30mul1v3vmmp ... .jpg?raw=1[/img]
[img]https://www.dropbox.com/s/9sx1467p41dd3 ... .jpg?raw=1[/img]

[miglo]

Having used FSRs for about a year now, I just never really liked them. They always seemed to vary so much, and I've tried everything. I mean when I touched the bed, it *seemed* pretty sensitive. But during calibrations, it would vary from reboot to reboot. Results were usually pretty inconsistent.

So I decided to use the poorman's microswitch as a z-probe. Couldn't find any good STLs so I made my own. Added magnets to to make attaching it easier and painless.

Result was a significantly better calibration - and on the cheap!

[img]http://i.imgur.com/WSa6z3H.jpg[/img]
[img]http://i.imgur.com/w3EbRBX.jpg[/img]

[bubbasnow]

Having used FSRs for about a year now, I just never really liked them. They always seemed to vary so much, and I've tried everything. I mean when I touched the bed, it *seemed* pretty sensitive. But during calibrations, it would vary from reboot to reboot. Results were usually pretty inconsistent.

So I decided to use the poorman's microswitch as a z-probe. Couldn't find any good STLs so I made my own. Added magnets to to make attaching it easier and painless.

Result was a significantly better calibration - and on the cheap!

[img]https://drive.google.com/uc?export=down ... Th4Q2RRN1E[/img]
[img]https://drive.google.com/uc?export=down ... FdNVkNjdFk[/img]

you dont keep your zero lash straps in that location during use do you???????????

Proper installation is in the first picture on the first post here. http://forum.seemecnc.com/viewtopic.php?f=54&t=1934

in your current placement your parallelograms are no longer parallelograms.

[miglo]

Ya ooops...I thought I saw some pics awhile back with them in that location, I never bothered to readjust them. I guess I should!

[geneb]

Please host images on imgur (or upload them directly to the forum) - hosting with googleusercontent spews tracking cookies all over the place and nearly makes it impossible to load this thread due to Privacy Badger throwing warning dialogs.

g.

[Xenocrates]

I'm with you there Geneb. Privacy badger combined with the choice of image hosts is a pain in the rear.

[miglo]

Sorry geneb, never even realize that it was doing that or that it was a problem at all.

fixed.

[Zipper]

I have used this fork for some time now and want to thank Pilot626 and all the other people contributing to this fantastic Smoothie fork. The calibration routine works absolutely perfect for me and i can finally use all of the print surface. Initially i had some issues and was forced to use the sep version but after designing a new kinematic coupling probe switch the latest version runs fine. The only thing I'm missing now is a way to automatically set the bed height with a button on the panel. Maybe this is doable now but i can't figure out how to use G30 for this.

[Eaglezsoar]

miglo , would it be possible for you to share with us the stl files and the source of the magnets that you used and the part number?

Hate to ask but it is the only way to possibly get the files! Great job!

[626Pilot]

Heard by way of the Smoothieware support group that there is a new version of S3D: https://www.simplify3d.com/software/release-notes/

Of particular interest:

Improved filtering for small movements during toolpath generation
Improved support for latest Smoothieware firmware and USB communication protocols

This may mean that the microscopic movements bug has been squashed.

[Jimustanguitar]

miglo , would it be possible for you to share with us the stl files and the source of the magnets that you used and the part number?

Second.

[Jimustanguitar]

miglo , would it be possible for you to share with us the stl files and the source of the magnets that you used and the part number?

Hate to ask but it is the only way to possibly get the files! Great job!

Found it!
http://www.thingiverse.com/thing:1547303