Heuristic (AI) calibration for delta printers on Smoothie

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Sat Sep 26, 2015 6:01 am

My fan wiring shorted out and blew the FET on my Smoothieboard. I tried to remove it, but they used some kind of weird Martian space solder. My soldering iron turned all the way up to 480C AND a desoldering iron put together were enough to get the solder to soften, but not melt. The solder pads lifted from excess heat, meaning that the board is functionally destroyed. Even if they hadn't lifted, I'm sure some nearby components got cooked. The board got so hot that I couldn't hold it.

I will not spend any more money on Smoothieboards until they come out with version 2 next year. This means that I have no way to test any new releases. If someone wants me to merge in the latest upstream changes, I can do that and upload a binary for testing.

I'm thinking about going with the BeagleBone Black and a CRAMPS board. I just need to look over the source again and make sure I want to do this.

I have a feeling this migration is going to be like a root canal.

Leo13
Noob
Posts: 2
Joined: Sun Sep 27, 2015 4:55 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by Leo13 » Tue Sep 29, 2015 8:58 am

Hello,

Sorry for my bad english, i hope that you will understand me.

I have make a kossel max, i use the board MKS Sbase (like a smoothieboard) with your firmware 626pilot ( July 2015 version). All worked perfectely before the new release firmware (9 days ago). It is the same with the original smoothieware firmware.

I have check alls, config file, conectic....

But the arms don't move correctly, with the commande G29, the Z axe down and not the two others. With the G28 the X and Z go at home position but not the Y.


Where can I find the previous version of your 626pilot's firmware ? I need to use two SD card, one that work well and one to test the news.

In the futur : It is possible to create a heuristic calibration plugin that can be merge with the orignal smoothieware ?


Thank you for your good job

Best regards
Leo

Leo13
Noob
Posts: 2
Joined: Sun Sep 27, 2015 4:55 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by Leo13 » Thu Oct 01, 2015 3:09 am

It's good, i have find the problem, a connexion.

All work with the last firmware.

Thanks

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Thu Oct 01, 2015 5:59 am

I'm glad you figured it out. Smoothie doesn't have support for plugins.

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Mon Oct 05, 2015 9:52 pm

I'm getting ready to merge in all the recent upstream changes to Smoothieware (better PID tuning, etc.) to prepare for what is probably my last release for version 1 of Smoothie. I can merge in future changes down the line, if there is enough interest (read: people asking me to do it here).

Did anyone else try my latest firmware, and if so, did it set the tower scale to any value other than 1.0? Since that was an experimental feature that never did anything for me, I want to remove it if it really is useless. If you haven't tried it and want to help out, now would be the time!

RMLynch
Printmaster!
Posts: 53
Joined: Fri Jul 31, 2015 6:48 am
Location: Ottawa, Ontario, Canada

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by RMLynch » Mon Oct 05, 2015 10:05 pm

I am interested, as are a couple of people I know.

On one of our 3D Printers we are testing out the latest Calibration Routines.

Do you have a SmoothieBoard at the moment that works?
My Spool Holder for the Rostock MAX v2 on Repables & Thingiverse.

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Mon Oct 05, 2015 11:15 pm

I have *a* Smoothieboard, some of which was heated to ~500C while I was trying to desolder a blown FET. I have no idea if it would even boot.

...Actually, now that I come to think of it, I have another with a buggy thermistor port. I can still use that to test calibration.

EskimoRuler
Plasticator
Posts: 14
Joined: Mon Feb 09, 2015 2:38 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by EskimoRuler » Wed Oct 07, 2015 4:53 pm

I want to start off by saying Thank You to 626Pilot for all the work that you have done for the Smoothieboard. Your firmware is the reason that I decided to purchase a smoothieboard, and possibly why I have used my printer so much since installing it (I was having the hardest time using the Rambo and Repetier to print all over the bed making me not want it any longer.)

A little background on my setup. I bought a RSMv2 in February and had a blast with the build. Printing started off well with the Fan Shrouds, and also with a few smaller prints about the size of the shrouds. But when I started to try and print larger objects, I noticed that I was having high spots between the towers. I was able to get a good calibration once by shimming the bed, and messing with the tower rotation. But it led to other problems with dimensions, and after cleaning the glass one day, I could not get it back to where it was once I put the glass back on.

So I decided to give 626Pilot's firmware for the smoothieboard a go. I purchased everything to make 626Pilot's Hall-O Probe (Did the best I could printing the parts) and installed the Smoothieboard. After a couple attempts calibrating the probe, adjusting the speeds, I did a full calibration and was very satisfied the first time around. I was getting energies around .070-.100, but I did feel that the calibration had the potential to be much better. Time to Upgrade.

Right around this time was when Brian released the Max Metal Platform. I purchased the Platform, some Trick Trucks, 325mm CF Arms, 0.9 degree steppers, and a Prometheus. After Making these upgrades I have been getting great calibrations since.

It is very sad to see the troubles you have been going through with the smoothieboard and even more sad to see you leave to another platform, but I will say, I'll follow you to the next device because your Heuristic Calibration has been amazing and kept me interested in my printer. This post is my attempt to hopefully get you to keep updating your Smoothieware fork in the future :D

So thank you for what you have put into the Smoothieboard, and for this community here. It is really appreciated!!

Below is the output of my last calibration using the build from 6/12/2015.

Printer is now online.
>>> G29 E
SENDING:G29 E
[PR] Repeatability test: 10 samples (S)
[PR] Acceleration (A): 90.0
[PR] Debounce count (B): 0
[PR] Smooth decel (D0|D1): True
[PR] Eccentricity test (E): On
[PR] Probe smoothing (P): 1
[PR] Probe priming (Q): 5
[PR] Feedrates: Fast (U) = 70.000, Slow (V) = 35.000
[PR] 1 step = 0.00625 mm.
[BH] Determining the probe-from height.
[BH] Probe-from height = 300.712
[BH] Bed height set to 310.675
[PR] Test 1 of 10: Measured 1592 steps (9.950 mm)
[PR] Test 2 of 10: Measured 1592 steps (9.950 mm)
[PR] Test 3 of 10: Measured 1592 steps (9.950 mm)
[PR] Test 4 of 10: Measured 1592 steps (9.950 mm)
[PR] Test 5 of 10: Measured 1592 steps (9.950 mm)
[PR] Test 6 of 10: Measured 1592 steps (9.950 mm)
[PR] Test 7 of 10: Measured 1592 steps (9.950 mm)
[PR] Test 8 of 10: Measured 1593 steps (9.956 mm)
[PR] Test 9 of 10: Measured 1593 steps (9.956 mm)
[PR] Test 10 of 10: Measured 1592 steps (9.950 mm)
[PR] Stats:
[PR] range: 1 steps (0.0063 mm)
[PR] mu: 1592.200 steps (9.951 mm)
[PR] sigma: 0.400 steps (0.002 mm)
[PR] Repeatability: 0.0063 (add a little to be sure)
[PR] This is your best score so far!
[PR] This score is very good!

>>> G32
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: 325.000
[PK] Delta radius: 133.500
[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)
[ES] Endstops: Difference => 0.162 (want 0.030), out of tolerance by 0.132.
[ES] Setting endstops to {-0.211, 0.000, -0.122}.
[DR] Delta Radius - Depths: Center=9.950, Tower average=9.967 => Difference: -0.017 (want 0.030), within tolerance.

[IC] Iteration 2 (max 20)
[ES] Endstops: Difference => 0.050 (want 0.030), out of tolerance by 0.020.
[ES] Setting endstops to {-0.162, 0.000, -0.130}.
[DR] Delta Radius - Depths: Center=9.837, Tower average=9.871 => Difference: -0.033 (want 0.030), out of tolerance by -0.063.
[DR] Changing delta radius from 133.500 to 133.433

[IC] Iteration 3 (max 20)
[ES] Endstops: Difference => 0.012 (want 0.030)
[ES] Endstops are within tolerance.
[DR] Delta Radius - Depths: Center=9.856, Tower average=9.863 => Difference: -0.006 (want 0.030), within tolerance.

[PK] Current kinematic settings:
[PK] Arm length: 325.000
[PK] Delta radius: 133.433
[PK] Endstop offsets: {-0.162, 0.000, -0.130}
[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] All done! Save settings with M500.
>>> M500
SENDING:M500
Settings Stored to /sd/config-override

>>> G31 OPQRS U0.01
SENDING:G31 OPQRS U0.01

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

[HC] Data source: Probe
[HC] Active tests: [Endstops (O), mul=1.00] [Delta Radius (P), mul=1.00] [Arm Length (Q), mul=1.00] [Tower Angle Offset (R), mul=1.00] [Virtual Shimming (S), mul=1.00]
[HC] Inactive tests: (none)
[HC] Set geom during/after (J): During
[HC] Keep last settings (K): False
[HC] Annealing tries (T): 50
[HC] Max temp (U): 0.010
[HC] Binary search width (V): 0.100
[HC] Overrun divisor (W): 2.000
[HC] Zero all offsets (Y): False

[HC] Baseline kinematics updated.
[HC] Keeping existing kinematics.
[HC] Depth-mapping the print surface...

[PK] Current kinematic settings:
[PK] Arm length: 325.000
[PK] Delta radius: 133.433
[PK] Endstop offsets: {-0.162, 0.000, -0.130}
[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

[DM] Depth to bed surface at center: 1580 steps (9.875 mm)

[PD] 0.013
[PD]
[PD] [ -- ] 0.031 -0.006 -0.050 -0.094 -0.131 [ -- ]
[PD]
[PD] [ -- ] 0.031 -0.006 -0.038 -0.025 -0.044 [ -- ]
[PD]
[PD] 0.081 -0.019 0.006 0.000 0.019 -0.025 -0.056
[PD]
[PD] [ -- ] 0.000 0.000 0.031 0.013 0.006 [ -- ]
[PD]
[PD] [ -- ] -0.013 -0.044 0.006 -0.006 0.000 [ -- ]
[PD]
[PD] -0.056
[PD]
[PD] Best=0.000, worst=0.131, min=-0.131, max=0.081, mu=-0.008, sigma=0.032, energy=0.030

[HC] Generating carriage positions for a printer with this configuration.

[HC] Starting test configuration: Arm Length=325.000, Delta Radius=133.433

[HC] ***** Simulated annealing pass 1 of 1 in progress *****
[HC] Existing calibration has energy 0.030

[HC] Reticulating splines...
[HC] Try 0 of 50, energy=0.028 (want <= 0.010)
[HC] Try 5 of 50, energy=0.023 (want <= 0.010)
[HC] Try 10 of 50, energy=0.022 (want <= 0.010)
[HC] Try 15 of 50, energy=0.022 (want <= 0.010)
[HC] Try 20 of 50, energy=0.020 (want <= 0.010)
[HC] Try 25 of 50, energy=0.019 (want <= 0.010)
[HC] Annealing has stalled - aborting.

[HC] End of annealing pass (energy=0.019)
[HC]
[HC] Heuristic calibration complete (energy=0.019)

[PK] Current kinematic settings:
[PK] Arm length: 324.998
[PK] Delta radius: 133.357
[PK] Endstop offsets: {-0.190, 0.000, -0.157}
[PK] Radius offsets (ABC): {0.000, 0.076, 0.076}
[PK] Angle offsets (DEF): {-0.075, 0.021, 0.099}
[PK] Virtual shimming: {0.012, 0.006, -0.037}, vector={0.000, 0.000, 1.000}, d=0.003, Enabled
[PK] Depth (Z) correction: Disabled

[HC] Checking calibration. If it's worse than it was before, you may have to run this several times!
[DM] Depth to bed surface at center: 1574 steps (9.837 mm)

[PD] 0.013
[PD]
[PD] [ -- ] -0.038 -0.044 -0.044 -0.044 -0.044 [ -- ]
[PD]
[PD] [ -- ] 0.019 -0.013 -0.025 -0.013 -0.019 [ -- ]
[PD]
[PD] 0.069 -0.038 -0.013 0.000 -0.006 -0.031 -0.062
[PD]
[PD] [ -- ] 0.025 0.000 0.031 0.000 -0.019 [ -- ]
[PD]
[PD] [ -- ] 0.019 -0.031 -0.006 -0.019 -0.044 [ -- ]
[PD]
[PD] -0.050
[PD]
[PD] Best=0.000, worst=0.069, min=-0.062, max=0.069, mu=-0.009, sigma=0.023, energy=0.028

[HC] You can run this command again to see if it gets better, or type M500 to save.

>>> G31 OPQRS U0.01
SENDING:G31 OPQRS U0.01

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

[HC] Data source: Probe
[HC] Active tests: [Endstops (O), mul=1.00] [Delta Radius (P), mul=1.00] [Arm Length (Q), mul=1.00] [Tower Angle Offset (R), mul=1.00] [Virtual Shimming (S), mul=1.00]
[HC] Inactive tests: (none)
[HC] Set geom during/after (J): During
[HC] Keep last settings (K): False
[HC] Annealing tries (T): 50
[HC] Max temp (U): 0.010
[HC] Binary search width (V): 0.100
[HC] Overrun divisor (W): 2.000
[HC] Zero all offsets (Y): False

[HC] Baseline kinematics updated.
[HC] Keeping existing kinematics.
[HC] Depth-mapping the print surface...

[PK] Current kinematic settings:
[PK] Arm length: 324.998
[PK] Delta radius: 133.357
[PK] Endstop offsets: {-0.190, 0.000, -0.157}
[PK] Radius offsets (ABC): {0.000, 0.076, 0.076}
[PK] Angle offsets (DEF): {-0.075, 0.021, 0.099}
[PK] Virtual shimming: {0.012, 0.006, -0.037}, vector={0.000, 0.000, 1.000}, d=0.003, Enabled
[PK] Depth (Z) correction: Disabled

[DM] Geometry hasn't changed since last depth map - keeping it!


[PD] 0.013
[PD]
[PD] [ -- ] -0.038 -0.044 -0.044 -0.044 -0.044 [ -- ]
[PD]
[PD] [ -- ] 0.019 -0.013 -0.025 -0.013 -0.019 [ -- ]
[PD]
[PD] 0.069 -0.038 -0.013 0.000 -0.006 -0.031 -0.062
[PD]
[PD] [ -- ] 0.025 0.000 0.031 0.000 -0.019 [ -- ]
[PD]
[PD] [ -- ] 0.019 -0.031 -0.006 -0.019 -0.044 [ -- ]
[PD]
[PD] -0.050
[PD]
[PD] Best=0.000, worst=0.069, min=-0.062, max=0.069, mu=-0.009, sigma=0.023, energy=0.028

[HC] Generating carriage positions for a printer with this configuration.

[HC] Starting test configuration: Arm Length=324.998, Delta Radius=133.357

[HC] ***** Simulated annealing pass 1 of 1 in progress *****
[HC] Existing calibration has energy 0.028

[HC] Reticulating splines...
[HC] Try 0 of 50, energy=0.021 (want <= 0.010)
[HC] Try 5 of 50, energy=0.020 (want <= 0.010)
[HC] Try 10 of 50, energy=0.018 (want <= 0.010)
[HC] Try 15 of 50, energy=0.018 (want <= 0.010)
[HC] Try 20 of 50, energy=0.019 (want <= 0.010)
[HC] Try 25 of 50, energy=0.017 (want <= 0.010)
[HC] Annealing has stalled - aborting.

[HC] End of annealing pass (energy=0.017)
[HC]
[HC] Heuristic calibration complete (energy=0.017)

[PK] Current kinematic settings:
[PK] Arm length: 324.961
[PK] Delta radius: 133.316
[PK] Endstop offsets: {-0.155, 0.000, -0.203}
[PK] Radius offsets (ABC): {0.000, 0.118, 0.118}
[PK] Angle offsets (DEF): {-0.125, 0.116, 0.097}
[PK] Virtual shimming: {0.049, 0.024, -0.025}, vector={0.000, 0.000, 1.000}, d=-0.021, Enabled
[PK] Depth (Z) correction: Disabled

[HC] Checking calibration. If it's worse than it was before, you may have to run this several times!
[DM] Depth to bed surface at center: 1581 steps (9.881 mm)

[PD] 0.087
[PD]
[PD] [ -- ] 0.000 -0.019 0.000 0.006 0.000 [ -- ]
[PD]
[PD] [ -- ] 0.013 0.000 -0.006 0.025 0.006 [ -- ]
[PD]
[PD] 0.019 -0.062 -0.006 0.000 0.019 -0.013 -0.056
[PD]
[PD] [ -- ] 0.000 0.006 0.050 0.019 0.000 [ -- ]
[PD]
[PD] [ -- ] 0.025 -0.013 0.038 0.006 -0.013 [ -- ]
[PD]
[PD] 0.000
[PD]
[PD] Best=0.000, worst=0.087, min=-0.062, max=0.087, mu=0.003, sigma=0.021, energy=0.018

[HC] You can run this command again to see if it gets better, or type M500 to save.

>>> G31 AF
SENDING:G31 AF
[!!] Forcing re-probe.

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

[DC] Probing bed for depth correction...
Prefix: Max prefixes exceeded (5)
[DC] Depth to bed surface at center: 1594 steps (9.962 mm)

[PD] 0.087
[PD]
[PD] [ -- ] -0.013 -0.006 0.006 -0.013 -0.025 [ -- ]
[PD]
[PD] [ -- ] -0.019 -0.038 -0.044 -0.013 0.013 [ -- ]
[PD]
[PD] 0.013 -0.038 0.013 0.000 0.031 -0.006 -0.069
[PD]
[PD] [ -- ] 0.013 0.006 0.025 -0.006 -0.006 [ -- ]
[PD]
[PD] [ -- ] 0.025 0.013 0.056 0.069 0.044 [ -- ]
[PD]
[PD] 0.025
[PD]
[DC] Probing at circle's edge for extrapolation...

[DC] Complete output, with extrapolated points (numbers in brackets):

[PD] 0.087
[PD]
[PD] [ 0.067] -0.013 -0.006 0.006 -0.013 -0.025 [-0.131]
[PD]
[PD] [-0.026] -0.019 -0.038 -0.044 -0.013 0.013 [-0.093]
[PD]
[PD] 0.013 -0.038 0.013 0.000 0.031 -0.006 -0.069
[PD]
[PD] [-0.003] 0.013 0.006 0.025 -0.006 -0.006 [-0.044]
[PD]
[PD] [-0.001] 0.025 0.013 0.056 0.069 0.044 [-0.142]
[PD]
[PD] 0.025
[PD]
[PD] Best=0.000, worst=0.142, min=-0.142, max=0.087, mu=-0.005, sigma=0.041, energy=0.026

[DM] Surface transform saved to SD card.
[DM] Checking calibration...
[DM] Depth to bed surface at center: 1592 steps (9.950 mm)

[PD] -0.044
[PD]
[PD] [ -- ] -0.044 -0.038 -0.044 -0.019 -0.013 [ -- ]
[PD]
[PD] [ -- ] -0.006 0.019 0.019 -0.006 -0.050 [ -- ]
[PD]
[PD] -0.031 -0.044 -0.044 0.000 -0.044 -0.019 -0.025
[PD]
[PD] [ -- ] -0.013 0.000 0.013 0.006 -0.006 [ -- ]
[PD]
[PD] [ -- ] -0.019 -0.025 -0.013 -0.062 -0.031 [ -- ]
[PD]
[PD] 0.000
[PD]
[PD] Best=0.000, worst=0.062, min=-0.062, max=0.019, mu=-0.012, sigma=0.019, energy=0.025

[DM] /!\ IMPORTANT /!\ Type M500 to save!
>>> M500
SENDING:M500
Settings Stored to /sd/config-override
Last edited by EskimoRuler on Thu Oct 08, 2015 1:19 am, edited 1 time in total.

EskimoRuler
Plasticator
Posts: 14
Joined: Mon Feb 09, 2015 2:38 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by EskimoRuler » Wed Oct 07, 2015 5:35 pm

For my Probe, I have been using 626Pilot Hall-O with a modified Top and Bottom that matches the 3 main mounting holes

I have also "Flipped" the magnets so that they push away from each other. I feel that this keeps the rod more steady, and also makes sure that the rod doesn't get pulled up and you end up crashing the probe during calibration.

03_ZProbe.jpg

02_ZProbe.jpg
Attachments
626_Remix_Top_fixed.stl
(88.46 KiB) Downloaded 9 times
626_Remix_Bottom_fixed.stl
(107.6 KiB) Downloaded 8 times

RMLynch
Printmaster!
Posts: 53
Joined: Fri Jul 31, 2015 6:48 am
Location: Ottawa, Ontario, Canada

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by RMLynch » Wed Oct 07, 2015 10:32 pm

I agree with the above post...

I wanted to get the SmoothieBoard from reading all it can do, but it was because of 626Pilot's efforts that pushed me to purchase one, actually two of them!

So I second the praise for 626Pilot's efforts and hope to see him continue his great efforts in supporting the Smoothieware for those of us with them.

Maybe we could all pitch in a little bit of $$$ to get 626Pilot a new SmoothieBoard to allow him to continue helping us out here with future firmware upgrades?

As long as someone teaches him how to solder properly? :D
My Spool Holder for the Rostock MAX v2 on Repables & Thingiverse.

User avatar
teoman
ULTIMATE 3D JEDDI
Posts: 1783
Joined: Sat May 24, 2014 5:43 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by teoman » Thu Oct 08, 2015 4:14 pm

Somehow I doubt he needs any schooling on soldering.
When on mobile I am brief and may be perceived as an arsl.

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Fri Oct 09, 2015 1:17 am

EskimoRuler wrote:I want to start off by saying Thank You to 626Pilot for all the work that you have done for the Smoothieboard. Your firmware is the reason that I decided to purchase a smoothieboard, and possibly why I have used my printer so much since installing it (I was having the hardest time using the Rambo and Repetier to print all over the bed making me not want it any longer.)

[...]


It is very sad to see the troubles you have been going through with the smoothieboard and even more sad to see you leave to another platform, but I will say, I'll follow you to the next device because your Heuristic Calibration has been amazing and kept me interested in my printer. This post is my attempt to hopefully get you to keep updating your Smoothieware fork in the future :D

I'm glad you got some use out of it, and your calibration numbers are really good too. I'm still on good terms with the Smoothie team - or at least Arthur Wolf, who explained the cause of the problems and (more importantly) what he was doing to fix them.

Right now I'm working on an update to my Smoothie release. There are updates to some stuff that have happened elsewhere in the firmware over the summer (better PID tuning, etc.), so a couple of days ago I did a merge to pull them all in. I'll be hooking up a mostly-working Smoothieboard to test the release, to make sure the calibration still runs.

For the time being, I'm going to use a Duet. If the included auto-calibration system isn't as good as mine - and it may not be, as it doesn't adjust as many variables - I'll port mine over. There is a possibility that I'll work on the Replicape (BeagleBone Black solution) next year, pending availability of an expansion board that lets me drive more steppers. Eventually the Smoothieboard 2 will come out, and I'd like to develop for that as well, although it's about a year off and I don't know how everything will pan out. I would certainly like to develop for it, as the Smoothie code base is the best 3D printer firmware I've ever seen. It's a lot better laid out than the rest of what I've seen, and it doesn't rely on uploading crap from the Arduino IDE at paint-dryingly-slow ISDN speeds.

EskimoRuler wrote:For my Probe, I have been using 626Pilot Hall-O with a modified Top and Bottom that matches the 3 main mounting holes

I have also "Flipped" the magnets so that they push away from each other. I feel that this keeps the rod more steady, and also makes sure that the rod doesn't get pulled up and you end up crashing the probe during calibration.

Nice work! You should publish it to Thingiverse as a remix if you haven't already.

RMLynch wrote:Maybe we could all pitch in a little bit of $$$ to get 626Pilot a new SmoothieBoard to allow him to continue helping us out here with future firmware upgrades?

It is a nice thought, but I have a board I can develop on. I just can't have two thermistors connected to it at the same time because all but one analog port is fried. (Totally my fault, too.) Good enough for testing calibration. Smoothie is not going to be my production firmware because I need six stepper drivers, something that will hopefully be rectified with the version 2 board.

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Fri Oct 09, 2015 5:56 am

OK, I just pushed an update. Changes:

Code: Select all

Merged with upstream changes done over the summer (better PID tuning, etc.)
Added probe_ignore_bed_temp setting for people who don't have a (working) heated bed
Probe repeatability test aborts if you have more than 5 steps of range
Took out obsolete code for per-arm length
Took out code for tower scale, since it does nothing for me and nobody ever told me it did anything for them
Minor updates to README.creole
Minor bugfixes to G29 probe calibration

Before you had to type G31 HOPQRS, but since tower scale is gone you'll just type G31 HOPQR.

There is a new config option, probe_ignore_bed_temp. If you don't have a heated bed, set this to true. You can see how it looks here.

You can download the updated firmware.bin here.

EskimoRuler
Plasticator
Posts: 14
Joined: Mon Feb 09, 2015 2:38 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by EskimoRuler » Wed Oct 14, 2015 1:42 am

626Pilot wrote:OK, I just pushed an update. Changes:

Code: Select all

Merged with upstream changes done over the summer (better PID tuning, etc.)
Added probe_ignore_bed_temp setting for people who don't have a (working) heated bed
Probe repeatability test aborts if you have more than 5 steps of range
Took out obsolete code for per-arm length
Took out code for tower scale, since it does nothing for me and nobody ever told me it did anything for them
Minor updates to README.creole
Minor bugfixes to G29 probe calibration

Before you had to type G31 HOPQRS, but since tower scale is gone you'll just type G31 HOPQR.

There is a new config option, probe_ignore_bed_temp. If you don't have a heated bed, set this to true. You can see how it looks here.

You can download the updated firmware.bin here.



So I tried to run the calibration on the latest firmware but I keep getting a "ERROR: Couldn't query bed temperature!" when I try to run the simulated annealing. I have the probe_ignore_bed_temp config set to false. I have even tried it set to true and I get the same output. I have also tried different temps from 90-100C. I feel like I am missing something simple.

My config file
https://www.dropbox.com/sh/i3al9uvem78whhf/AAADhav9w1B18mQ-vRAd53-Ua?dl=0

The output from Pronterface
Error.txt
(12.54 KiB) Downloaded 14 times



But i did run the calibration again from your 6/12 build and I still get great results
Results_2015-10-12.txt
(14.54 KiB) Downloaded 9 times

3D52
Noob
Posts: 3
Joined: Wed Oct 14, 2015 10:07 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 3D52 » Wed Oct 14, 2015 1:25 pm

EskimoRuler wrote:
626Pilot wrote:OK, I just pushed an update. Changes:

Code: Select all

Merged with upstream changes done over the summer (better PID tuning, etc.)
Added probe_ignore_bed_temp setting for people who don't have a (working) heated bed
Probe repeatability test aborts if you have more than 5 steps of range
Took out obsolete code for per-arm length
Took out code for tower scale, since it does nothing for me and nobody ever told me it did anything for them
Minor updates to README.creole
Minor bugfixes to G29 probe calibration

Before you had to type G31 HOPQRS, but since tower scale is gone you'll just type G31 HOPQR.

There is a new config option, probe_ignore_bed_temp. If you don't have a heated bed, set this to true. You can see how it looks here.

You can download the updated firmware.bin here.



So I tried to run the calibration on the latest firmware but I keep getting a "ERROR: Couldn't query bed temperature!" when I try to run the simulated annealing. I have the probe_ignore_bed_temp config set to false. I have even tried it set to true and I get the same output. I have also tried different temps from 90-100C. I feel like I am missing something simple.

My config file
https://www.dropbox.com/sh/i3al9uvem78whhf/AAADhav9w1B18mQ-vRAd53-Ua?dl=0

The output from Pronterface
Error.txt



But i did run the calibration again from your 6/12 build and I still get great results
Results_2015-10-12.txt

I got the same error when I tried it.

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Wed Oct 14, 2015 6:44 pm

EskimoRuler wrote:So I tried to run the calibration on the latest firmware but I keep getting a "ERROR: Couldn't query bed temperature!" when I try to run the simulated annealing. I have the probe_ignore_bed_temp config set to false. I have even tried it set to true and I get the same output. I have also tried different temps from 90-100C. I feel like I am missing something simple.

My config file
https://www.dropbox.com/sh/i3al9uvem78whhf/AAADhav9w1B18mQ-vRAd53-Ua?dl=0

According to that file, you have this line in your config:

Code: Select all

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

That word should be true, not false.

3D52, can you post your config too?

3D52
Noob
Posts: 3
Joined: Wed Oct 14, 2015 10:07 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 3D52 » Wed Oct 14, 2015 7:26 pm

Here is my config file. I tried both true and false in this line: (comprehensive-delta.probe_ignore_bed_temp false # Don’t wait for bed temp to stabilize (ONLY use this if you DON”T have a heated bed!)
Attachments
config.txt
(22.69 KiB) Downloaded 12 times

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Wed Oct 14, 2015 9:08 pm

That file also has the value set to false.

Do this:

  • Shut off the printer completely (including removing the USB cable from the board)
  • Put the SD card into a reader
  • Edit the file, setting the value to true
  • Save & close
  • Use your OS's "safe eject" feature (may be called eject, unmount, etc.) to unmount the card so that it isn't visible as a drive anymore
  • Remove the card from the reader, put it in your printer and boot it up

That should fix it. I don't know why it wouldn't.

WZ9V
Printmaster!
Posts: 65
Joined: Sat Jan 25, 2014 8:49 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by WZ9V » Fri Oct 16, 2015 10:54 pm

I tried your latest firmware.bin and when I do the probe test it seems to slam into the bed. It's like It's trying to move past the bed. I know the FSR's are working because they trigger with the regular Smoothie firmware. What should I look into?

User avatar
bubbasnow
ULTIMATE 3D JEDDI
Posts: 1064
Joined: Fri Aug 02, 2013 4:24 pm
Location: Dayton, WA

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by bubbasnow » Sat Oct 17, 2015 4:31 pm

WZ9V wrote:I tried your latest firmware.bin and when I do the probe test it seems to slam into the bed. It's like It's trying to move past the bed. I know the FSR's are working because they trigger with the regular Smoothie firmware. What should I look into?


make sure the pin assignments didn't change between the two firmware? if you change the z plunge speed and force a trigger before the end hits the plate and nothing happens then i would check with switch connection/pin assignment.

User avatar
626Pilot
ULTIMATE 3D JEDDI
Posts: 1719
Joined: Tue May 14, 2013 12:52 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by 626Pilot » Sat Oct 17, 2015 8:12 pm

WZ9V wrote:I tried your latest firmware.bin and when I do the probe test it seems to slam into the bed. It's like It's trying to move past the bed. I know the FSR's are working because they trigger with the regular Smoothie firmware. What should I look into?

That means Smoothie is not reading a signal when your Z probe triggers. You can test the status with M119. If you run the tip down to the glass, M119 should show your Z probe as being triggered once it touches the glass. If that doesn't happen, your probe setup needs to be fixed. The probe signal should go to one of the pins on the Z MIN endstop port. I never tried using FSRs. My microswitch probe has one end connected to +5V and another connected to the rightmost pin on Z MIN, but your FSR setup might be different. Just make sure you don't accidentally wind up shorting 5V to ground or anything like that.

DaveS
Noob
Posts: 2
Joined: Sun Sep 27, 2015 8:51 am

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by DaveS » Wed Oct 21, 2015 7:40 am

Just want to say Thank You to 626Pilot for this great calibration firmware.
I have been using an older version on my Kossel with SmoothieBoard, and just tried the 09 October release.today, and it seems to be working fine.
Thank You

WZ9V
Printmaster!
Posts: 65
Joined: Sat Jan 25, 2014 8:49 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by WZ9V » Sat Oct 24, 2015 10:59 am

I'll check it again but it works with the main Smoothie so I know it's registering. Must be something in the config not right.

626Pilot wrote:
WZ9V wrote:I tried your latest firmware.bin and when I do the probe test it seems to slam into the bed. It's like It's trying to move past the bed. I know the FSR's are working because they trigger with the regular Smoothie firmware. What should I look into?

That means Smoothie is not reading a signal when your Z probe triggers. You can test the status with M119. If you run the tip down to the glass, M119 should show your Z probe as being triggered once it touches the glass. If that doesn't happen, your probe setup needs to be fixed. The probe signal should go to one of the pins on the Z MIN endstop port. I never tried using FSRs. My microswitch probe has one end connected to +5V and another connected to the rightmost pin on Z MIN, but your FSR setup might be different. Just make sure you don't accidentally wind up shorting 5V to ground or anything like that.

WZ9V
Printmaster!
Posts: 65
Joined: Sat Jan 25, 2014 8:49 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by WZ9V » Sat Oct 24, 2015 4:04 pm

Got it working it as I suspected my FSR setup is long n repeatable measurements. Using DC42's IR probe I got better numbers but since it's offset so far from the hot end. I need to figure out how to make the probing radius/grid smaller to keep it on the plate.

WZ9V
Printmaster!
Posts: 65
Joined: Sat Jan 25, 2014 8:49 pm

Re: Heuristic (AI) calibration for delta printers on Smoothi

Post by WZ9V » Sat Oct 24, 2015 9:01 pm

626Pilot, does you calibration set the maximum Z-Height or do you still need to do that separately like with the stock algorithm?

Post Reply

Return to “Smoothieboard and variants”