I just upgraded to Duet Wifi+Teensy LC and trying to do the Auto Bed Compensation but the nozzle is probing in the air about halfway down the full length. Can someone point me in the right direction of how to set these parameters properly/have a file I can download? I believe it has something to do with the .9 steppers as everything is halved.
I set the steps per mm in config.g to 160 and it went all the way to the bed but it probed and skipped across the bed three times before moving to the next probe point. After the third it wouldn't touch the bed.
Duet AutoBed Compensation issue
Re: Duet AutoBed Compensation issue
Do you really mean bed compensation, or do you mean auto calibration as I suspect? If it's auto calibration, see https://duet3d.com/wiki/Calibrating_a_d ... parameters and the following sections on that page.
Re: Duet AutoBed Compensation issue
My apologies, Autobed leveling found in the Duet Wifi interface. I came across that guide late last night and planned to go through it soon, I was just wondering if anyone had a quick fix/file. I was also thinking of going back to stock motors just in case it gets too overwhelming trying to set these parameters.
Re: Duet AutoBed Compensation issue
You may find it useful to use the tool at reprapfirmware.org to generate the configuration, homing and bed probing files. But increase the dive height temporarily before you run auto calibration for the first time.
Re: Duet AutoBed Compensation issue
Went back to the 1.8 deg steppers and did auto cal, Probe wasn't triggering at all. Found out I was using the Teensy LC which is not 5v compatible...Ordered the Teensy 3.2 this time. I have FSRs+JohnSL board, just haven't had time to make a mount for the bed. I have both just in case I can't figure one out, I have something as backup. I eventually want to have FSRs from what I hear.
Re: Duet AutoBed Compensation issue
I've been thru what your doing in the last few weeks, tried both accelerometer interface boards and couldn't get consistent triggering and "triggering before errors" and I finally put the FSR's under the bed. After tweeking the config line, it works like a charm. I printed the FSR mounts, cleaned them up so the caps would glide better. Then put a silicone pad above and below the FSR in the mount, the above one should be stuck to the bottom of the cap. Screw the mount cap to the bottom of the heated bed snowflake. Make sure the FSR mounts are adjusted distance wise so that the bed slides up and down with no friction between the mount and the cap. Pressing on the bed should make the led's light up on the JohnSL board. Add the 2 lines below in your config.g M558 and M557.
; Z probe and compensation definition
M558 P4 X0 Y0 Z0 H10 F50 T6000 ; F50 very slow, but works great, higher values causes the bed to move down or shift during probing
; grid compensation enables/setup G29 probing map
M557 R135 S15
The John SL board should be connected to the E0 endstop on the duet wifi. The P4 in the M558 designates this. P5 would instruct the firmware to use the Zprobe port.
Set your bed height (HXXX.XXX in the line below) in the config.g file to about 10 less then your actual z height. Your homed height will probably be much higher than mine, I have 360mm arms. I think I did this because originally the HE280 accelerometer bed.g file I had was lowering the hot end to above the first point before probing and my HXXX.XXX (homed height) was too high and it would hit the bed before it actually started probing. If your bed.g doesn't have the gcode entry to lower the hotend to above the bed before the probing gcode then you dont need to change the HXXX.XXX
M665 R130.738 L360.270 B135 H252.800 X-0.178 Y0.035 Z0 ; set delta radius, diagonal rod length, printable radius and homed height
Goto http://escher3d.com/pages/wizards/wizardbed.php
Enter values 10, 6, 6 and put in your printable radius somewhere around 135. If you have short diagonal rods like 269 on the V2 you might have to lower the radius to like 120. It basically just tells the wizard how far to the perimeter to make the probe points. Click generate bed file. Copy the text into the bed.g file in the duet system. The file should have these 2 lines also above what you pasted:
M561 ; clear any bed transform, otherwise homing may be at the wrong height
G28 ; home the printer
then your pasted stuff from the wizard.
Once all setup click G-code console
Type G32 in the Send g-code field. This is the same as pressing the Auto Delta calibration button.
If all is setup correct, it should probe the bed in 16 points then move up off the bed.
Type M665 in the send gcode field and write down the values returned.
Type M666 in the send gcode field and write down the values for endstop corrections.
The values written down are the new settings to replace in the config.g file on the M665 line above. You are gonna change the RXXX.XXX (delta Radius) HXXX.XXX (Z-height) and the X, Y , and Z values reported.
Change this line in the config.g to adjust the enstops with the values returned from typing M666:
M666 X-0.35 Y0.72 Z-0.36 ; put your endstop adjustments here from auto calibration
The G32 can be run multiple times until the change reported converge. I ran it 3 times and then the values kept changing slightly up and down.
Bed Compensation
Change the M558 line in the config.g to have H3 instead of H10. This sets the z lift 3mm after each probe point, which will speed things up slightly when probing all the points.
M558 P4 X0 Y0 Z0 H3 F50 T6000
Run the G32 one last time without re-homing when done. Type G29 in the send gcode field. The printer will do the bed compensation probing. It will take a while because the speed is set at F50. If it completes successfully it will generate a bed compensation file and colorful picture.
Add this line to config.g file somewhere at the top:
G29 S1 ; load bed leveling compensation
This will instruct the firmware to use the compensation file whenever the printer is booted.
At this point if you print the bed will be level, but in my case the print was too close to the bed for the first layer and it smushed the plastic flat, backing up the flow of plastic from the extruder.
Adjust the HXXX.XXX in the M665 line of the config.g file to be lower. In my case I subtracted about 0.3 from the number there. The results are below.
Really even first layer across the entire bed:
[img]https://photos.smugmug.com/Other/3D-Pri ... 456-X3.jpg[/img]
I'm now able to print using the entire bed
[img]https://photos.smugmug.com/Other/3D-Pri ... 237-X3.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 852-X3.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 2227-M.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 5727-M.jpg[/img]
; Z probe and compensation definition
M558 P4 X0 Y0 Z0 H10 F50 T6000 ; F50 very slow, but works great, higher values causes the bed to move down or shift during probing
; grid compensation enables/setup G29 probing map
M557 R135 S15
The John SL board should be connected to the E0 endstop on the duet wifi. The P4 in the M558 designates this. P5 would instruct the firmware to use the Zprobe port.
Set your bed height (HXXX.XXX in the line below) in the config.g file to about 10 less then your actual z height. Your homed height will probably be much higher than mine, I have 360mm arms. I think I did this because originally the HE280 accelerometer bed.g file I had was lowering the hot end to above the first point before probing and my HXXX.XXX (homed height) was too high and it would hit the bed before it actually started probing. If your bed.g doesn't have the gcode entry to lower the hotend to above the bed before the probing gcode then you dont need to change the HXXX.XXX
M665 R130.738 L360.270 B135 H252.800 X-0.178 Y0.035 Z0 ; set delta radius, diagonal rod length, printable radius and homed height
Goto http://escher3d.com/pages/wizards/wizardbed.php
Enter values 10, 6, 6 and put in your printable radius somewhere around 135. If you have short diagonal rods like 269 on the V2 you might have to lower the radius to like 120. It basically just tells the wizard how far to the perimeter to make the probe points. Click generate bed file. Copy the text into the bed.g file in the duet system. The file should have these 2 lines also above what you pasted:
M561 ; clear any bed transform, otherwise homing may be at the wrong height
G28 ; home the printer
then your pasted stuff from the wizard.
Once all setup click G-code console
Type G32 in the Send g-code field. This is the same as pressing the Auto Delta calibration button.
If all is setup correct, it should probe the bed in 16 points then move up off the bed.
Type M665 in the send gcode field and write down the values returned.
Type M666 in the send gcode field and write down the values for endstop corrections.
The values written down are the new settings to replace in the config.g file on the M665 line above. You are gonna change the RXXX.XXX (delta Radius) HXXX.XXX (Z-height) and the X, Y , and Z values reported.
Change this line in the config.g to adjust the enstops with the values returned from typing M666:
M666 X-0.35 Y0.72 Z-0.36 ; put your endstop adjustments here from auto calibration
The G32 can be run multiple times until the change reported converge. I ran it 3 times and then the values kept changing slightly up and down.
Bed Compensation
Change the M558 line in the config.g to have H3 instead of H10. This sets the z lift 3mm after each probe point, which will speed things up slightly when probing all the points.
M558 P4 X0 Y0 Z0 H3 F50 T6000
Run the G32 one last time without re-homing when done. Type G29 in the send gcode field. The printer will do the bed compensation probing. It will take a while because the speed is set at F50. If it completes successfully it will generate a bed compensation file and colorful picture.
Add this line to config.g file somewhere at the top:
G29 S1 ; load bed leveling compensation
This will instruct the firmware to use the compensation file whenever the printer is booted.
At this point if you print the bed will be level, but in my case the print was too close to the bed for the first layer and it smushed the plastic flat, backing up the flow of plastic from the extruder.
Adjust the HXXX.XXX in the M665 line of the config.g file to be lower. In my case I subtracted about 0.3 from the number there. The results are below.
Really even first layer across the entire bed:
[img]https://photos.smugmug.com/Other/3D-Pri ... 456-X3.jpg[/img]
I'm now able to print using the entire bed
[img]https://photos.smugmug.com/Other/3D-Pri ... 237-X3.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 852-X3.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 2227-M.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 5727-M.jpg[/img]
Re: Duet AutoBed Compensation issue
This is exactly what I needed THANKS SO MUCH! I got the leveling down to .25 correction consistently with a little drop <.2 on the Z side but I'll tweak it later.
I did have a question regarding the wifi conversion, I noticed I have the PS-on (blue) and the i2c connectors not going to anything. is this right? Also my three cooling fans are not working after a certain height and I tried manually turning them on in duet interface but not response.
I did have a question regarding the wifi conversion, I noticed I have the PS-on (blue) and the i2c connectors not going to anything. is this right? Also my three cooling fans are not working after a certain height and I tried manually turning them on in duet interface but not response.
Re: Duet AutoBed Compensation issue
Since you have FSR's connected to the duet, the i2c and blue wires aren't used anymore. The i2c wires were the communication to the accelerometer on the He280, which has been replaced by the FSR's which don't need to be talked to. The blue ps -on wire was the output pulse from the accelerometer to trigger the duet that the accelerometer sensed the nozzle hit the bed, so it is not used anymore either.
As far as the fans I think we would need more info on what you mean by certain height. Could be a bad connection at the HE280 connector. I replaced the connector on my board and whip with a locking one and crimp terminals. The original whip has push in wires that are just held by spring tension of a bent metal tab. Just a note, the heavier gauge heater wires were a little tight for the crimp and needed a little persuading.
[img]https://photos.smugmug.com/Other/3D-Pri ... 151-X2.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 823-X2.jpg[/img]
As far as the fans I think we would need more info on what you mean by certain height. Could be a bad connection at the HE280 connector. I replaced the connector on my board and whip with a locking one and crimp terminals. The original whip has push in wires that are just held by spring tension of a bent metal tab. Just a note, the heavier gauge heater wires were a little tight for the crimp and needed a little persuading.
[img]https://photos.smugmug.com/Other/3D-Pri ... 151-X2.jpg[/img]
[img]https://photos.smugmug.com/Other/3D-Pri ... 823-X2.jpg[/img]
Last edited by wguibas on Tue Mar 07, 2017 5:05 pm, edited 1 time in total.
Re: Duet AutoBed Compensation issue
Thanks for that, I just wanted to make sure I wasn't going crazy about the extra wires. As far as fans, I found that another user had the same issue and it was the orange wire should be on the negative or left side of the two pin connector. I emailed mhackney to verify. I swapped sides and fans work like a charm again.