markniu
Hitting the bed and "Endstop z still triggered after retrect" error on klipper
Hi!
So i managed to get BD Sensor installed, wired and installed on Klipper following your instructions step by step.
Everything seems to be fine, Calibration seems fine with the first value beeing 230 and the second beeing 257.
> 21:18:29
$ M102 S-5
21:18:29
230
21:18:29
257
21:18:29
289
21:18:29
321
21:18:29
356
21:18:29
391
21:18:29
426
21:18:29
460
21:18:29
499
21:18:29
517
21:18:29
562
21:18:29
610
21:18:29
642
21:18:29
681
21:18:29
719
21:18:29
743
21:18:29
763
21:18:29
814
21:18:29
881
21:18:29
920
21:18:29
958
21:18:29
980
21:18:29
980
21:18:29
980
21:18:29
1015
21:18:29
1015
21:18:29
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:30
1015
21:18:31
1015
21:18:31
1015
21:18:31
1015
21:18:59
$ M84
21:19:36
$ M119
21:19:36
x:open y:open z:TRIGGERED
As you can see Z Endstop is triggering correctly and reports open when the bed is a few cm away from the Nozzle.
The Issue i`m having is that the nozzle is hitting the bed ever so slightly during homing, the bed moves dowm a few mm and klipper is giving me a "Endstop z still triggered after retract" error. So homing correctly is still not possible.
Looked through printer.cfg but couldn`t figure out where the issue could be.
[BDsensor] sda_pin: PE13 # Don't use aliases for the board pins scl_pin: PC5 # -clk pin example of connecting to main board Creality V4.2.7 #scl_pin:MKS_THR:gpio20 # example of connecting to CAN module like MKS THR42 #sda_pin:MKS_THR:gpio11 # example of connecting to CAN module like MKS THR42 delay: 20 # 20us per pulse, this value should be >=20 but must be below 50 z_offset: 0 # this
z_offsetmust be set to 0. #z_adjust: 0.0 # z axis adjustment, replace the z_offset function. within -0.3 to 0.3mm x_offset: 38 y_offset: -1 #no_stop_probe: # enable this for fast probe, the toolhead will not stop at the probe point. position_endstop: 0.8 # the Z axis will stop at this position (mm) while homing z, recommend value is 0.4~1.0 speed: 0.8 # this speed only works for the z tilt and PROBE_ACCURACY command.
[safe_z_home] home_xy_position = 162,162 speed = 150 z_hop = 10 z_hop_speed = 8
[force_move] enable_force_move: true
[bed_mesh] speed: 150 horizontal_move_z: 2 mesh_min: 38,5 mesh_max: 320,320 probe_count: 6,6 mesh_pps: 2,2 fade_start: 1 fade_end: 10 algorithm: bicubic bicubic_tension:0.2
[z_tilt] speed: 150 horizontal_move_z: 2 retries: 10 retry_tolerance: 0.0040 z_positions: 396,186 -71,186 points: 285,186 0,186
Any ideas on this?
I Have to mention that i`m on a corexy machine and using a steel pei sheet with a magnet sticker on a glass bed
What's the homing speed of Z axis, how about slow down more the homing speed of Z axis?
also set the horizontal_move_z in the bed_mesh section form 2 to 1 , because the actual distance it can measure is 2.2mm from your data from 230 to 980.
I am having this same issue also, my number are:
[stepper_z] step_pin: PD7 dir_pin: !PD6 enable_pin: !PF10 rotation_distance: 40 gear_ratio: 60:20 microsteps: 16 endstop_pin: probe:z_virtual_endstop homing_speed: 8 second_homing_speed: 0.8 position_max: 270
[BDsensor]
sda_pin: PB2 # Don't use aliases for the board pins
scl_pin: PB1 # # example of connecting to main board Creality V4.2.7
delay: 20 # 20us per pulse, this value should be >=20 but must be below 50
z_offset:0 # this z_offset must be set to 0.
z_adjust:0.0 # z axis adjustment, replace the z_offset function. within -0.3 to 0.3mm
x_offset: 0
y_offset: 0
no_stop_probe: # enable this for fast probe, the toolhead will not stop at the probe point.
position_endstop: 0.8 # the Z axis will stop at this position (mm) while homing z, recommend value is 0.4~1.0
speed:0.8 # this speed only works for the z tilt and PROBE_ACCURACY command.
[safe_z_home] home_xy_position:125,105 #speed:100 z_hop:10
[bed_mesh] horizontal_move_z:1 # 0.7~1.0mm is recommended algorithm: bicubic # this should be bicubic if the bed mesh points count is > 6*6 mesh_min: 35, 6 mesh_max: 240, 198 probe_count: 5, 3
Someone fixed it by re-install and re-flash the mcu, you can try that. also try to slow down the home speed from 8 to 4 in section stepper_z
So i changed some things that you mentioned @markniu e.g. home speed to 5 (with 6 i got same error on homing again) Also horizontal_move_z from 2 to 1 which was a mistake when i posted my config.
Also try to lower the z_hop in safe:z_home section @rolohaun Dont know if this is directly related to this issue but it semms to help in imy case. And as a rook mk1 owner, building a rook 2020 right know im glad that you are playing around with BD Sensor aswell :)
Homing, z_tilt , bed mesh in general and KAMP are working now but i get very inconsistent first layer when printing with pla (60°C bed). First time the nozzle was way to close to the bed leaving some marks on my PEI sheet. Second and third prints where way to close to the bed aswell but don`t damaged my bed at least
How can i adjust that? My understand is that BD Sensor should care for first layer itself?
Also i cant use TMC Autotune any longer since i installed the patched klipper version for bd sensor. I know, totally another topic but can you confirm at least that this isnt working because of the "dirty" klipper install?
https://github.com/andrewmcgr/klipper_tmc_autotune
Homing, z_tilt , bed mesh in general and KAMP are working now but i get very inconsistent first layer when printing with pla (60°C bed). First time the nozzle was way to close to the bed leaving some marks on my PEI sheet. Second and third prints where way to close to the bed aswell but don`t damaged my bed at least
How can i adjust that? My understand is that BD Sensor should care for first layer itself?
You can adjust the offset by changing the value z_adjust in BDsensor section but the recommend way is adjust the nozzle at the zero position and calibration again.
When (and why?) exactly do i have to calibrate again?
For example if the motor is unlocked before calibration the zero position maybe changed while doing calibration, because of the movement inside the step motor from unlock to lock. and you can also use "G28" and "G1 Z0" to check if the zero position is as expected after calibration.
I don't think of that. the install of BDsensor is only copy the source code file BDsensor.py and BD_sensor.c into the klipper folder and let it can be compiled into the firmware, and only if there is a senction `BDsensor in the printer.cfg file that will run that source code, and also only in doing probing process.
Also i can
t use TMC Autotune any longer since i installed the patched klipper version for bd sensor. I know, totally another topic but can you confirm at least that this isnt working because of the "dirty" klipper install?https://github.com/andrewmcgr/klipper_tmc_autotune
When (and why?) exactly do i have to calibrate again?
For example if the motor is unlocked before calibration the zero position maybe changed while doing calibration, because of the movement inside the step motor from unlock to lock. and you can also use "G28" and "G1 Z0" to check if the zero position is as expected after calibration.
Okay, i got that. So after G28 i did G1 Z0 and the nozzle is hitting the bed slightly. I did M102 S-6 and these are the results
11:38:14 $ M102 S-6 11:38:14 // Calibrating, don't power off the printer 11:38:15 // Please Wait... 11:38:44 // Calibrate Finished! 11:38:44 // You can send command BDSENSOR_READ_CALIBRATION to check the calibration data 11:38:54 $ M102 S-5 11:38:54 224 11:38:54 225 11:38:54 228 11:38:54 236 11:38:54 262 11:38:54 293 11:38:54 328 11:38:54 362 11:38:54 397 11:38:54 431 11:38:54 468 11:38:54 505 11:38:55 541 11:38:55 577 11:38:55 612 11:38:55 632 11:38:55 669 11:38:55 700 11:38:55 721 11:38:55 786 11:38:55 799 11:38:55 880 11:38:55 889 11:38:55 949 11:38:55 973 11:38:55 973 11:38:55 973 11:38:55 1015 11:38:55 1015 11:38:55 1015 11:38:55 1015 11:38:55 1015 11:38:55 1015 11:38:55 1015 11:38:56 1015 11:38:56 1015 11:38:56 1015 11:38:56 1015 11:38:56 1015 11:38:56 1015
Second data is only one point more than the first one, so not good right? :( I´m hesitant to start a print with that result because my build plate is already slightly damaged.
What potentially could be the issue here? Steel Plate too thin? Position of the probe overall?
Edit: I did Calibration with a super clean nozzle and build platze of course ;)
seems that the nozzle is still touch the bed after it move up 0.3mm. the sensor will use the first point as the zero position so that will cause problem.you can let nozzle a little higher about 0.1mm from the bed and calibrate again until it get good look data and the zero position is as expected.
okay with Z at 0,10mm i still got bad results from calibration but with 0,20mm i got that
12:17:22 452 12:17:22 467 12:17:22 525 12:17:22 564 12:17:22 602 12:17:22 639 12:17:22 652 12:17:22 715 12:17:22 753 12:17:22 789 12:17:22 797 12:17:22 849 12:17:23 887 12:17:23 925 12:17:23 966 12:17:23 966 12:17:23 966 12:17:23 1015 12:17:23 1015 12:17:23 1015 12:17:23 1015 12:17:23 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:24 1015 12:17:25 1015
When i do G1 Z0 now what should it look like? I still don´`t get the part where the zero position is after calibration. Nozzle touching the bed?
Started a print and the nozzle is digging into the bed when the print is starting
I`m done for now...
if it still too close to the bed, then use the z_adjust that is used for replaceing the z_offset
Can i still do babystepping over the web ui or klipperscreen?
It can only be settled by manual now, but you can use the web ui to do babystepping and find what's the value that should be set
Someone fixed it by re-install and re-flash the mcu, you can try that. also try to slow down the home speed from 8 to 4 in section stepper_z
changing from speed 8 to 4 fixed it for me
With adjusting Z Offset values back and forth in printer.cfg i got one good print with a nice first layer but the next one wirhout changing anything and same GCode was way to close to the Bed. So not consistent at all for me. Don't know If more testing is worth it since i still don't know what my issue is.
With adjusting Z Offset values back and forth in printer.cfg i got one good print with a nice first layer but the next one wirhout changing anything and same GCode was way to close to the Bed. So not consistent at all for me. Don't know If more testing is worth it since i still don't know what my issue is.
Could you check the bed mesh data and map displayed on the mainsail or fluidd is consistent?
I have the same problem, my version is V1.1 pandapi3d.com
What i found out of my observations:
- You should calibrate only when hot (bed and maybe nozzle too).
- You must home hot too, otherwise it crashes into bed.
- KAMP is now working, which is great.
- I made calibration with 0.1 from the bed and the numbers were good, but it's still scary to home the printer.
- When i do endstop check in fluidd in the tune menu, z triggers before probe (there's a slight delay).
And one suggetion @markniu can you please make the window on z_adjust to more than 0.3 -0.3 to maybe 0.5 -0.5 ? Because it's not enough.
With adjusting Z Offset values back and forth in printer.cfg i got one good print with a nice first layer but the next one wirhout changing anything and same GCode was way to close to the Bed. So not consistent at all for me. Don't know If more testing is worth it since i still don't know what my issue is.
Could you check the bed mesh data and map displayed on the mainsail or fluidd is consistent?
Unfortunately not right now. Im in the process of switching my toolhead on that printer what will take me 1 or 2 days. But i will replay to that for sure. I still wanna get that sensor running relaible, even with tzhe new toolhead. I did not take any screenshots from my bed mesh but it wasnt looking unusual as far as i renember
would you tell me the hardware version? or the version after the M102 S-1
V1.1 pandapi3d.com for me aswell. Bought it at melow store over an aliexpress a couple of weeks before
Same issue, also have a v1.1 Sometime would home fine, cold bed, but nozzle pressing on the bed lightly, even with recalibration with nozzle 0.2mm above bed, same result. Light turns on well above the bed at least 1-2mm above. Other random time ( as in testing homing right after the previous one ) with no change whatsoever would do a hard crash and keep digging into the bed until I hit emergency stop.
i know there is a v1.3 out, so maybe that would solves those issues?
Same issue I had with v1.0 one year ago and reverted to cr touch.
Sendt fra Outlook for Androidhttps://aka.ms/AAb9ysg
From: Squirrelf @.> Sent: Thursday, December 28, 2023 6:06:50 PM To: markniu/Bed_Distance_sensor @.> Cc: Subscribed @.***> Subject: Re: [markniu/Bed_Distance_sensor] Hitting the bed and "Endstop z still triggered after retrect" error on klipper (Issue #64)
Same issue, also have a v1.1 Sometime would home fine, cold bed, but nozzle pressing on the bed lightly, even with recalibration with nozzle 0.2mm above bed, same result. Light turns on well above the bed at least 1-2mm above. Other random time ( as in testing homing right after the previous one ) with no change whatsoever would do a hard crash and keep digging into the bed until I hit emergency stop.
i know there is a v1.3 out, so maybe that would solves those issues?
— Reply to this email directly, view it on GitHubhttps://github.com/markniu/Bed_Distance_sensor/issues/64#issuecomment-1871350261, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AWOSXTWL4DV43MT7S2SXJ53YLWRKVAVCNFSM6AAAAABBDTGCGWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNZRGM2TAMRWGE. You are receiving this because you are subscribed to this thread.Message ID: @.***>
ok update for me, I have version 1.1 and I send my z_homing to 2 and I seem to have reliable first layers, Quad gantry level sometimes does not work but it seems ok. this sensor does seem to work, I wish the distance from the bed was greater it must be so close to the nozzle height. very easy to hit a printed part
I am sorry for that bad experience and thank you for feedback this problem and support. the hardest thing for us is that we still do not know the actual reason since that we have not repeated this problem, and at the same time get some good feedback from some users, maybe it is the problem of the quality from the factory. now we have the new hardware version v1.3 since last month. anyone who met this problem can send email to pandapi or mellow webstore and request replacement.
- I made calibration with 0.1 from the bed and the numbers were good, but it's still scary to home the printer.
you can try to increase the position_endstop: 0.8. for example from 0.8 to 1.2
I have the same problem, my version is V1.1 pandapi3d.com
What i found out of my observations:
- You should calibrate only when hot (bed and maybe nozzle too).
- You must home hot too, otherwise it crashes into bed.
- KAMP is now working, which is great.
- I made calibration with 0.1 from the bed and the numbers were good, but it's still scary to home the printer.
- When i do endstop check in fluidd in the tune menu, z triggers before probe (there's a slight delay).
And one suggetion @markniu can you please make the window on z_adjust to more than 0.3 -0.3 to maybe 0.5 -0.5 ? Because it's not enough.
Thanks for this good information. I will consider increasing a little more about the z_adjust.
- I made calibration with 0.1 from the bed and the numbers were good, but it's still scary to home the printer.
you can try to increase the
position_endstop: 0.8. for example from 0.8 to 1.2
There's a cap on 1.0 in there. I tried 1.2 and it gave me an error that i'm tryint to print lower than the bed. Maybe you can increase it too?
Okay, it crashes into bed 1 time out of 5. I asked mellow store for a replacement, and they sent it. One question, why can't it stop when homing, when the diode first lights up?