BLTouch working
-
Thanks. That's good then! Yes a diagram would help but I understand it's early days. I have it setup currently on my Bigbox (Rumba) so looking forward to switching.
-
If anyone could knock up a quick diagram of connecting the BLTOUCH to a duet wifi I would appreciate it. I am getting everything ready to change over on Saturday evening.
-
I have set my BL Touch up, but I am having a lot of repeatability issues that I am hoping someone could help me with. I have followed all the instructions in this thread as best as I could. I am using the resistor, plugging into the z probe pins, pin 1 and 2 on the expansion, and heater3 pin. I have tried changing the trigger value a bit which didnt do anything. All my connections seem fine as they are crimped and the resistor is soldered.
Here are the results (just the Z coordinate) from four 5-point auto bed compensations that I did in a row, without the heated bed or anything turned on.
[ 0.000] [-0.190] [0.425] [0.112] [0.158]
[ 0.082] [-0.170] [0.540] [0.503] [0.285]
[ 0.010] [-0.522] [0.780] [0.635] [0.293]
[-0.122] [-0.637] [0.322] [0.115] [0.197]As you can see there is a big difference between each new probe sequence. Here are some relevant config files and my bed file.
; BL Touch
M558 P5 X0 Y0 Z1 H4 F450 T4000
G31 P25 X36 Y0 Z1.5; Auto Bed Compensation
M561
M98 Pdeployprobe.g
; Probe the bed at 5 points
G30 P0 X15 Y15 H0 Z-99999
G30 P1 X15 Y175 H0 Z-99999
G30 P2 X185 Y175 H0 Z-99999
G30 P3 X185 Y15 H0 Z-99999
G30 P4 X100 Y100 H0 Z-99999 S
M98 Pretractprobe.gThanks in advance for any help.
-
Few things to check,
Make sure your config_override.g isn't changing the value of the G31 P25…command.
I'm actually running G31 P1 (vs P25) and have been. My theory is that the even shorter pulse detection gives the Z axis less time to move once triggered. I've had really consistent results. Granted if the z axis overshoots the same each time it shouldn't matter. But still perhaps worth a try.
Edit: Not that it matters but M401 can be used to call the deploy.g and M402 to call the retract. Just a bit cleaner.
Jeff
-
So your not wanting to run the grid based compensation , G29 ?
-
I remember I used to have a config override file but after updating a few times it seems to have disappeared. Is this a problem?
I tried a lower trigger value but haven't had any better results. I also decreased my Z acceleration as it was set pretty high. Also ironically shortly after I posted, I set up a macro for grid compensation. The probe points produce different values each time.
Any other suggestions? Thanks a lot for the help. I read your previous post about M401/2, but just forgot to change it, thanks though!
-
I remember I used to have a config override file but after updating a few times it seems to have disappeared. Is this a problem?
I tried a lower trigger value but haven't had any better results. I also decreased my Z acceleration as it was set pretty high. Also ironically shortly after I posted, I set up a macro for grid compensation. The probe points produce different values each time.
Any other suggestions? Thanks a lot for the help. I read your previous post about M401/2, but just forgot to change it, thanks though!
Hmm, mechanically sound printer ? In other words do you feel it's nice and solid and should be producing the same results each time ? For the grid what's the map look like, is it random junk ? Did you try changing the height of the BLTouch based on the docs that came with it ? 8m gap from bottom of BLtouch shell to bed.
Jeff
-
Yep my printer is pretty solid and prints consistently. I took it apart to make sure nothing got loose, tightened a few things, didn't help sadly. I am still getting deviations of 0.5mm maximum.
The grid is pretty random, I can upload the map if you want but I dont think it will help much.
I actually didnt get any docs with my BLTouch,, although it looked legit. I got it from spool3d.com. I modelled my own mount, and the distance from probe (retracted) to tip of nozzle is just 1.61mm. I read a distance of 2mm is optimal, so I don't think my mount height is causing any issues. Honestly at this point im thinking the probe is defective.
Thanks a lot for the help, I really appreciate it.
-
I thought the probe I got was defective at one point as well. It turned out my config override was causing the issues. You've ruled that out already. See if you can get it to probe the same exact spot over and over to P0 P1, P2, etc and see how much the values vary.
Jeff
-
If you are not getting consistent results from a Z probe, using a lower probing speed may help. It's the F parameter in the M558 command.
-
I probed the same spot 20 times and I ended up getting a standard deviation of 0.054.
Hopefully we can get this solved, I am losing hope though lol.
EDIT: Thanks for the input dc42. I dropped the probing speed to 100. I had already tried that with bad results, but this time I got a standard deviation of 0.026.
I will lower the speed again.I dropped the speed again to 50. Standard deviation is 0.032.
Any other ideas?
-
If the BL touch is genuine then contact the manufacturer and ask what repeatability they expect. Could be a hardware fault.
-
I just ran a bed probe:
This was on a dirty piece of glass with some left over print residue.
35 points probed, mean error -0.125, deviation 0.055On a different piece of semi clean glass:
35 points probed, mean error -0.050, deviation 0.076and again on the same peice:
35 points probed, mean error -0.170, deviation 0.073What command did you run to get the multi point probe ? I can try that.
Edit: Whatever my results I'm getting really nice flat first layers with even widths. Probes were done at F300.
Jeff
-
So I will take the 5v and Gnd from the expansion port (pin 1 & 2) also taking the Servo pin from the same source (pin 8 (heater 3)).
Now when wiring the bltouch black and white wires into the DuetWifi Z- probe pins do they go to 3.3v and Gnd or Z- probe in and GND?
Also is it OK to crimp the resistor in with the pin connectors?
-
So I will take the 5v and Gnd from the expansion port (pin 1 & 2) also taking the Servo pin from the same source (pin 8 (heater 3)).
Now when wiring the bltouch black and white wires into the DuetWifi Z- probe pins do they go to 3.3v and Gnd or Z- probe in and GND?
Also is it OK to crimp the resistor in with the pin connectors?
Z probe in and gnd. Be sure to use probe type 5 in M558. Also I dont see any problem crimping the resistor in the pin connector, as long as they are connected properly across. It might be easier to strip some insulation further up the wire though and put a bit of solder, but either or should work.
–---
And (In)Sanity here is my probing files, I have been using G30 S-1 to probe the single spot, the centre of my bed. For the G32 probe (auto bed compensation) my bed.g file contains this:
M98 Pdeployprobe.g
; Probe the bed at 5 points
G30 P0 X25 Y15 H0 Z-99999
G30 P1 X25 Y175 H0 Z-99999
G30 P2 X185 Y175 H0 Z-99999
G30 P3 X185 Y15 H0 Z-99999
G30 P4 X100 Y100 H0 Z-99999 S
M98 Pretractprobe.gMy mesh leveling macro is as follows:
G91
G1 Z6 F200
G90
M98 Pdeployprobe.g
G29 S2
M557 X20:180 Y20:180 S40
G29
M98 Pretractprobe.gI have contacted Antclabs, but if anyone has any other suggestions please let me know! Thanks
-
Do you have to define the servo pin anywhere? I'm thinking back to the way you use pins_Rumba.h to define the pins used in marlin, is there anything similar.
-
The servo pin is defined by the P3 parameter in the M280 commands in the deployprobe.g and retractprobe.g files.
-
Excellent! Do you define the bed limits for probing anywhere? Does this all currently look on track?
Config.g:
; *** BLTouch and Compensation
M558 P5 X0 Y0 Z1 H3 F200 T5000 ; BLTouch Z probe, used for homing Z axis, dive height 3mm, probe speed 200mm/min, travel speed 5000mm/min
G31 P1 X-25.0 Y58.0 Z-1.0 ; Set the Z-probe position and thresholdBed.g:
G28 ;home all
G91 ;relative mode
G1 Z4 F200 ;move Z up 4mm for clearence
G90 ;absolute mode
M98 Pdeployprobe.g ;call macro
G29 S2 ;clear bed compensation
G29 ;run auto bed mesh level
M98 Pretractprobe.g ;call macro
G1 X10 Y10 F5000 -
this is from my own config
M557 X20:290 Y60:280 S60 ;probe boundaries with roughly 16 points (290-20=270/4=67.5 280-60=220/4=55)bounding box of x20 to x290, y60 to y280 with 60mm spacing
-
Deckingman said "I currently use DC42s excellent probe" what is this probe? I am in the market for my next printer…
