bltouch probe

[jesse]

I dusted off my Rostock Max v1 from Indiegogo 2012. The arms are a bit loose and move a few millimeters if I jostle the effector.

It has Smoothieboard 32-bit electronics.

I've been trying to get auto-leveling to work. Adjusting the screws on each tower doesn't work well enough anymore.

I added a bltouch v3.1 to the Smoothieboard and have the process working. However, it does not level the bed.

Code: Select all

G28; #home to max endstops
M280 S3.24; #deploy the probe
G32; #3-point delta calibration
M500; #save results of 3-point[attachment=0]IMG_6681.JPG[/attachment]
G31; #delta grid probing
M374; #save results of delta grid
M280 S7.36; #retract the probe

I mounted the probe in one of the corner holes of the triangular effector. It's about -50mm on the Y-axis, relative to the bed. The maker of Smoothieboard said that the probe needs to be in the same position as the hot-end in the X,Y to get accurate results.

Anyone have advice? The bed is still off about 1mm when moving from X100 to X-100 and Y-100 to Y100.

[geneb]

Oblig: "That thing belongs in a museum!"

Is the firmware you're running have a good auto-calibration system for delta printers? I don't know anything about it.

g.

[jesse]

Yes, the Smoothieboard firmware appears to be trying to do the correct operations. It's the newest firmware version.

Maybe I should measure the distance between the effector and the bed to see if it's parallel. Another idea is to make the leveling screws exactly the same height before trying to auto-level.

Here are the details of auto-leveling on the Smoothieboard: http://smoothieware.org/zprobe

GcodeV31.png (220.98 KiB) Viewed 14475 times

I was able to get to the delta leveling grid that saved the SD card.

notepad++_5OhU8DesjX.png (80.15 KiB) Viewed 14475 times

Unfortunately it's in a serial binary format and I don't have the layout. If I can decode/encode the table, I might be able to put my own measurements to correct for the high/low spots.

Oblig: "That thing belongs in a museum!"

Yes, it's the first one I bought. I'm trying to get it working for my Etsy store: https://www.etsy.com/shop/FantasticGadgets. Then I can sell 9 and 10 inch labyrinths.

[geneb]

I would really recommend you upgrade to the injection molded carriages and the ball & cup arms. The carriages are clear so they'll match the asthetic of the machine and the carriage/arm combination will take a dramatic amount of "slop" out of the system. Reach out to SeeMeCNC and they should be able to supply you with one of the v2 effector platforms that are designed for the ball assemblies - that way you won't have to fab your own.

Doing auto-cal with all three adjustment screws in the same place is the way to go.

g.

[jesse]

I think the autocalibration is needed because my bed is not flat.

I replaced the original MDF wooden fractal star with a solid disc of aluminum, thinking that it would help.

Maybe the acrylic has sagged over time or the aluminum disc is a little raised in the middle.

IMG_6699.JPG (332.99 KiB) Viewed 14464 times

IMG_6701.JPG (611.77 KiB) Viewed 14464 times

I put springs under each of the 6 mounting points and that helped enough for 0.5mm layers but not for 0.25mm layers.

Steve from SeeMeCNC sent me a few links to buy the parts to fix the looseness in the arms which will cost about $100 + shipping.

[geneb]

I suspect those springs are your problem. Even the Rostock MAX v0 didn't use springs under the bed - it was nylon spacers.

g.

[jesse]

I turned off auto-calibration and tried just using only the springs. It worked to about 0.5mm layers but not 0.25mm layers.

The maker of Smoothieboard (electronics) said that the auto-level probe is supposed to be in the same spot as the hot end in order to make an accurate delta probe grid. So I may need to 3d print something so that the bltouch can be mounted under the effector temporarily and then removed.

[geneb]

Ah, ok. This is why the SE300 or Duet Smart Effector is such a good choice - the nozzle IS the probe. I don't know if Smoothie board supports those or not though. (the SE300 uses the same circuit as the Duet Smart Effector.)

g.

[jesse]

Not a bad idea about changing to electronics that can detect acceleration of stopping on the bed. I have some TMC2208 and another electronics board where that is possible.

Here's what I came up with as far as a mount for the probe. I put a shroud around the probe so that it can't slam into the bed if something goes wrong.

IMG_6743.JPG (557.9 KiB) Viewed 14441 times

IMG_6744.JPG (553.98 KiB) Viewed 14441 times

openscad_prz8FjnGQo.png (147.2 KiB) Viewed 14441 times

openscad_GHwM5royJG.png (129.57 KiB) Viewed 14441 times

openscad_6YyFwueBKZ.png (111.68 KiB) Viewed 14441 times

[geneb]

Does it work well for you that way?

g.

[jesse]

Yes, it works now, mostly. The printer can make a 9-inch cylinder with 0.25mm layers.

A 10-inch cylinder would be nice, but it doesn't work. Not sure if it's because of the limits of the printer geometry.

The goal was to be able to sell larger labyrinths in an Etsy store. My Prusa clone maxes out at 8 inches.

Settings added to Smoothieboard electronics "config" text file (below).

Code: Select all

zprobe.enable                                true            # set to true to enable a zprobe
zprobe.probe_pin                             3.25            # pin probe is attached to if NC remove the !
zprobe.slow_feedrate                         30               # mm/sec probe feed rate
#zprobe.debounce_count                       100             # set if noisy
zprobe.debounce_ms                           1
zprobe.fast_feedrate                         300             # move feedrate mm/sec
zprobe.probe_radius                          127             # where to probe
zprobe.probe_height                          74               # how much above bed to start probe

switch.servo.enable                         true
switch.servo.input_on_command               M280           # M280 S7.5 Would be midway
switch.servo.input_off_command              M281           # Same as M280 S0 0% duty cycle, effectivley off
switch.servo.output_pin                     3.26                # Must be a PWM capable pin
switch.servo.output_type                    hwpwm               # H/W PWM output settable
switch.servo.pwm_period_ms                  20                  # Set Period to 20ms (50Hz) - Default is 50Hz

leveling-strategy.three-point-leveling.tolerance   0.05    # the probe tolerance in mm, default is 0.03mm
leveling-strategy.delta-calibration.enable   true            # basic delta calibration
leveling-strategy.delta-calibration.radius   127             # the probe radius
leveling-strategy.delta-grid.radius        127
leveling-strategy.delta-grid.enable        true
leveling-strategy.delta-grid.size        15
leveling-strategy.delta-grid.do_home         true
leveling-strategy.delta-grid.probe_offsets  0,0,0  # probe offsets x,y,z  (Z should always be 0)
leveling-strategy.delta-grid.save        true
leveling-strategy.delta-calibration.initial_height 100        # height above bed to stop initial move
leveling-strategy.delta-grid.initial_height 100

The probe is attached to the effector using a set of three M3 screws and wingnuts.

I'm going to print this part on my resin printer next to prevent the filament from buckling in Steve's extruder: https://forum.seemecnc.com/forum/viewto ... 6&p=100698

[geneb]

Those labyrinths are great!

You should be able to print to nearly the edge of the glass, even on the "v0" when it's calibrated properly. What problems are you seeing?

g.

[jesse]

Those labyrinths are great!

Thanks, the manufacturer of the green glitter filament is out-of-stock for a while. I haven't found a replacement.
https://hobbyking.com/en_us/pla-premium ... yking.html

Tried:

• https://www.proto-pasta.com/collections ... reen-htpla
  Sparkles are too small

• https://www.amazon.com/ZIRO-Filament-1- ... 08LDMTZN4/
  Green is too light