COMPLETE BigBox + DuetWifi controller

Discussion in 'Guides, Mods, and Upgrades' started by elmoret, Jul 17, 2016.

  1. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    After a hotend mishap by having the fans reversed and causing a massive clog, I have the printerand the duet about 90% configured. Howvwer, I have two issues at the moment:
    1. I have included my config.g file below and it appears that my X and Y are off from what the bed shows. When I set the printer to X80 / Y50, it actually goes to X100 and Y60. I noticed this when I started a print and the hotend was off center when it was in the dump bucket. How do I get the bed dimensions and grid set correctly. Single Extruder Titan.
    2. Possibly related to the above, but I am trying to get G29 bed leveling set up. I have setup the probe using M558 P1, G31 P500 Z1.0, G92, G30 steps in the mini probe sensor according to the link in the Duet Wiki. I then tried to set up a 9 point grid following the G29 steps in https://duet3d.com/wiki/Using_mesh_bed_compensation. When the machine was turned off and on, the whole grid was erased. DO I need to add the M557 mesh grid to the config.g file or should it save using an M500 command? Lastly, when I did set up the 9 points and issued a G29, the sensor went off the bed (-x) when trying to probe the first point and nearly cracked the glass before I shut it off. This might be solved for by fixing #1 above, but I want to be sure.

    Ephemeris modified code is below as his is a dual and mine is a single.

    Code:
    ; *************
    ; config.g
    ; *************
    ; File: config.g
    ;       Configuration file for BigBox 3D printer updated
    ;       with Chase Wichert's Dual Direct Titan upgrade
    ; Type: Duet WiFi configuration gcode file
    ;       This is what configures the printer when it powers up.
    ; Date: February 3, 2017
    ; From: Ephemeris
    ;       2017.01.31 Increased X & Y currents to 1275 mA (75%) from 600 & 800
    ;       2017.02.01 Increased X & Y accelerations from 800 to 1000
    ;       2017.02.02 Clarified heater fault settings
    ;       2017.02.10 Added M376 bed compensation taper
    ;       2017.02.11 Added M581 PANIC button trigger
    ;       2017.02.13 Added M572 pressure advance code
    ;       2017.02.14 Increased maximum instantaneous speed change from E20 to E300:300
    ;       2017.02.16 Added M94 (time until idle)
    ; Note: A few commands, if used, must come after others:
    ;       M558 before G31, and M584 before M350 and M906
    
    ; *** General Settings
    M111 S0                                  ; Debug off
    M555 P2                                  ; Set command responses sent to the USB port to be Marlin compatible
                                             ; Pn = 0 or 1 RepRap_Firmware, 2 Marlin (default), 3 Teacup, 4 Sprinter, 5 Repetier
    
    M550 PBB Pro Titan                       ; Machine name and Netbios name (can be anything you like)
    ;M551 Ppassword123                       ; Optional machine password
    
    G21                                      ; Work in millimetres
    G90                                      ; Send absolute coordinates...
    M83                                      ; ...but relative extruder moves
    
    ; *** Wifi Networking
    ;M540 P0x5c:0xcf:0x7f:0xff:0xff:0xff     ; Your Duet WiFI's actual MAC Address - This is probably unneeded
    M552 S1                                  ; Enable WiFi interface
    
    ; *** PanelDue Touch Display
    M575 P1 B57600 S1                        ; Assign serial channel #1 (P1)
                                             ; Set baud to 57.6k and only accept commands
                                             ; with valid checksum (parameter S1)
                                    
    ; *** Set X,Y,Z Travel Limits
    M208 X300   Y245   Z295                  ; Set axes maxima (mm)
    M208 X-21     Y0    Z-0.2  S1            ; Set axes minima (mm) - S1 parameter means set minima
                                             ; Adjust minimum X and Y to make X=0 and Y=0 the edge of the bed
                                             ; Adjust minimum Z so we can overtravel slightly during Z-probe setup
    
    ; *** Limit Switch Configurations
    M574 X1 Y1 Z0 S1                         ; Set type of each axis limit switch:
                                             ; 0=no switch, 1=low limit, 2=high limit
                                             ; S1 indicates active high inputs
                                             ; (e.g. normally-closed switches)
                                             ; Because we are using an infrared Z-probe
                                             ; instead of a limit switch, Z=0, and the
                                             ; Z-probe is set up below with M558 and G31.
                                          
    ; *** Z-Probe Configuration
    M558 P1 X0 Y0 Z1 H3 F200 T5000           ; Escher3D infrared Z-probe, used for homing Z axis,
                                             ; Set dive height H=3mm, probe speed F=200mm/min, travel speed T=5000mm/min
                                             ;
    G31 X31.7 Y15.3 Z1.05 P500               ; Set the Z-probe position and threshold -
                                             ; Z parameter was measured with paper as 2.40 mm so it should have been 2.5 mm
                                             ; G31 Z parameter set per https://miscsolutions.wordpress.com/mini-height-sensor-board/
                                             ; If you slightly increase the Z parameter Layer 1 will be correspondingly thinner.
                                             ; If layer 1 is too tight, decrease the Z parameter to loosen
                                             ; X and Y parameters define position of Z sensor with respect to left nozzle (T0)
                                             ; Was G31 X12.0 Y26.0 Z1.34 P500 before switching to SMD IR probe and twist mount
                                             ;     G31 X34.5 Y16.0 Z1.34 P500 measured with white cube
    
    ; *** Panic Button                       ; Assign an action for when Z stop switch closes (not used as endstop because Z-probe is installed)
                                             ; Switch is normally open so trigger on falling edge (S0) when switch closes
                                             ; instead of rising edge (S1, default), or ignore input (S-1)
                                             ; C0 = trigger at any time (default) not C1 = only when printing a file from SD
    M581  Z:S0 T0 C0                         ; Trigger 0 - Emergency stop. Same as if M112 was received
                                             ;           - Immediately turns off all motors and heaters
    ;M581 Z:S0 T1 C0                         ; Trigger 1 - Pause print     Same as if M25  was received,
                                             ;           - Before pausing, pause.g will execute
    ;M581 Z:S0 T2 C0                         ; Trigger 2 - User defined stop
                                             ;           - Executes sys/trigger2.g
    
    ;     _______________________________________________________________________________
    ;               Stepper          Step Size   Holding   Max Current       Phase
    ;               Model            (degrees)   Torque     per Phase      Impedance
    ;     Extruder  OMC 17HS08-1004S    1.8     13.0 N cm   1.0 amps    3.5 ohm, 4.5 mH Not Used.  Mini Stepper
    ;     Extruder  Wantai 42BYGHM809   0.9     40.0 N cm   1.7 amps    1.8 ohm, 4.0 mH
    ;     X-Axis    Wantai 42BYGHM809   0.9     40.0 N cm   1.7 amps    1.8 ohm, 4.0 mH
    ;     Y-Axis    Wantai 42BYGHM809   0.9     40.0 N cm   1.7 amps    1.8 ohm, 4.0 mH
    ;     Z-Axis    Wantai 42BYGHW609   1.8     40.0 N cm   1.7 amps    1.5 ohm, 2.8 mH
    ;
    ;     DC42 recommends running steppers at 50% to 85% of their current ratings, he also
    ;       said setting current below 30% of stepper RATED current causes carriage jumping.
    ;     Due to low torque, pancake steppers in the extruders are set to run at 100%
    ;     _______________________________________________________________________________
    
    ; *** Stepper Polarity
    ;     The stepper connectors on Duet WiFi are labeled, from left to right,
    ;     Red-Blue-Green-Black. My corresponding stepper cables are all
    ;     Blue-Red-Green-Black
    M569 P0 S1                               ; X-axis (Drive 0) goes forwards
    M569 P1 S0                               ; Y-axis (Drive 1) goes forwards
    M569 P2 S0                               ; Z-axis (Drive 2) goes forwards
    M569 P3 S1                               ; Extr-1 (Drive 3) goes backwards
    ;M569 P4 S1                               ; Extr-2 (Drive 4) goes backwards
    
    ; *** Set the microsteps/step and microsteps/mm for each motor
    M350 X16   Y16   Z16   E16   I1      ; Set microstepping to 16x for motion and 16x for extrusion
                                             ; The I parameter turns 256x interpolation on (1) or off (0)
                                             ; Interpolation to 256x microstepping only works if you set
                                             ; the microstepping to 16x and use the I1 parameter.
                                             ; The I1 parameter does nothing for any other microstepping value.
    
    M92 X160  Y360 Z1600  E880               ; Set microsteps/mm for each stepper motor
                                             ; X motor is 400 steps/rev driving a 20 tooth GT2 pulley
                                             ;    So expect 16/(1/400*20*2) = 160 microsteps/mm
                                             ; Y motor is 400 steps/rev driving a 16T GT2 pulley that belt drives a 36T
                                             ; on the Y drive shaft which has 20T pulleys that drive the Y carriages
                                             ;    So expect 16/(1/400*16/36*20*2) = 360 microsteps/mm
                                             ; Z motors are 200 steps/rev driving 2 mm pitch, 2 mm leadscrews
                                             ;    So expect 16/(1/200*2) = 1600 microsteps/mm
    
    ; *** Set stepper motor drive parameters
    M906  X800  Y600   Z800 E1000 I30        ; Set stepper motor currents (mA) and idle current to 70%
    ;M906  X1275  Y1275   Z800 E1000 I30     ; Set stepper motor currents (mA) and idle current to 70%
                                             ; I parameter is the percentage by which the currents to all steppers
                                             ; are reduced when the printer is idle. (Idle time is set by M84)
    M84   S300                               ; Reduce current to all steppers after 300 seconds of inactivity on all axes.
                                             ; Use S0 to disable the idling feature altogether.
                                             ; The idle current is set above by the M906 I-parameter
    M201  X1000  Y1000    Z15 E1000          ; Accelerations  (mm/s^2)
                                             ; X and Y can be much quicker (2000+) but noise was objectionable at 1200
    M203 X15000 Y15000   Z360 E3600          ; Maximum speeds (mm/min) Marlin had Z at 360 mm/s
    M566 X600     Y600    Z30 E300           ; Maximum instant speed changes mm/minute (sometime miscalled 'jerk')
                                             ; Alex9779 uses M566 X480 Y480 Z24 E300:300 Elmoret had E20
                                             ; DC42 once said in a discussion of pressure advance (M572) that
                                             ;    "200mm/min extruder jerk is unusually low, so the associated acceleration
                                             ;     limit for printing moves that it causes might have an effect."
    
    ; *** Set thermistor and ADC parameters  ; The Duet WiFi needs no H or L parameters "because the ADC is calibrated automatically"
    
    M305 P0 T100000 B4300 R4700              ; First generation (2016) E3D Varipower bed uses a Semitec 104GT2 thermistor
                                             ; Datasheet says thermistor resistance @ 25 C = 100 kohm, beta = 4300 K
                                             ; Duet WiFi has series resistance of 4700 in the measurement circuit.
    
                                             ; Hotends are instrumented with E3D PT100 platinum resistance temperature detectors (RTD)
                                             ; They are wired into a Duet PT100 temperature sensor daughter board that uses a
                                             ; Maxim MAX31865 RTD-to-Digital Converter chip. The daughter board outputs are
                                             ; routed to channels 200 and 201.
    M305 P1 X200                             ; PT100 Tool 0 (left) , RTD1, channel 200
    M305 P2 X201                             ; PT100 Tool 1 (right), RTD2, channel 201
    
    ; *** Set Maximum Heater Temperatures
    M143 H0 S135                             ; Maximum temperature - Heated bed    , default is 125C
    M143 H1 S290                             ; Maximum temperature - Left  extruder, default is 262C
    M143 H2 S290                             ; Maximum temperature - Right extruder, default is 262C
    
    ; *** Set Heater Fault Time Limits       ; A fault will be declared if a heater can't reach a commanded
                                             ; temperature within the number of seconds given by S parameter.
                                             ; H parameter is heater number (bed is 0, extruders are 1 and 2)
                                             ; P parameter sets how many seconds a temperature anomaly may persist
                                             ;   before a heater fault is declared (default is 5 seconds)
                                             ; T parameter specifies the permitted temperature excursion
                                             ;   from the heater setpoint (default 10C)
    
    M570 S360  H0  P5  T10                   ; E3D Varipower heated bed with Clever3D PEI/Al bed, takes take 5:30 to reach 130 celsius.
                                             ; Temperature climb from room temperature in one minute steps: 58, 84, 102, 117,128 C
      
                                             ; Without socks on, hot-ends take 2:37 to reach 285 celsius. 180 seconds should be generous
    M570 S180  H1  P5  T10                   ; Left  extruder (T0)
    ;M570 S180  H2  p5  T10                   ; Right extruder (T1)
    
    ; *** Tool (Extruder) Definitions
    M563  P0 D0 H1                           ; Define tool 0 (Left  extruder) as using drive 0 and heater 1
    G10   P0 S0 R0                           ; Set tool 0 operating and standby temperatures to zero celsius
    
    ;M563  P1 D1 H2                           ; Define tool 1 (Right extruder) as using drive 1 and heater 2
    ;G10   P1 S0 R0                           ; Set tool 1 operating and standby temperatures to zero celsius
    
    ; *** Pressure Advance                   ; Compensates for elasticity of extrusion system
                                             ; extrusion_speed = requested_extrusion_speed + (S * current_extruder_acceleration)
                                             ; The more the system can lag, the greater the compensation required
                                             ; Typically S = 0.05 for a direct extruder
                                             ;           S = 0.10 for a Bowden with a 400 mm tube
    M572  D0 S0.05                           ; Set pressure advance coefficient for tool 0 (Left  extruder)
    ;M572  D1 S0.05                           ; Set pressure advance coefficient for tool 1 (Right extruder)
    
    ; *** Assign Fan Channels
    M106  P0 F10                             ; Part cooling fan (P0 is PWM Fan 0 on the Duet WiFi)
                                             ; Some found this blower style fan's speed could not be reduced much
                                             ; without it stopping. Doing as Tim Elmore suggested and lowering
                                             ; the PWM frequency from 250 to 10 Hz per.
                                             ;
                                             ; Control the hot-end fans thermostatically in bang-bang mode.
                                             ; E3D recommends turning the hotend fans on at 40 celsius.
    M106 P1 T40 H1                           ; Trigger hotend 1 (left) fan at 40 C  (P1 is PWM Fan 1 on the Duet WiFi)
    ;M106 P2 T40 H2                           ; Trigger hotend 2 (right)fan at 40 C  (P2 is PWM Fan 2 on the Duet WiFi)
    
    ; *** Load bed height map and activate bed compensation
    ;G29 S1                                   ; Load the height map from the last time you used G29 to probe the bed.
                                             ; The default height map file is /sys/heightmap.csv.
                                             ; Comment this out if you are not using automated mesh leveling.
                                             ;
    ;M376 H10                                 ; Taper off bed compensation taper
                                             ; Hnnn Height (mm) over which to taper off the bed compensation
                                             ; Setting H0 leaves compensation always on
                                          
    ; *** Compensate for out of square axes  ; Assuming you printed a cube with side lengths given by S parameter
                                             ; Measure the distance by which the printed
    ;M556 S100 X0 Y0 Z0                       ; Snnn size of test article
                                             ; Xnnn is how far the printed X-axis is from being perpendicular to the printed Y-axis
                                                    This is positive if the angle between the printed X and Y is greater than 90
                                             ; Ynnn is how far the printed Y-axis is from being perpendicular to the printed Z-axis
                                                    This is positive if the angle between the printed Y and Z is greater than 90
                                             ; Znnn is how far the printed Z-axis is from being perpendicular to the printed X-axis
                                                    This is positive if the angle between the printed Z and X is greater than 90
    
    ; *** Set initial bed temperature (to get rid of the -273C reading)
    M140 S0
    
    ; *** Select the left hotend as the default active tool
    T0                                       ; Select left hotend
    Any assistance or guidance is appreciated. So close I can almost smell the PLA melting!!


    Edit: There are option on the Machine Control page for auto bed and setting up the mesh, but I could not find any instructions on their use. Is it easier to use those than to try to set it up manually?
     
    #321 Steven Burns, Aug 24, 2017
    Last edited: Aug 24, 2017
  2. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    1. Set axes minima and maxima correctly and adjust your end stops. Take care with your Z-safe levelling spot (if you're still using head-mounted IR) and choose somewhere definitely safe, like say, 100,100.
    2. Once you're sure you've got your X Y maxima and minima set properly and end stops are adjusted to your liking, change your Z safe spot (actually I think this is not specified in config.g but in your homez.g and/or homeall.g command)
    3. Look for the G31 command. This is important for Z levelling. This command defines where in relation to the nozzle, the IR sensor point is. So, if your IR sensor is to the left of the nozzle (which I suspect), then G31's X parameter will be negative. For me, X is negative, and Y is 0. I suspect my Y is not absolutely zero, but I don't see a simply way to measure the actual trigger point as I don't see in IR. In any case, you're trying to get this roughly right, so don't get OCD about the millimetres (unless you want to).

    Finally, go and follow (perhaps even print out) Eph's guide for mesh levelling.

    I think the machine control options are essentially just providing quick access to the commands Eph details. If you go through his guide first, you'll appreciate all the steps and can make a decision about those options. Once you have things set, you'll probably only rerun G29 from time to time; say you change the nozzle or the bed.
     
    Steven Burns likes this.
  3. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    I am very close now with only one (I believe) issue left. I am going to post snippets of my code if that will help:

    Code:
    ; Mesh Bed Leveling Macro
    M557  X60:250  Y50:170 S40   ; Define a  6x4 point grid for measuring the bed
    G28 Y                        ; Always home Y axis first to avoid dock crash
    G28 X                        ; Always home X before Z
    G28 Z                        ; Always home Z last
    G29 S0                       ; Run probing sequence defined by M557 above
    M374                    ; Save calibration data. Default is sys/heightmap.csv
    G1 X87 Y200 F3000           ; move to position in front of dock
    G1 X87 Y240 F3000           ; move into dock
    Config.g Snippet
    Code:
    ; *** Set X,Y,Z Travel Limits
    M208 X300   Y245   Z295                  ; Set axes maxima (mm)
    M208 X-68   Y0     Z-0.2  S1             ; Set axes minima (mm) - S1 parameter means set minima
                                             ; Adjust minimum X and Y to make X=0 and Y=0 the edge of the bed
                                             ; Adjust minimum Z so we can overtravel slightly during Z-probe setup
    
    G31 X60.0 Y15.3 Z0.83 P500                ; Set the Z-probe position and threshold -
                                             ; Z parameter was measured with paper as 2.40 mm so it should have been 2.5 mm
                                             ; G31 Z parameter set per https://miscsolutions.wordpress.com/mini-height-sensor-board/
                                             ; If you slightly increase the Z parameter Layer 1 will be correspondingly thinner.
                                             ; If layer 1 is too tight, decrease the Z parameter to loosen
                                             ; X and Y parameters define position of Z sensor with respect to left nozzle (T0)
                                             ; Was G31 X12.0 Y26.0 Z1.34 P500 before switching to SMD IR probe and twist mount
                                             ;     G31 X34.5 Y16.0 Z1.34 P500 measured with white cube
    So this is my meshlevel macro my M208 and G31 commands. I have the bed now probing without it being off the bed. My initial issues was that all of the examples seem to be from dual machines with the left nozzle obviously much closer to the sensor. As it stands the X-68 is actually overstated as the sensor to nozzle gap is closer to 60-61. If I set it to that then the sensor is not over the bed.

    The issue now lies in the docking code. The carriage is now way past the right side of the dock so it crashes into the back. I know there is just a small set of changes left to balance this out so that I am over the bed for probing AND still within the dock range when the probing, or printing, is complete.

    Edit: I realize that I could edit the dock code to move it over -X xxx mm, but I should not need to right?
    What confuses me even more is that my S3D End Script is:
    ;move to dock

    T0 ; select extruder 0
    G1 X95 Y200 F3000 ; move to position in front of dock
    G1 X95 Y235 F3000 ; move into dock

    Which is even farther to the right than the macro code.


    Thoughts?
     
    #323 Steven Burns, Aug 26, 2017
    Last edited: Aug 26, 2017
  4. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    There's nothing sacred about the dock position specified in the start and end scripts. Adjust them to your needs. I recommend Homing the machine, then using the Machine Control Tab manually move the head until you find a good entrance point for the dock. Note the X, the Y will be 200. Now Dock it (G1 Y240), and then move the X until you're happy with the new exit position. Note that, then go and update your scripts, which are now personal to you and your machine. Save your changes in a factory file so you can return to this starting point any time in the future.
     
  5. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    I understand, but why would they have changed? My existing S3D scripts have worked fine for hundreds of prints on the Rumba.

    I guess my real question is my M208 or G31 commands off? Should I keep the G31 at -60 as that is the distance between the nozzle and the sensor and the put the M208 to 0 or 1?

    I'm just confused it would be so different between the Rumba and the Duet as far as coordinates go. Something has to be off....
     
  6. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    You're right. X,Y shouldn't really have changed much, and 95,200 sounds like a sensible entry point for the dock. Here's a question ... after you home if you try this ... G1 X0 Y0, do you actually have the head at where you think 0,0 should be?
     
    #326 Spoon Unit, Aug 26, 2017
    Last edited: Aug 26, 2017
  7. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    G1 X0 Y0 was not sitting at 0,0 so I changed the config.
    M208 X300 Y245 Z295
    M208 X1 Y1 Z-0.2 S1

    G31 X60.0 Y15.0 Z0.75 P500

    Mesh is set as follows:
    M557 X60:250 Y50:170 S40

    Now when I do the G1 X0 X0 it is sitting at 1, 1, which is fine for now.

    When I start the meshlevel macro, the first probe point is at X40, Y40 on the bed grid markings.

    So, M208 is setting the nozzle to the correct point with respect to the bed, but now the sensor is back to hanging off the left side of the bed and will crash if I do not use the panic button.

    I can't imagine that I would need to over compensate on the G31 and set X60 to X100 and Y15 to Y 55.
     
  8. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    I don't think you're quite there yet. Your lines say this to me ...

    M208 X300 Y245 Z295

    When you push your head all the way to the right, the nozzle is over X=300 (I'm guessing it's actually further to the right than this)
    When you push your head all the way to the back, the nozzle is at 245 (this is probably right, when it's in the dock)
    If you move your bed all the way to the bottom, Z is 295, which is probably safe, and in fact, it probably could move further if you have no obstructions.

    M208 X1 Y1 Z-0.2 S1

    When you push your head all the way to the left, X is at 1 (this is possible, but I suspect it's not true. If it true, you might have an obstruction of some sort, could be cables, or maybe the end stop is unscrewed too far)
    When you pull your head all the way to the front, Y is at 1 (ideally you should screw your y end stop screw in a little so you can get to Y=0
    And the furthest you can move your bed is Z=-0.2 (I find this a dangerous idea, as once you have things tweaked correctly, this will mean the bed is pushing the whole head up 0.2, this might be fine for a temporary override if you're sorting out where Z=0 actually is, but don't leave it this way I think, but ultimately your choice, your machine)

    G31 X60.0 Y15.0 Z0.75 P500

    Your IR sensor is positioned 60mm right (really) of the nozzle, and 15mm further back (could be), when the nozzle is touching the bed, the IR is 0.75mm higher than the bed (for me this is really too close, you should be able to move it up somewhat from here ... if you follow Eph's guide you'll end up with it at 1.4 to 1.9 or so), and will trigger when the reading is 500 (mine's the same)

    M557 X60:250 Y50:170 S40

    That's a very safe mesh. Fine as a starting point. You could extend this out a little to cover more of the bed once it works properly.

    I don't understand your point about the IR hanging off the edge of the bed? Is this happening during G29? If so, I suspect your IR position is not configured correct with G31. Definitely hit the panic, but once you have this set right, you can actually set a grid too large and the Duet (at least on my firmware) will disregard attempts to probe off-bed).
     
  9. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    @Spoon Unit

    You helped me solved it. When you said the sensor was to the right "(really)", it made me realize that I had that backwards. The line should have been:
    G31 X-60.0 Y15.3 Z0.83 P500

    When I changed it the head did not crash and the mesh routine kicked off without an issue and it docked without any problems once I changed it to X95 versus X87.

    Thank you!!!
     
    Spoon Unit likes this.
  10. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    Sorry to be a pain, but I have a question...

    I was getting very bad reading from the sensor on the glass bed due to sagging (even with the knurled knob support) so I thought I would give BuildTak a try. That way I get a nice flat bed with consistent readings all across the bed surface. Or so I thought.

    I installed the BuildTak and then started running through the sensor calibration for the mini sensor. M558, G31, G92 (after getting the nozzle close), but the sensor WILL NOT TRIP. I put a piece of paper under the sensor and the light comes on. I put a piece of builtak under the sensor and nothing happens, even when the sheet is making direct contact with the underside of the sensor.

    I know folks are adding aluminum beds with PEI, but after dropping all of his money on the upgrade, this is all I can afford.

    Any thoughts as to why the sensor will not trip using black BuildTak? @dc42

    EDIT: I tried sanding a piece of the sheet with 400 grit sandpaper and then cleaning with alchohol, but it still will not trip the sensor at any height. Very frustrating!!

    Edit #2: After looking at the sensor I have, it appears to be different than the current sensor sold. The current sensor has small square LEDs next to the capacitor (I am assuming that's what it is) while mine has more convenstional round taller LEDs. Could this be an issue with the older sensors. I thought I read something about this on the Duet forum or online somewhere. Should I buy a new sensor? I only want to try this is I am fairly sure it is going to resolve the issue.
     
    #330 Steven Burns, Aug 31, 2017
    Last edited: Aug 31, 2017
  11. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    Hi Steve,

    I don't think sagging will affect the ability of the IR to sense the glass. People have had issues with the glass bed reflecting, the challenge is compounded with an auto-bed-level process which then expects an equally accurate reading at all points. If you're not getting a reading from the untreated surface, it might be worth trying a thin coating of UHU, but of course this is going to potentially impact the results. If you're careful though, you'll have a thin covering and have a completely opaque surface, which should then reflect.

    I'm surprised BuildTak doesn't work, but the surface isn't even on BuildTak, and given the matt finish, you can imagine the optical characteristics are going to be to reflect light in many directions, which is also going to affect the IR reflection. Just found this from dc42 re. BuildTak (http://forums.reprap.org/read.php?340,486842). One of the things he suggests there is, when not using BuildTak, put a thin sheet of black paper behind the glass to control reflections.

    The PEI plate definitely reflects consistently across the whole bed.

    I also just pulled out one of my original glass plates and added PrintBite. I ordered this stuff a while back, but it was delayed due to them making a new formulation. I have the black, matt surface, and this runs as well as the PEI plate from the perspective of IR bed leveling.
     
  12. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    I tried without glue and with a heavy coating of glue and the results were the same on bare glass. The height map was looking like there were giant holes in the glass even though I had manually leveled it to within a business card thickness across the entire bed. If I supported the grid under the bed just enough to reduce the sag then my readings changed drastically. This tells me that the underlying heat bed was having an impact on the readings and that the sensor was not truly looking only at the glass. This is not typical behavior.

    So now with the buildtak as the primary surface above the glass I should be getting stable readings across the entire surface, but instead the sensor is not picking up the new surface at all. This leads me to believe that the sensor is bad.

    I purchased a new one from Filastruder and will see what happens when I get it. I suspect that, based in the article you posted that I had read last night, that the sensor I had was early enough that it was one of the ones that would not detect the new print surface.
     
  13. Springfield-Jack

    Springfield-Jack Well-Known Member

    Joined:
    Dec 20, 2015
    Messages:
    67
    Likes Received:
    6
    I don't know if will be any help but I had a problem with the sensor,it would react to everything but the glass bed.I eventually solved it by very carefully adjusting the position of the LEDs.(instructions on the troubleshooting page). They had been knocked out of alignment at some stage.
     
  14. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    Couldn't hurt. At this point all options are good options.
     
  15. dc42

    dc42 Well-Known Member

    Joined:
    Aug 16, 2016
    Messages:
    526
    Likes Received:
    221
    The more recent design of IR sensor has a wider dynamic range than the original design, so it can handle both dark surfaces like buildtak and bright surfaces better. Even so, I am surprised that the original sensor isn't working with buildtak, because it is one of the surfaces that I tested them with.
     
  16. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    While I am glad that I ordered a new sensor, I did finally manage to get it to work with the Buildtak. It took quite a bit of trial and error moving the outside led and I had to reduce the sensor height to 1mm from the bed.

    I am happily printing my first test piece with the Duet. I cannot believe how quiet it is. The fans make more noise than anything else.

    Thank you, everyone, for all of your help!!! I truly cannot thank you enough.
     
  17. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    589
    Likes Received:
    100
    With more recent design. Are you talking about the sensors with SMD ir components or a newer revision?
     
  18. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,561
    Likes Received:
    483
    You can tune those somewhat too :)
     
    Steven Burns likes this.
  19. Steven Burns

    Steven Burns Well-Known Member

    Joined:
    Sep 4, 2015
    Messages:
    195
    Likes Received:
    55
    The SMD version, I believe.
     
  20. dc42

    dc42 Well-Known Member

    Joined:
    Aug 16, 2016
    Messages:
    526
    Likes Received:
    221
    Correct.
     
    Stian Indal Haugseth likes this.

Share This Page