IR Mini Z Sense Height?



  • Hello,

    HW Config

    • AnyCubic Linear Plus delta (seriously unhappy with the quality grade of components and their tech support), with MGN-12 (mm) linear-rails

    • DUET3D EtherNet Ver 1.02

    • PanelDUE w/7" LCD/Touch Panel, mounted in a TenTec slopped instrument enclosure (nice), interconnected using the four wire cable to the DUET3D controller board.

    • DUET3D Controller EtherNet port is attached to my local network & InterNet via a switch (in lieu of a router or EtherNet hub).

    I am surprised at how well the WebControl interface is designed (designer understands Human Factors) and works great too…!!

    SW Config (M115)

    • FIRMWARE_NAME: RepRapFirmware for Duet 2 WiFi/Ethernet

    • FIRMWARE_VERSION: 1.21

    • ELECTRONICS: Duet Ethernet 1.02 or later

    • FIRMWARE_DATE: 2018-03-21

    PanelDUE FW:? (as it arrived from FilaStruder in USa, I seem to recall that they had upgraded the FW at the time I bought, 2017).

    As a general FYI…the printer has never printed, and little by little, I replacing components with better quality parts, eg, to fix all the problems, but it was fun to assemble.

    The question

    How far from the bed is the IR Z Probe (from the edge of the PCB below the IR sensors) designed to detect (at '536')?

    Background

    I milled the AnyCubic Linear+ aluminum effector plate and made an adjustable adapter to mount the IR Z-Probe. The nozzle is about 25mm from the IR sensors.

    Using URL vector:
    https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer
    as a guide, I have attempted to measure three areas adjacent to each tower on the bed print 'edge'. The trigger heights are not the same
    for each of the measurement areas, for instance (after executing G28, M208 S1 Z-3 & M564 S0 & G92 Z0, etc):

    Bed Center: G1 F1800 X0 Y0 Z0
    Towers:

    • XT: G1 X-46 Y-28 Z-18.65 = Z-probe trig: 1.68
    • YT: G1 X42 Y-23 Z-14.60 = Z-probe trig: 1.54
    • ZT: G1 X-1 Y38 Z-13.40 = Z-probe trig: 1.81

    One might notice that the effector height varies significantly as indicated by Z parameter in the G1 commands. Definately bowl-shaped. This was just a first run test to see if there is some mechanical problem that needs fixing.

    However, I went back and checked to see how high the nozzle was located after measuring the trigger height, and I found that all of the nozzle heights are around 3 ~ 7 mm off the bed? Is this normal or is the IR sensor board mounted too low?

    I seem to recall that when I first acquired the IR Z-Probe, there was some information that said something to the effect, that the nozzle should be located around 2 or 3 mm below the sensor PCB.

    Any suggestions will be appreciated.

    Thanks in advance.
    3mm
    ps its a kewl setup, eg, the DUET3D & PanelDUE…the on screen edit/upload of the system files...very convenient


  • administrators

    Is the bed surface just plain aluminium, or is there something on top of it? If it is plain aluminium, is it bright or anodised?

    Please check that when you do G30 S-1 to measure the trigger height, the red LED on the sensor is not blinking rapidly. That would indicate that the phototransistor is saturating, due to either the bed being too reflective or because direct bright sunlight is reflecting off the bed into the sensor.



  • Hi,

    The bed is the back-side of an aluminum substrate bed heater. It does not appear to have been anodized, but seems to be just matt grey aluminum, but, of course the surface is very smooth, and does appear to be a bit reflective.

    Initially I took all the measurements on the surface of the bare aluminum bed, but later in an another measurement test, I covered the bed surface with strips of 7,6cm/3" wide blue-painters tape and then retook the ZT (adjacent 'Z-Tower' area) measurement that I listed. The measurement was essentially the same for the bare aluminum as it was for the tape covered tests.

    Perhaps I'll spray paint it as you've suggested somewhere?, with black BBQ paint & cured in an oven. I don't know if it'll fit, (24cm/9.5" dia) in my homebrew SMD reflow oven or not?

    Re; trigger LED flashing, –I didn't notice it flashing. How fast is fast? Need I hookup a photodiode on a scope probe type fast? Other wise, the answer is 'no flashing seen'.

    Should the nozzle touch the bed just as the LED trig indicator illuminates? I suppose that there is an offset command somewhere that could be used to correct the height difference?

    I very much appreciate the time and effort you give in assisting all of us. You are likely a stranger around home?! 🙂

    walt - 3mm


  • administrators

    If the bed is dull grey aluminium then the sensor should work with it OK, as long as you don't have very bright sunlight (no clouds in front of the sun) reflecting off it into the sensor.

    The LED flashing is designed to be easily visible, it's about 8 flashes per second.



  • Hi,

    All the tests were done inside my den after 9pm. There was no sunlight or bright light during the tests. The room is illuminated by a 17Watt LED luminaire located on the ceiling, however, for the second series of tests, and on a hunch, I covered the entire aluminum bed surface with blue-painter's tape. It now exhibits low reflectance but the results are the same.

    So the question exists, how far below the PCB board edge of the Mini-IR does the Z-Probe sensor detect? 1mm? 3mm? 10 mm? What is the expected detection distance from the photo-transistors? If I have the sensor mounted incorrectly it could be possible that the sensor could detect the surface while the nozzle could be .GT. 3 ~ 7 mm from the surface, true?

    I'm considering replacing the entire effector as the diagonal-rods are terminated using miniature Chinese 'pressed' Heim-Joints/rod-end-bearings that are sloppy pigs, so much so that should I bump the table the effector moves up and down!!! I've been trying to get this AnyCubic running for a long time now as budget and time permits. Apparently the AnyCubic vendor sold me an early Linear-Plus Kossel prototype as a finished 'kit' product, and I haven't been able to cause them elect to assist me with the problems. I tried writing them again last week the first time was six months ago, and their answer this time was to have me view a youtube video for a different model than that which they sold me, and then offered to sell me a new printer. I hear good things about their printers…but so far my experience has been an unpleasant one and their tech-support is at best non-existent. However the frame seems to be ok, its everything else that was cr.ap. But it was fun to build??! I've not got hundreds of Dollars into this $300 delta and it still doesn't run!!

    So, if I employ the DC42 Smart-Effector then the matter of the bed surface type will be inherently resolved AND I will likely be rid of the rod-end bearing sloppiness by replacing them using mag-ball joints. However, since I don't have a running printer, I can't print the necessary fan mount and duct components required. I feel like screaming..

    So? Where are we?

    3mm


  • administrators

    When the target is 3mm glass with black paper underneath, the IR sensor trigger height measured from the bottom edge of the sensor to the board will be between 2.5mm and 3.5mm. With an opaque target, it will be a little higher. I recommend you mount the sensor with the bottom edge about 1.5mm higher than the nozzle, that way the trigger height measured from nozzle to be will be no less than 1mm.

    It's possible to calibrate your delta without a Z probe. Just set the Z probe type to 0 (P0 in M558), set up bed.g as usual, run G32 and follow the prompts in DWC (or in PanelDue if you are using latest PanelDue firmware).



  • Hi,

    Well…I'd rather cause the IR Probe to work. Re; the edge to detection distance...the 2.5mm ~ 3.5mm is how the probe PCB is mounted...above the nozzle. So maybe that painter's tape is still too reflective? Something screwy is happening here? It should be working.

    Also, heh heh, at URL vector:
    https://duet3d.dozuki.com/Wiki/Smart_effector_and_carriage_adapters_for_delta_printer
    h t t p s : / / duet3d.dozuki.com/Wiki/Smart effector and _carriage _adapters _for_delta_printer

    Under Heading, "If you use our design for a print cooling fan attachment, then you will also need:"

    (The two items listed below cite the same URL )

    1 Printed fan bracket Print it yourself, see https://www.thingiverse.com/thing:2467663
    1 Printed fan duct Print it yourself, see https://www.thingiverse.com/thing:2467663

    (your US vendor 'FilaStruder' has copied this verbatim, sans the 'alternative' print component vector.

    FYI

    Thanks for your help…

    3mm


  • administrators

    1. Please check that your IR sensor has only one trigger height. Occasionally we find a sensor has more than one trigger height, the normal one and a second one much higher. So bring a piece of paper up under the sensor and check that it triggers just once.

    2. If the trigger height varies with position on a delta printer with an opaque bed surface, that normally means that the effector is tilting by a varying amount depending on XY position. You can check this by dropping the nozzle so that it just touches the bed or just grips a piece of paper, then use feeler gauges to measure the height of the bottom of the sensor board above the bed surface. It should be the same all over the bed. Another good way to detect tilt is to mount a small circular spirit level on the effector.

    3. The two thingiverse links are the same because that page contains the files for both the bracket and the duct.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.