Endstop on one of the axis all time on hit state



  • Hi, I am new in duet world.
    Three days ago I completed replace old MKS Sbase by Duet WIfi on Tevo Little monster.

    There is one problem that I can not resolv. I can't homed delta because Z-axis endstop all time said triggered state and carriage don't move. But I can manuly move Z-axis via G1 commands.

    In this case, the electronics of the board and the LED on endstop indicate that no operation is triggered. With the coin test, the endstop is triggered(LED on both side off) , but the state in the microcode (GCODE M119) does not change:

    Endstops: X: not stopped; Y: not stopped; Z: at max stop; Z probe: not stopped.

    I tried to replace the wires, replace the optical sensor, switched Z-motor to E1. Nothing helps.

    0_1534550259927_config.g

    config.g in attachment. I don't know what I can to do yet.

    Firmware: 2.01
    DuetWifi: 1.02

    Can you help me?


  • administrators

    First, check that you have the correct M574 command in config.g. For a delta printer using active high optical endstops, it should be: M574 X2 Y2 Z2 S1.

    If you have the correct M574 line then my guess is that the optical endstops you are using are marginal when used with a 3.3V supply. It's possible for an endstop to pull down the endstop input enough to light the endstop LED at less than full brightness, but not enough for it to register as a low logic level to the microcontroller. If you very carefully short the outer two pins of the Z endstop connector together, you will probably find that the brightness of the LED increases and M119 changes from Z triggered to not triggered.

    If this is the case, a solution is to reduce the value of the opto switch LED series resistor on the optical endstop. See https://duet3d.dozuki.com/Wiki/Connecting_endstop_switches#Section_3_3V_compatible_optical_endstop for more.



  • David, thanks for the answer.

    I read the wiki before.

    I do not think that the problem is in the endpoints themselves. I changed them in a circle and the problem always remains on the Z axis.

    In order to completely eliminate the problems with the endstops, I decided to simply simulate them with two wires. Unfortunately, the opening and closing of the wires did not affect the status of the program trigger.

    I also came to the conclusion that the remap of the motors between the axes in the firmware does not lead to a remap of the terminal endstops. Moreover, I did not find in the firmware to make endstops remap.

    Do I need re-compile firmware for endstop remap?


  • administrators

    Remapping motors with M584 does not remap the endstops. Your Z endstop must be connected to the Z endstop connector even if you have connected the Z motor to a different motor output.

    Shorting the two outer pins of the Z endstop connector together should change the status from triggered to not-triggered in both M119 and in the Machine Properties page of DWC. Did you do that test?

    All the endstop inputs are tested before a Duet leaves the factory, but it is not impossible for a fault to develop subsequently. In the case of the Z endstop input, if the endstop LED is working then the most likely cause of such a fault would be resistor R58 missing, cracked or not properly soldered.



  • Shorting the two outer pins of the Z endstop connector together should change the status from triggered to not-triggered in both M119 and in the Machine Properties page of DWC. Did you do that test?

    Yes, I checked it. Unfortunately, the state of the trigger did not change, so I wanted to pimp the pins to the neighboring endstop.

    Also, I measured the voltage on the outer pins of all the endstops and found out that on the pins of the Z axis, the voltage is below 1.8V. On all the others it is 3.2-3.3V.

    Apparently there is a faulty component on the board.



  • @dc42 said in Endstop on one of the axis all time on hit state:

    In the case of the Z endstop input, if the endstop LED is working then the most likely cause of such a fault would be resistor R58 missing, cracked or not properly soldered.

    R58 on the place and has ~1kOm


  • administrators

    Does the Z endstop LED light up (like the X and Y ones) when the endstop is plugged in but not triggered?



  • @dc42 said in Endstop on one of the axis all time on hit state:

    Does the Z endstop LED light up (like the X and Y ones) when the endstop is plugged in but not triggered?

    Yes, LED light up when endstop is plugged in and not triggered and itโ€™s light off when endstop triggered but M119 and DWC everytime stay in endstop hit.

    When I check R58 I noticed that the inductance L3 looks suspicious, but I have no idea how it can affect the Z end-stop only.


  • administrators

    Please connect the two outer pins of the Z endstop connector on the Duet together and check the reading in M119 or DWC machine properties. The LED should light and the endstop status should be "not triggered". Also send M574 without parameters and check that the Z endstop setting reads the same as X and Y (i.e. active high input).

    If the M574 parameters read the same for all endstops but the status reads as "triggered" with the outer pins shorted together, then please ask your supplier to replace your Duet under warranty.



  • @dc42 said in Endstop on one of the axis all time on hit state:

    Please connect the two outer pins of the Z endstop connector on the Duet together and check the reading in M119 or DWC machine properties. The LED should light and the endstop status should be "not triggered". Also send M574 without parameters and check that the Z endstop setting reads the same as X and Y (i.e. active high input).

    Already try to do that. No good news.

    please ask your supplier to replace your Duet under warranty.

    It will be difficult. The duet2 plate was a gift. ๐Ÿ˜‰

    Will try to replace L3 and check again.


  • administrators

    L3 has nothing to do with the endstop circuit.

    I guess it's quite an old Duet because R58 is 10K in revision 1.03 or later. The other thing you could check is that pin 108 of the ATSAM4E chip looks properly soldered.



  • Yes, it is version 1.02 and judging by the scheme with github the value of this resistor should be 1kOm. I checked for adjacent on the X and Y axes they are the same. Printer is out of reach, try to ping the line from 108 feet to C42 tomorrow.

    Thanks in advance, David!



  • @dc42
    The problem was poorly soldered 108 pin. Soldered it yourself and checked under a microscope. Hope this was the first and the last problem ;)))

    David, thank you so much for your help!


Log in to reply