β€Ί β€Ί

Convert (adapter) PCB on A5S

Recently I had trouble with mainboard on A5S and now I am looking for what caused the damage.
X-axis endstop slot on original board does not react any more on signals, it is locked in NC position. Rest of hardware on printer is ok.
After playing around with Lerdge X and few small succesufull prints, same problems started on Lerdge like with original board before it had damage.
After deep inspection of everything, only possible and logical cause of problems can be Convert PCB.
As picture shows, x and z axis endswitch have common ground and that is in certain scenarios cause for printers strange behavior.
And on PCB is written A3s Convert board, I am not sure is it design /hw borrowed from previus A5 or it is mistake in factory assembling?

Your opinion guys?



Comments

  • Stephen ToddStephen Todd Posts: 281🌟 Super Member 🌟
    Common ground is not a problem - laptops are far more sensitive and they have dozens of components tied to a common ground

    As I said previously, the standard procedure is to re-flow with 60/40 solder - you can do it quickly and easily with a soldering iron and liquid flux - then clean up the PCB with isopropyl alchohol and cotton buds

    Lead free solder is the worst thing ever invented for electronics assembly
  • DukeLanderDukeLander Posts: 23Member
    edited March 17
    Stephen, I know what is common ground and how it works. All contacts on board works as intended, checked more times.
    But think for a second: when x endstop switch let the signal (+) through common ground to the mainboard, automatically z slot on mainboard gets signal too.
    So i.e. when x is homed, mainboard "thinks" z is homed too.
    They are divided sections on main board with divided + and -  lines, common ground between endstops and mainboard does not fit at all.
    Here is video what is goin on, everything is connected as original :

    https://drive.google.com/file/d/1KUheFth8fcn1HMHDzjediqGRHiyf8PIO/view?usp=drivesdk
    Post edited by DukeLander on
  • Stephen ToddStephen Todd Posts: 281🌟 Super Member 🌟
    edited March 18
    Sorry, you are barking up the wrong tree - the problem is not shared ground - but it could be that your printer has a floating ground - have you checked that you have a direct connection to the wall socket ground from the metal controller case - I have seen problems when people use extension cables/blocks that break the earth path between the printer and wall socket

    If you are convinced the problem is shared ground - simply cut the track with a utility knife and make two direct wire connections to the metalwork of the controller case from the PCB tracks


    Post edited by Stephen Todd on
  • Stephen ToddStephen Todd Posts: 281🌟 Super Member 🌟
    Common ground and 2 open switches - closing one switch does not effect the other switch + side

    Anyway, we are only changing logic 1 to logic 0 on the controller sensing pins

    Also, how many of the controller board sockets are common ground?


  • Stephen ToddStephen Todd Posts: 281🌟 Super Member 🌟
    Question is, how is the sensor pin getting pulled to ground when the switch is not being closed?
  • DukeLanderDukeLander Posts: 23Member
    edited March 18
    I dont understand barking thing, but anyway I am not here to argue.
    Your scheme is correct.
     Can you please explain me in this case when we have common ground, why do we need 2 negative wires than?

    Interesting, I did not found earlier this thread:


    So there are more same cases like mine with same sympthoms and errors.

    Post edited by DukeLander on
  • Stephen ToddStephen Todd Posts: 281🌟 Super Member 🌟
    edited March 18
    Normally, 2 sets of wires is to use thinner wires to pass the current, i.e. share the load - but I wouldn't expect much current to be going through those wires, so likely to be simply good practice/clarity

    Have you measured a voltage drop on the + side of the open switch with a voltmeter when you close the other switch manually?

    Edit: If both those + voltages originate from the same source, and has insufficient current, one switch closing to ground could pull the + voltage for the other down as well, which could then be below the logic 0 threshold, and flip the logic state of the input pin
    Post edited by Stephen Todd on
  • DukeLanderDukeLander Posts: 23Member
    edited March 19
    I gave the PCB to my colleagues in company, electronics department ( because we have a lot of automatization and robotics ), they have checked and everything is ok. But colleague also said he does not understand why they did common ground like thit. He got this limited conclusion based only on PCB, I am gonna take the mainboard tomorrow to them to take a look what is really goin on.

    So in this thread,mate with same problem with board have posted pictures with multimeter testing.
    There are pictures of endstop switch slots on board. All black (negative) wires on board and from switches are connected to common ground on PCB.

    And PCB detailed scheme looks like on picture below, where + wires are directly connected from mainboard to switches, all - goes trough common ground.


    P.S.: I have forgot to mention earlier: first thing what have I tried was direct wires to switches, it works without problem.


    Thanked by 1Samuel Pinches
    Post edited by DukeLander on
  • Stephen ToddStephen Todd Posts: 281🌟 Super Member 🌟
    Really strange that printer works OK when switches are directly wired

    The electronics department will be able to test for "noise" and "pulses" which could be making the board unstable
Sign In or Register to comment.