I can see the wire break in the cable chain :'(

  • @SzethFriendOfNimi
    link
    English
    16
    edit-2
    1 month ago

    I think I need more coffee. Does the chart show where the thermistor failed?

    I’m kind of worried that it’s not obvious and if it happened to me I’d miss it.

    Or is it that it doesn’t show? And, if so, that’s worrying since thermal runaway protection may not trigger properly.

    • Odious
      link
      English
      151 month ago

      +1. No idea what’s supposed to lead to this conclusion. I’d say that’s a bad PID tune.

      • @IMALlamaOP
        link
        English
        121 month ago

        The drops especially are way too steep for this to be a PID tune. There’s too much thermal mass to loose that kind of temperature basically instantly.

        Lots of wandering around? PID tune seems like a good starting point, especially if something on the printer changed. Defying the laws of physics? Something else is going on.

      • @ikidd
        link
        English
        41 month ago

        If you got used to looking at your graph and saw those dropouts, you start looking for a physical problem.

      • @asbestos
        link
        English
        41 month ago

        The short timeframe and variations that are way to high even for a mediocrely tuned PID

      • @SzethFriendOfNimi
        link
        English
        31 month ago

        That’s worrying then. He can see the wires broken but it still has enough connectivity to keep chugging.

        On the plus side if the thermistor broke enough that it would keep pumping heat into the system then thermal runaway protection should trigger at least.

        • @IMALlamaOP
          link
          English
          21 month ago

          This is basically spot on. It’s not full open, but it’s close. This is the first print I’ve noticed it on. It wasn’t a long print, so I figured I would let it go and see what happened. If Klipper loses it for too long it will go into thermal shutdown. Ask me how I know :(

    • @IMALlamaOP
      link
      English
      61 month ago

      There’s no way the hot end’s temperature is changing that much and that rapidly in such a small time period. Rapid 1-2 °C fluctuations? Sure. This is significantly bigger than that.

      • @SzethFriendOfNimi
        link
        English
        21 month ago

        Ahh. I see that now. Not cycling like it normally would.

  • @[email protected]
    link
    fedilink
    English
    41 month ago

    Too familiar with that plot, at least you diagnosed it quick and it was apparent, sometimes they show up super intermittent and are a pain to chase.

    I think I recall you having a break a while back ago on [email protected], same spot?

    • @IMALlamaOP
      link
      English
      41 month ago

      That was me, but in that case it was the heater wire and the break was in the y-chain. This break is in the x-chain. Based on a reply on the Voron design forum, I may not have left enough slack in my wire runs. The breaks always happen on the inside radius, right where a wire touches one of the links in the chain, so this seems like a decent theory.

      I think I’m going to completely rewire the hot end. On the fence about going to USB or CAN vs discreet wiring. Since I tend to print big things, I don’t think an umbilical is a great idea - it seems like it would not work that well in high z situations. My reverse bowden already snags things sometimes.

      • @[email protected]
        link
        fedilink
        English
        21 month ago

        I left probably too large loops in those places, but I’ve seen some nasty cable breaks when I worked with pipeline inspection tools so I’m kinda paranoid about having enough slack.

        Umbilical folds on itself pretty well, I probably have mine too high tbh, but I could see a tophat mod being a thing. What you gain in less wires with can, you do add complexity to your overall setup, ran into some timeout issues after upgrading my SBC which seems to be related to this crowsnest issue thread that I’ve got sorted. Usb toolheads are a good idea, I went can because I’m using that for an ercf anyhow and already had bunch of usb devices.

        • @IMALlamaOP
          link
          English
          21 month ago

          I had no idea about cable chain slack when building :( I wouldn’t call the wires tight, but they’re not super duper loose either.

          I work in automotive and have a bit too much experience with CAN. It is appealing to have fewer wires in the chains, but so far wire breaks have been super easy to find/fix. I’m somewhat tempted to give chains with discreet wires another go. It shouldn’t take more than 4-5 hours to have a cut and crimped to lenth harness installed.

          I am tempted to start researching a BOM for USB + top hat though.

          • @[email protected]
            link
            fedilink
            English
            21 month ago

            Last time I used can was for battery controllers on one tool family back in pipelining, people kept losing or forgetting the terminators.

            There’s the nighthawk toolhead for Stealthburners that’s one of those usb umbilical boards, and did look at this printable tophat in the past.

            I definitely ran spares in my cable chain as the less I need to open them the better, if I did discreet wiring on a new build though I very much might consider running something like a multi conductor cable through the cable chains with a smaller one for power+ground, build in some allowances for spares just to be safe.

            • @IMALlamaOP
              link
              English
              21 month ago

              The nighthawk looks interesting. I’m somewhat tempted to give it a go, but I’m not sure about ditching the chains for a top hat yet. All the multi conductor cables I looked at that are USB rated are not rated for the tight bend radii found in the Voron chains :(