Heat bed failure

Discussion in 'Calibration, Help, and Troubleshooting' started by K3LAG, Mar 13, 2016.

  1. K3LAG

    K3LAG Active Member

    Joined:
    Oct 9, 2015
    Messages:
    43
    Likes Received:
    21
    I finished building my printer yesterday and got everything dialed in. I printed a couple of benchy's with no problem then started a longer print before I went to bed. About 1/3 of the way through the print, the print stopped with a Thermal Runaway error on the heat bed.

    The heat bed will no longer heat up. I check all connections and they are fine. The resistance across my bed is only 1.5 ohms, so I think the bed killed the heat bed output on the Rumba. The reason I think the Rumba heat output is dead is because the heat bed LED no longer turns on when heat is called for.

    Is there anything else I should check, or is a new heat bed and Rumba going to be the answer like it was for 69charger?

    I've already emailed support so I'll report here when I have a resolution if it comes from there first.

    The print quality looks great. Hopefully I can keep printing some PLA while this get sorted out.

    Larry
     
  2. 69charger

    69charger Active Member

    Joined:
    Sep 5, 2015
    Messages:
    29
    Likes Received:
    4
    They did get me a 3rd Rumba, new bed, and wiring harness. I chose to still limit max bed power to 60% just to avoid another problem. Bed still heats fine with no risk of killing another board.
     
  3. K3LAG

    K3LAG Active Member

    Joined:
    Oct 9, 2015
    Messages:
    43
    Likes Received:
    21
    I have a new bed and Rumba on the way! I've got to say that while these minor problems are unfortunate, the customer support has been fantastic. The parts probably aren't all that expensive in the big picture, but the international shipping has got to add up.

    I'm so glad I can swap the Rumba without taking the bottom off the printer. I had to take it off twice to get my RPI working and I don't want to do it again any time soon. ;)

    It would probably be worth adding a step or two to the initial power up test to have folks test that the Pi and camera are working before they button up the bottom.

    In the long run, a small access panel for the Pi would be nice in BigBox v2.

    Larry
     
  4. 69charger

    69charger Active Member

    Joined:
    Sep 5, 2015
    Messages:
    29
    Likes Received:
    4
    Support has been absolutely stellar! Kudos to E3D!
     
  5. Dr Jeep

    Dr Jeep Well-Known Member

    Joined:
    Sep 12, 2015
    Messages:
    530
    Likes Received:
    169
    I don't really get how heated beds can be going low resistance once in service. It'd take a short across two or more parallel tracks to shorten the path enough to make that much difference.

    More likely I think is that there are a few with low resistance (due to track placement or sizing errors during fabrication) are making it past QA, the variable track width is a relatively unusual request so I can see how once it went to a board fab it might be getting missed in QA. I'm not about to suggest any QA step which might further delay my kit being dispatched but it might be worth adding steps to the build manual to check the heater resistances before wiring up (with the correct room temp values) least that way it saves frying Rumba's.

    Also it seems there is an issue with the sizing of the polyfuse. They are relatively hard to get right as there is a big gap between their nominal rating and their break rating, they are more useful for protecting against wiring fires etc than protecting the electronics, but at the point that we are blowing tracks off the board without the polyfuse tripping I'd say there is a sizing issue.
     
  6. K3LAG

    K3LAG Active Member

    Joined:
    Oct 9, 2015
    Messages:
    43
    Likes Received:
    21
    My failure mode was somewhat different than Charger69's. While my heat bed resistance is low, which I suspect is the cause of the dead Rumba, my Rumba failure is different. My polyquest fuse did not fail, and my board appears to be in good shape. However, the MOSFET used to control the heat bed is dead and it appears it may have taken that AVR pin out with it since there is no buffer of any kind between the AVR and the gate of the MOSFET. I suspect that the heat bed was faulty from the beginning and it just took some time to kill the Rumba. I did notice that the bed seemed to heat really fast, but it now has me wondering if only part of the bed was heating and the thermistor happened to be in a good part of the bed.

    I think the MOSFET may have overheated due to the fact that the polyquest fuse was kind of bent over above it likely blocking some of the air from the fan. Had the MOSFET not failed first, I may have ended up with the same failure Charger69 did.

    I too wondered about the idea to add a quick resistance check for the bed to the instructions before installing the bed to catch these situations before destroying more Rumbas.
     
  7. Dr Jeep

    Dr Jeep Well-Known Member

    Joined:
    Sep 12, 2015
    Messages:
    530
    Likes Received:
    169
    Least we know the psu is up to the job, your bed was trying to draw 16A and the supply didn't trip :)

    Unfortunately it's not unusual with controllers like the rumba to omit buffers or even better opto's in the name of cost or simplicity these days. If you're oldschool like me it feels wrong, but the fact is hardly anyone would bother to fix or even diagnose the fault to component level these days.

    So what does it matter if a failed fet takes out the micro as well...The board is going in the bin and being replaced in 90% of cases so adding protection on a £50 board for anything other than safety hazards and turning it into a £75 one doesn't make much sense......sign of the times :(
     
  8. K3LAG

    K3LAG Active Member

    Joined:
    Oct 9, 2015
    Messages:
    43
    Likes Received:
    21
    Everything is back up and running. The new heat bed is 2.3 ohms instead of 1.4. I had previously measured 1.5 but that was with it installed and included the wires.

    I may pull the FET off and see if that AVR pin is still alive. If it is, I'll replace the FET and keep the board as a spare.
     
  9. Old_Tafr

    Old_Tafr Well-Known Member

    Joined:
    Mar 7, 2016
    Messages:
    560
    Likes Received:
    75
    Power MOSFETs (the type used to heat the bed) are susceptible to static. Such devices don't fail immediately but over time when exposed to static so simply touching the terminals, heater wires when wiring the bed etc. can expose these components to static.

    It may be over cautious but simply grounding yourself on a radiator pipe or the screws on the back of your PC or even wearing an anti-static wrist strap suitable grounded will help, these precautions take seconds, but could save days or weeks of problems.
     

Share This Page