Burnt Driver Part 2



  • After dealing with a loose wire burning my original Y-axis driver, I ordered a new one from Digi-Key and soldered it back on to the board. the soldering was fine, but as soon as I clicked "move" it fried. What could cause this? Both drivers are TMC2660-PA. The Original drivers are TMC2660-PA followed by "1612 07005" while my burnt Y-axis is a TMC2660-PA "–-6 06814" (The three missing digits are burnt) I couldn't find anything other than a TMC2660-PA driver, no variants leading me to believe it's some sort of date or batch code. But I also hope I bought a driver designed for 5 or 12 volts as opposed to 24V-30V like the standard Duet Wifi drivers.

    Thanks in advance!

    Burnt Driver Part 1: https://www.duet3d.com/forum/thread.php?id=392


  • administrators

    If your wiring and soldering were OK, the only other thing I can think of is ESD damage prior to soldering the device that caused gate insulation breakdown on one of the output mosfets. The datasheet includes this warning:

    The TMC2660 is a ESD-sensitive CMOS device and sensitive to electrostatic discharge. Take special care to use adequate grounding of personnel and machines in manual handling. After soldering the device to the board, ESD requirements are more relaxed. Failure to do so can result in defects or decreased reliability.
    Note: In a modern SMD manufacturing process, ESD voltages well below 100V are standard. A major source for ESD is hot-plugging the motor during operation. As the power MOSFETs are discrete devices, the device in fact is very rugged concerning any ESD event on the motor outputs. All other connections are typically protected due to external circuitry on the PCB.



  • Did you verify the motor is not shorted by measuring the resistance at the connector to the duetwifi and compare it with the other motors that are running? You should also rotate the shaft on the motor to make sure there is nothing mechanically damaged inside. It should rotate freely. If the resistance value is good, you should plug it into another axis and verify the motor works perfectly. If you didn't measure the resistance, you might blow the other controller also. At least this way, you know for sure the problem is on the PWB side and not your motor and wiring.


  • administrators

    I guess another possibility is that another component attached to the TMC driver is burned out, such as one of the current sense resistors. But I expect you would have noticed that.



  • I tested the board by

    1. powered on the board to test X and Z axis motors (Y axis is the driver in question)

    2. Verified X and Z functioned.

    3. Cut power to the Power Supply, waited until all lights were off.

    4. Hooked the X axis motor to the Y axis, due to the first driver being killed via shorting a wire I placed the wires in connectors to eliminate shorting. Also visually checked the wires were not going to short.

    5. Clicked "Move right 1 mm"

    6. Experienced the scent of burning silicon and dying a little inside.

    Just in case, how could I map the Y axis to an extruder driver?



  • @dc42:

    I guess another possibility is that another component attached to the TMC driver is burned out, such as one of the current sense resistors. But I expect you would have noticed that.

    Is it possible it's burnt but not visually noticeable? Would that cause the driver to self destruct?



  • MOS family semiconductors are be pretty sensitive and can certainly suffer rather catastrophic modes of failure when damaged. It could have been any number of things, ESD damage, some kind of latch-up event, heated beyond tolerance or for too long, etc.

    Assuming none of the supporting passive devices related to that driver are bad, it was likely damaged when fitted.


  • administrators

    To map the Y axis to Z different driver, see the M584 command in the gcode wiki at reprap.org.

    Are you quite certain that the soldering on the new driver was OK? Have you checked the current sense resistors for continuity?



  • Your test does not prove the Y motor is not at fault. It is obvious the Y controller chip is dead, so hooking up the X motor to to Y controller port makes no sense. You may still burn up your new chip if the Y motor (and wiring/connectors) is hooked up to it again if that is the cause. I had bad connectors and intermittent connections that ended up damaging the controller on a CNC machine. You should inspect the connectors and wiring to make sure that its seated properly and there are no loose connections also.

    You need to know if the Y motor is working properly by checking resistance and hooking that motor up to a controller that is working. If you verified the Y motor is working on the other controller (X or Z), you can focus attention on soldering, ESD or other potential causes for the second failure. If not, you may fry the control chip again since you never identified the root cause.



  • To clarify, I only used the X axis motor, I know it works and is wired up to a 4 pin connector. I hadn't yet wired the Y-axis to a connector so I used the X-axis motor. The motors are not the issue, this much I know.


Locked
 

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