Could StallGuard be used to detect _lack_ of load?



  • Could StallGuard of the TMC2660 on DuetWifi somehow be used to detect a LACK of load? Reason for asking is a recent experience with an almost broken drive belt on a Delta machine that normally prints quite fast.

    Bu i guess the answer is "No, StallGuard can only detect excessive load"?



  • What it does is measure what trinamic refers to as the load angle; so "yes" its possible to determine-(ish) the amount of load.

    But if its possible with current RRF I couldn't tell you; but the same caveats applies to this as it does with sensorless homing. Its not 100% reliable, especially at high and low speeds, so ultimately not sure if its useful in your application.



  • @bearer said in Could StallGuard be used to detect _lack_ of load?:

    What it does is measure what trinamic refers to as the load angle; so "yes" its possible to determine-(ish) the amount of load.

    But if its possible with current RRF I couldn't tell you; but the same caveats applies to this as it does with sensorless homing. Its not 100% reliable, especially at high and low speeds, so ultimately not sure if its useful in your application.

    That pretty much sums up what i thought as well.. And i realize that apart from a sensing idler or photo-gate, I cant see a good way to detect a broken drive belt.

    Ok, apart from a multiaxis accelerometer on the effector that compares the measured motion to the expected motion. Hmmm, could Duet3 and its CAN bus...



  • Without knowing much of your mounting, a strain gauge could be a solution if you can mount it where it matters. HX711 modules to interface strain gauges are cheap as chips.



  • @janjoh perhaps one of those wheeled micro switches against the tensioned belt. Would release once the belt broke and lost tension and can easily be wired into a trigger. Quite simple if it could be applied.
    4180574d-1e83-438c-8886-275ebdd30935-image.png



  • I would think this would be the perfect application for an optical endstop, assuming its slot is wide enough.



  • @Danal said in Could StallGuard be used to detect _lack_ of load?:

    I would think this would be the perfect application for an optical endstop, assuming its slot is wide enough.

    I see one potential problem. You seem to be assuming that the belt would fall out of the slot upon failure. I can see many scenarios where that would not happen. Or, do you see a way to actually detect the teeth passing through an optical sensor?

    I think the least impractical ways would be a a microswitch as previously mentioned, or to detect rotation of the idlers.



  • @littlehobbyshop said in Could StallGuard be used to detect _lack_ of load?:

    @janjoh perhaps one of those wheeled micro switches against the tensioned belt. Would release once the belt broke and lost tension and can easily be wired into a trigger. Quite simple if it could be applied.
    4180574d-1e83-438c-8886-275ebdd30935-image.png

    Yeah, that is probably the most simple solution, especially since it would be so easy to wire three NC in serial.

    What bugs me about my failure is that I didnt get any advance warning in form of fraying of belt that i had seen.. and lets face it, we all watch most of our first layers, right?



  • Good point, the belt may not fall out of an optical. But that raises the same q for a microswitch: Where is the microswitch positioned that the belt is guaranteed to go slack?


  • administrators

    @janjoh said in Could StallGuard be used to detect _lack_ of load?:

    Could StallGuard of the TMC2660 on DuetWifi somehow be used to detect a LACK of load? Reason for asking is a recent experience with an almost broken drive belt on a Delta machine that normally prints quite fast.

    Bu i guess the answer is "No, StallGuard can only detect excessive load"?

    To answer the original question, in theory StallGuard could be used to detect lack of load. But it would only be possible while executing a move for which the load is significant - for example, accelerating a heavy print head, such that the load due to the inertia of the print head is greater than the load due to rotor inertial if the motor.


Log in to reply