Two 18v17 G2 high power motor drivers have failed during testing, both when running a drill, unloaded apart from the gearbox, using a 3S LiPo. The first has failed on the high current side. The DIR pin on the second has failed (it seems to be shorted to ground). Although hobby vehicle ESCs take quite a bit more programming (to get around reverse offsets etc), I’m leaning towards using them as they seem more robust (I have 2 working now) - but I’d like to be convinced otherwise.
My project will (hopefully) use 12v drills to drive winches which control kite lines. Initially this will be RC but will be developed to an autopilot. The kite will tow a boat.
I had not, as recommended, added a capacitor to the power supply on the first driver - so I’ll own that. Having said that, as the system was unloaded, should one expect this mode of failure? I had done quite a bit of successful testing (eg system responding appropriately to RC input signals) with that setup before it failed.
Failure of the 2nd controller cost me some time - the appropriate signal was going to the DIR pin of the controller. Subsequent troubleshooting identified that the DIR PIN was low regardless of the input signal. Visual inspection indicates that the header pin and capacitor soldering is clean and good - and the board worked fine for some time (testing) before failing.
I have been using an Arduino Uno. SLP, PWM and DIR signals to the controller are from suitable GPIOs. The Uno and controller have a common earth. I’ve supplied the Uno either with USB from laptop or from the controller’s VM pin (the 3S LiPo has been kept below 12v when the VM pin has been used). Maximum duty cycle is close to 100% (eg analogWrite(PWM_PIN,250); Initial testing was done at lower speeds - but each board had run successfully at maximum speeds (but no load).
The power supply lines are about 250mm/4.5mm diameter from battery to XT60 connector then 100mm/2.5mm diameter to controller. Another 100mm/2.5mm diameter wire from controller to motor. (Diameters are inclusive of insulation - I’ll strip and measure copper diameters if that’s important.) I had a 220uF (50V) capacitor across the power supply (ie on the 2nd controller).
The initial maximum power requirement will be to move a 5kg load through 300mm in 1 second ~about 15W mechanical. Allowing even for 25% efficiency, this would be a 5A current - 3A is a more likely upper estimate (at 12v). There is a 10A fuse in the circuit. Usage is intermittent - 300mm (so about 1 second) would be the longest travel and periods between such inputs would be several seconds. But, as indicated above, no loads have as yet been applied to the drills.