I had an 18v7 fail on me today, not 100% sure why, but I suspect its a due to the motor being disconnected while driving. I had wired up a 3-position reversing switch between the output and the motor, with a central OFF position. I thought I’d been careful to always stop the motor first, but maybe not. Or it could have been the loose wire I later found on the screw terminal.
Smoke came out of the board in the area of the FETs - I scrambled to kill the power but too late (obviously).
I was running on a 24V regulated supply, and the motor was loaded on start when it failed, so maybe it’s possible inrush/stall current >30A killed the 18v7 - I’m struggling to get complete specs for the motor.
Am I correct in assuming disconnecting the load during operation can kill the board ? Or can overcurrent also cause failure.
Would it be possible for me to repair the board by replacing the FETs ?
The board still talks happily to the ‘simple motor control centre’ software, but I just hear various high-pitched ringing noises when I try to drive (interestingly the noise is different between forward; & reverse)
I think i had the same problem with a jrk 12v3.
It seems that they are not protect again over-courrent and the 24v of your power supply are too much for them.
Try to see the power supply specifications of your motor.
To know what part of your card goes wrong try to get a thermal photograph of your card when connected to power supply to see if any component become hot.
In my case the motor driver chip get hot and that means it is damaged.
You can see my problem on post “Problems using Pololu jrk 21v3 USB motor controller and Polo”
In general, disconnecting things while they are powered is not good. I do not think disconnecting a motor while it is powered would damage the Simple Motor Controller, but I cannot say this for sure. I think an inadvertent short from a loose wire or excessive inrush current are much more likely to be responsible for the damage (the SMC does not have overcurrent protection).
If the MOSFETs are damaged, replacing them might fix the board. Do you have appropriate equipment and experience to perform surface-mount rework?
The motor driver on the jrk is different from the motor driver on the Simple Motor Controller. The jrk’s motor driver does have built-in overcurrent protection, though it is still possible for large current spikes to damage the driver before the protection can engage.
no evidence of that; just one slightly loose screw terminal with the wire still firmly in place, could explain interrupted load but not a short. I used crimp terminals on all wires.
I didn’t realise there was no overcurrent protection That now seems the most likely cause - I’ll need to buy a different board as a replacement, but am still keen to try to repair the failed one.
I have solder paste, tweezers, a hot air gun and nothing to lose.
What MOSFETs are suitable ?
Could I potentially put ‘bigger’ (higher current rating) ones ?
You might consider the Simple Motor Controller 18v15 as a replacement, as it can handle more current. If you contact us at support@pololu.com with your order information and reference this forum thread, I might be able to help get a you discount towards a replacement.
If you are going to try and replace the MOSFETs on your board, a suitable replacement would be the DMN3404L-7 MOSFETs made by Diodes inc. I also do not expect there to be many MOSFETs in that package that are capable of more power.
Thanks Grant - I’ll get some and try, at ~12c each I’ve nothing to lose, although I don’t understand the current rating of the FET seems to be 4.2 or 5.8A, c.f. the 18v7 rating of 7A.
I may revert to using the VNH5019 with this motor, then I can implement some current protection in software, and possibly use both channels if the current is too high.
The 18v7 Simple Motor Controller uses two MOSFETs in parallel per leg to get an overall higher current carrying capacity.
If you decide you want to keep using the Simpler Motor Controller, you might consider using a fuse to protect it from over-current. However, if you do decide to use the VNH5019 driver, you probably do not need to implement your own software over-current protection, as the VNH5019 driver already has relatively robust over-temperature and over-current protection built-in.
That now seems the most likely cause - I’ll need to buy a different board as a replacement, but am still keen to try to repair the failed one.