Today was going to be our first day for driving practice on the robot (which we have somehow managed to finish before ship date). Previous to today, the robot was complete enough to drive, operate pneumatics, etc. The electronics functioned correctly.
Well, we take the robot to an open area and get it there by pushing it. I turn the robot on and push the joysticks forward. It drives about 5 feet and then we hear POP! POP! and see 2 bright flashes. I hit the E-stop button but it doesn’t kill the right-side track. (Good thing we mounted the main breaker in an accessible place).
Long story short, one of the victors got fried and always supplied power to the drive motor it was connected to even when it had no signal from the RC. I unplugged the power to the RC and turned the robot on, the right side track still spun. ALL of the victors had the flashing yellow LED indicating that they had no signal.
After analysis, we find that 3 of the black things with the 3 leads coming out of them attached to the victor (are they called mosfets?) blew holes in themselves.
Has anyone heard of a victor failing this way? I was under the impression that the victors would supply no electricity if they were damaged and not working. Also, is it possible that pushing the robot (and thus supplying electricity back to the victors) caused this one to fail? We’ve never had problems with this in the past.
EDIT: We did double-check that everything was wired correctly both before and after this happened.
Wow! Although we let the magic smoke out of a Victor in our first year (late night, hooked the input leads up backwards) we’ve used… well… a couple of dozen of them by now and never had anything like what you have described happen.
I’ve heard of them blowing when metal shavings fall into them from above during repairs and aren’t noticed before power-up (hint: cover your Victors when working from above) but have not heard of the “locked on” state that you describe.
I guess that given that something like 6,000 victors were sent to teams all over the world this year that it is possible that you got a bad one, but I wouldn’t be surprised if IFI wanted to take a look at the dead one to figure out how it ended up in such a dangerous failure mode.
I hope you are able to replace it and carry on with your testing without too much lost progress.
If you are interested in the exact cause of the MOSFET failure, take a look at this Web page: http://www.4qd.co.uk/serv/MOSFETfail.html
Especially look under the section entitled dV/dt failure–it probably describes your situation perfectly.
Also, while most MOSFETs fail open, occasionally one can fail shorted, which would apply continuous power to your motor. Because the motor was not drawing enough current to either burn out the shorted MOSFET or trip the circuit breaker, the motor continued to operate.
We had a speed controller blow like that last fall. We ran a demo over the summer at a local fairgrounds where we ran on the grass. After a full weekend of driving the grass had worn away & we were running on dirt. Needless to say everything on the robot was covered in dirt. We tried to blow everything out with some compressed air. I guess we didn’t clean out the speed controllers well enough. We then were showing the robot at our school open house in the fall & POP & spin happened. We hit the disable switch but the same thing happened, it just kept going. After we took the speed controller apart we discovered that it was still filled with dirt & had over heated.
I had a Victor fail that way on our 2005 robot the day it went into the crate.
I had just opened the robotics room and had just powered up the system to check on work the kids had finished late the night before (okay, a couple of hours earlier).
It controlled a telescoping arm, so POP and the arm began extending on it’s own.
Hit the Kill switch…it kept going.
Pulled the tether…it kept going.
Pulled the battery connection just in case the main breaker decided to fail on me too.