This is what happens to a victor884 when someone neglects to mind their metal shavings. We were testing with our practice robot a few weeks ago and it started pouring smoke. Like the kind of pouring smoke that you see when you open the lid of a barbecue; it was an intense lunge for the circuit breaker.
Anyway, we took the thing apart to assess the damage, and it is gruesome. One of the regulators was burned off, and the circuit board is completely toasted.
This is basically a public service announcement. If your mentors tell you to put a towel down when you're drilling, for the love of god, take their advice.
(Note that we took the victor home to photograph it, we didn't break GM's camera policy)
This seems to be a very common problem among teams. IFI slathers conformal coating all over the board except the FET heatsinks. Maybie they leave out the sinks for thermal reasons. If this is the case, why don’t they use some anodized aluminum heatsinks.
The heatsinks would provide improved thermals and the anodization would provide some shaving protection.
The IFI people are smart. I would think that they would work toward a way to fix this issue. Sure its the team’s fault, but competitor’s speed controller designs do not have exposed FETs.
Well what they really should do is put some sort of rubber above the circuit board and below the heatsinks so little shavings can’t slip into the holes in the case. Maybe we’ll find a way to modify them.
Or (to take the electrical engineer’s side) we could all just refrain from drilling, sawing, grinding, filing, and other chip-making activities in proximity to closely spaced electrical conductors!
What a marvelous picture! Wonder what might have been in the hole? There is a FET missing from the board. Once the shaving(s) were layed down and power applied, a hot spot/arc developed, then the FET set itself on fire or might have exploded. Then the board succumbed to the high heat. I bet it smells really bad as well. Conformal coatings are not meant for transistor packages that exist above the circuit board or those components that may get hot. In this case, a little insulation around the FET leads (as they entered the board) might have helped. Now you know why electrical guys are always complaining about metal shavings and inspectors are always watching for wiring errors and circuit breaker use. I bet the breakers were tripping but the wires leading to the controller were not damaged.
I wouldn’t be surprised if when we get to Atlanta, Team 4 discovers at least one of our Victors has a toasted FET.
When we powered on our robot at the start of our second semi-final match at the GTR, a cloud of “magic smoke” appeared. Prior to going onto the field for match #2, we had used our timeout to repair battle damage. In the first match, the sheet aluminum “scoop” for our ball collector had snagged on the edge of the ramp, folded under the robot and ripped a portion of the lexan sheet on which our electronic components are mounted. The sheet aluminum was removed and, in order to re-attach the lexan board with zip-ties, some holes were drilled some sheet aluminum above the board. In the heat of battle, the electronics were not covered nor blown out with compressed air before going back onto the field. I suspect a bit of aluminum may have found its way into a Victor and caused a short - poof!
Our #4 alliance (229, 217 and 4) managed to beat the #1 alliance (1114, 1503, 865) in the first semi-final match. In the second and third matches, our drive system was clearly hobbled. Teams 1114 and 1503 were an awesome scoring pair and if you give them a few seconds, they can rack up a big score. We couldn’t harass either of the two triplets effectively, so they rattled the chains over heads with a deluge of balls.
I’m sure out EE’s work work toward a fix, if we felt there WAS an issue.
It’s really not a common problem.
We shipped about 4000 Victors in the kit this year. I wonder how many of those were burnt out by teams. 1% = 40? Less?
I think everyone will agree it’s a very robust product.
Now would you rather pay more for every victor, for protection to solve a non-issue, or would you rather lay down a towel?
I’d imagine that the number of Victors that release the magic smoke for one reason or another is in the lower hundreds. Every team has that moment of foolishness, that moment of “Eh, surely we can’t blow a Victor doing this…”
That said, the cheap option remains to find a towel/box/big piece of bubble wrap for the purpose. (I think it’d be kinda awesome to have an aftermarket cover that fits over the Victor and around the wires for these kinds of fabrication events. Then again, that may just be me.)
A heavy fabric shop coat works well. The shavings easily brush off upon removal for quick cleanup and reuse. Towels and rags leave gaps for shavings to get through. One continuous covering to protect the electronics is better.
Another possibility is to apply a conformal coating to the exposed leads on the FETs. Putting electrical tape over the power and motor connections is a no brainer, but most teams won’t want to open up the Victor to brush on an insulating coating on the FET leads.
Dave,
This is not a sure fix with all the vibration and bumping of the FET cases. The coating would eventually break out. It is far simpler to protect with a shop coat or turn the robot on it’s side and let gravity take over. If your mechanical guys knew the cost of replacement, there would be more care taken around an expensive component. Remember too, it is not always just one wayward shaving. Many of the Victor failures I have examined were full of aluminum “dust” and shavings. Enough to wipe your finger across the board and come back visibly silver. Drilling is not the only operation, deburring creates as much or more and using compressed air makes matters worse. Use a vacuum when possible. Compressed air drives the metal into open frame motors, bearings, gears and electronics. Any of which will cause some form of failure.
Slightly off-topic, but this reminds me of the machine shop supervisor I worked with in college. He did not allow any use of compressed air to blow chips (a practice I have since observed in many shops) on the grounds that you can’t control where a blown chip will land. This supervisor rigidly enforced a policy of vacuuming all chips between cutting operations, and stopping work 30 minutes before shop-closing time to thoroughly vacuum the areas around all the machines.
Of course, this guy worked for the School of Electrical Engineering (now ECE) at Georgia Tech, and many of his jobs were for research in electronics and optics, so maybe he was a little more sensitive about chips than your average mechanical person.
