Swarf - The Sparkly Killer

I received a PD recently for failure analysis. I’ve attached a word document that contains 5 pictures of what I saw on first opening the case.

I’d like to take this opportunity to remind teams to protect (or remove!!) their electronics when working on the robot. This one “looks like it was placed under a drill press to catch the shavings”, according to one of my co-workers. The conformal coating of the boards greatly reduces the likelihood that they will be damaged by swarf. However, you really shouldn’t rely on it - belt and suspenders, as they say.

Surprisingly enough, this PD works just fine after being cleaned. The boards are designed to fail safe, so it likely protected itself. On the other hand, I am glad I wasn’t on the team that saw their robot grind to a halt mid-match.

PD Swarf.doc (666 KB)

PD Swarf.doc (666 KB)

At the end of the season, our team had a WGA that we thought was intermittent. It had dropped out a few times halfway through the GTR, and for the remainder of the season, we used loaners at the events.

A few weeks ago, we opened it up, and the PCB was absolutely covered in filth.

We generally take great care to cover up all our electronics whenever we remove material on the robot - regardless of whether it is near the electronics board or not. We always blow everything out with compressed air afterwards as well.

However, I thought this year the dirt/swarf problem seemed a lot worse. It seemed like the static buildup from running on these wheels/regolith/carpet turned our electronics into dust and dirt magnets.

Has anyone else bothered to crack open their WGAs? Was yours incredibly dirty, despite thinking your electronics housekeeping was pretty good?

After vacuuming, brushing, and blowing it clean, the WGA works great again!

You could just attach the .jpg images to your post, so we don’t have to use Word to see them :slight_smile:

Looks like normal dirt to me…I was expecting to see something like the floor of my shop under the drill press!



Sorry Jim, I got in a fight with my work computer and decided this was the easiest way to compress the images.

If you look in between C8 and C9 in the image you reposted, you can see some rather evil looking swarf.

I’m not sure where “between C8 and C9” really is; the one tiny dot I can make out there is dwarfed by junk elsewhere on the board. Without any depth to the image, it’s not easy for me to tell what’s debris and what’s just blemishes in the conformal coating. It probably looks a lot worse in person.

Alan, I think the biggest culprit are the two metal “hairs” that start in the gap between C8 and C9 (but closer, or even touching C9), run down and across the R38a label, and almost underlining the C8 label.

To me, of all the debris on the board, those particular pieces scream loudest in their readiness to release magical smoke.

Are we looking at the same photo? I don’t see anything remarkable between the C8 and C9, and I don’t see a R38a anywhere (though R38 is near). However, your description of “underlining the C8 label” does a good job of pointing to what I had originally interpreted as a wrinkle in the conformal coat – is that truly metal shavings? Yoicks.

My not-so-old eyes do deceive me…It’s R38, and the dirty via beside it looked like an “a”

Of course that would’ve been the first time ever I’ve seen a layout labelled with trailing letters, but I digress =).

We covered our control board with an old team T shirt a couple of times, but we might want to figure something else out- the shavings stick to to fabric.

Stephen Colbert swarf.jpg

Stephen Colbert swarf.jpg