Fried relay controllers on the Digital Sidecar

Technical Electrical
#1

So today, our season was brought to an unfortunate end by the poorly timed failure of the relay section of our digital sidecar in our division semi-finals. We tried replacing PWM cables, trying different ports for DIO/Relay, shorting across the pressure gauge switch, the only thing we didn’t try was replacing the spike).

In talking to the kids and a mentor from another team, it sounds like this is not an uncommon occurrence.
I’m curious if others have had such issues?
What could be possible causes? My guess is metal shavings, we had just done an emergency repair the previous match, but our DSC is also fairly tucked away, so shavings feels a bit like a cop-out.

But most importantly, is there anything that can be done to protect it, like potting material or something similar (shavings protection) or is it just sensitive to stress, and bound to fail no matter what you do?

#2

We had one die during regionals, we’ve yet to get back in the shop so I can tear it apart and see, but I’ll let you know when I do!

#3

We had a DSC die in competition from an unintentional short of the blinking light wires. When we took it apart, it looked pristine, with no signs of magic smoke release.

We used CAN, so it did not stop the robot, but we couldn’t compete until we swapped the side car and had a blinking light, again.

#4

My team had the same problem this year too. At a regional a sidecar’s RSL and relay ports broke however the rest remained functional. On the inside it looked good but something must have happened. Luckily, the we noticed the problem in between matches and were able to fix it.

#5

We had one on Wednesday at Championships. Mechanical had tipped the robot over to fix a wheel problem, and when they set it up, the compressor wouldn’t start anymore.

We troubleshot everything, praying it wasn’t the DSC, because it was rather awkwardly placed, we had an hour before we needed to queue for our first real match, and we had a lot of PWM and sensor connections. But with 40 minutes to go, we realized that was it. (The compressor was the only thing on a relay port, but since both climbing and shooting required air, it wasn’t exactly optional.)

We had to disassemble part of our new blocker to be able to unfold the robot and reach the DSC, and then the wire moving began. They finished moving all the wiring, confirming the compressor now worked, and checking the shooter and wheels for function, and got everything buttoned back up just as we needed to cart up and go. Then we realized the signal light wasn’t, so that got fixed hastily. They picked up the robot, and as I was about to run the cart under it, another mentor noticed a dangling wire, we froze, he tied it back in, we put the cart under, and the drive team headed for queuing only two minutes behind schedule.

Once they’d left, we took the DSC and tapped it upside down on our podium and took a sad look at the pile of shavings that came out, probably having arrived during one of our emergency mechanical repairs in Arkansas. I don’t quite know what we’re going to do to protect it next year, but we’re doing something. :frowning:

#6

We’ve had our electronics mounted vertically for the past 2 years, and the only sidecar we’ve blown was wiring the power backwards during build season in 2012.
[/LIST]

#7

Very awesome tip. Thank you, I will surely be considering this for next season.

Any thoughts out there on conformal coating or anything of the sort?

#8

Just to clarify, there are no relays (spikes) in the DSC. Are you referring to the digital I/O section of the DSC?

#9

I think by “Relay section” of the DSC they are referring to the section of the DSC labeled “Relay”:rolleyes:



DSC-Relays.jpg


DSC-Relays.jpg

#10

An astute observation :wink: it’s the relay control portion, where all relays go to get their instructions.

#11

There is some conformal coating but not what I have seen in the past. It is possible to open a power trace on the board that takes out part of the outputs but not everything. Can you tell me, did you look at the LEDs on the relay section of the board? They tell right away if an output is being commanded for relays.

#12

I did not see any active LEDs, but I wasn’t actively looking for them.

We did determine that the pressure switch hadn’t failed, and replaced the cables connecting both the pressure switch and the relay.

We also tried moving the relay and digital output to different ports.

#13

We were actively looking and the relay section was dead. No lights. All other sections including the power indicators to the board looked normal.

After the replacement, another mentor popped the cover off the bad DSC and didn’t see any obvious signs of magic smoke release (burn marks/odor), but on very close inspection reported the relay section traces were feathered with AL fragments.

#14

SWARF is evil!

#15

Yes, yes it is. Perhaps we could have shop vacs in the KOP in the future?

#16

We lost a DSC this season (we think) due to swarf… things stopped working on it in a match up in Duluth, and when we got back to the pit, the 5V LED wasn’t lit anymore. We may have been able to resurrect it by blowing it out… but rather than take the risk of not getting everything, or having it fail again, or spending a long time blowing it out with no results, we went in and replaced it with a backup. Testing to see if we can get it working again is still on our “todo” list!

#17

The big flashy light is driven off of the same chain of shift registers that the relays are. It makes sense that both of them could fail at the same time.