Fried Digital Side Car

[Brian Selle]

We've had a few Digital Side Cars burnout this season... wondering what may be causing them to fail (based on the picture below) and if they are repairable.

[CalTran]

Starting down at post 49 of that thread might help some diagnostic. Chip Q1 is a FET that protects reverse polarity powering of the sidecar. Check and make sure that your Wago connector is in fact wired properly.

[jee7s]

Did I read that schematic right? A 10 A breaker? Don't the FRC rules say to attach the DSC to use a 20 A breaker?

[cgmv123]

If you can find a replacement chip, any competent solderer should be able to fix the sidecar.

[cgmv123]

Originally Posted by jee7s

Did I read that schematic right? A 10 A breaker? Don't the FRC rules say to attach the DSC to use a 20 A breaker?

You read it right. I'm not sure why FIRST differs. That's a question for EricVanWyk.

[jee7s]

Well. the datasheet for the FET says it's good to 33A. But, that doesn't mean there isn't a lot of heat building up anyway.

Also, is that schematic for Rev 7 or Rev 8 of the DSC? Q1 has decidedly different packages on those two revs.

[EricVanWyk]

That schematic image is from Rev 7 or earlier. I think CB1 was removed before we went to actual production.

The updated FET part number is FDS8817NZ. There is also an added 10k gate resistor.

That part can blow in a few situations:
1) The DSC ground was used to return something else's current. This can happen if a wire shorts to a ground pin on the DSC. This is one of several reasons we do chassis isolation tests during inspection.

2) An ESD event hit the gate of the FET, and subsequently a large but nominally acceptable current was passed through the part (which is now probably a bad diode).

3) An ESD event hit, and then the DSC was plugged in with reversed polarity.

4) Aliens.

5) A short across the power inputs just down stream of this protection. This is rare: The switchers have very well behaved protection against short circuits at their loads, and the Big Flashy Light port has a breaker.

You can swap the part out with a decently hot iron and try again, but I would check to find where the bonus current is coming from. I'd turn it on attached to a current limited bench supply the first time after the repair.

Originally Posted by jee7s

Did I read that schematic right? A 10 A breaker? Don't the FRC rules say to attach the DSC to use a 20 A breaker?

I'd be happier with a 10 Amp breaker, but 20 is enough like 10 to be just fine. In the end, it wasn't worth the supply chain changes. There are vanishingly few cases where the difference between 10A and 20A breakers would matter to a team - At this point the first line defenses are already down and the board would need to be repaired or replaced.

[Brian Selle]

Thanks for the info. I'm thinking the issue may have been a PWM ground that momentarily touched a +12v spike input. Would that do it?

[EricVanWyk]

Originally Posted by btslaser

Thanks for the info. I'm thinking the issue may have been a PWM ground that momentarily touched a +12v spike input. Would that do it?

Definitely.

[RyanShoff]

If you just want to use it on a test bot, cut the FET off. Solder a jumper across the source and the drain. Don't connect the one pin that is the gate. Don't reverse the polarity or you will fry a previously fried sidecar.

[Al Skierkiewicz]

With the latest run of boards, the conformal coatings are not consistent. The damage looks like a piece of conductive material made it's way onto the board. With the kind of damage I see, replacement is the better choice. If it was one of kind, repair would be your only option but it may not fix the problems.

[amesmich]

We had the exact same thing happen to ours last week. I just bought the replacement. I was not around when it happened but I took it apart and it is identical to yours.

[Brian Selle]

Originally Posted by amesmich

We had the exact same thing happen to ours last week. I just bought the replacement. I was not around when it happened but I took it apart and it is identical to yours.

Our new protocol is to power down before touching ANY wiring. We always did this with the 12V wiring but got lazy hot-swapping PWM cables... not anymore.

[Jacob Bendicksen]

I have no idea what caused that, but a flipped PWM in the wrong (or right, depending on how you look at it) could probably do that. How'd it smell?

[Quix]

We have had his problem before - fried 3 DSCs and the same chip was fried.

It turns out that one of the victors was shorting on a a screw to the bellypan and this caused the chip to blow. I would check for any shorts, that is most likely your problem.