OT: strike from a welder (thanks)

[newguy]

Sorry for the completely off topic post, but I'm so completely thrilled.

I designed and have hand built several automotive components for a customer. This customer, in turn, had a customer that reported a unit that suddenly stopped working. From the symptoms that my customer relayed to me, I suggested that a communications cable was either compromised or the terminals had become somehow fouled/dirty.

Turns out I was right. A terminal in the housing was bent. He bent it back, and the unit started to function normally. However, he noticed that classic burned odor and swapped the unit for a new one. He removed the unit's cover and discovered that the connector was visibly scorched and then sent the unit back to me.

I replaced the connector and tested the unit. It functions perfectly. The only thing that was scorched was the connector itself.

This connector feeds power to downstream devices and also accomodates two CAN communication lines. All lines, save for ground, are protected with TVS, series ferrite chokes, and the main power out line also has a resettable fuse. The line that was the most damaged was the ground line, which lacks any protection at all. Since the unit has internal fusing on the power feed, and that fusing is for a small current value, I deduced that some external power source shorted to ground through the data/power cable, which must have been left hanging at the other end. I thought that perhaps it contacted a live battery (+12V) source, and relayed this information to my customer.

Upon hearing this, the customer did one of those "Ohhhhhhh. Yeah, that explains it. The cord was hanging right next to where they were welding."

!!!!?????!!!!!!! Welding!?!!!???? Holy %$&#!!!!! I'm so ABSOLUTELY STOKED that an electronic device that I designed survived an indirect or direct strike from a welder. I don't know about the nature of the strike; all I know is that a welder caused it. The ground line definitely can't be the only one that got hit; since it's unprotected it's the only one that could support an appreciable current. However, I'm so happy that the other lines (which surely must have absorbed something) survived.

Now for the point of this post: I have to thank the assembled masses here for all the ESD and general layout tips that I've learned over the years. There are too many people in particular who deserve thanks to be able to name them all, so thank you everyone. If there is anyone here struggling with ESD or general protection issues, just search this forum. There is a lot of really useful, really effective advice to be found here.

[ELCouz]

Congrats !

For sure that case wasn't the average ESD problem that many of us had !

Best Regards,

Laurent

[Ttelmah]

Well done newguy.
Lovely example.
I remember one I had a while ago, where a customer started complaining that the error log was showing a lot of 'retries' on communications. This was over a RS485 link. The reason turned out to be that a contractor, had connected one wire of the RS485 communications, to one of the 'live' lines on the local 440V supply...
The impressive thing was that the comms still ran pretty well. What was happening though was that as motors started on the supply, individual packets would get damaged.
In my case 'praise' for using opto-coupling.

Best Wishes

[Storic]

I would agree in somewhat protection,

Do the best you can to give good cost effective protection for normal operation and sell a replacement unit because the customer had damage there unit due to incorrect installation/service

good effective protection means cost effective parts that the customer is willing to hold addition units just in case.

5 of $200 item sell better than 1 of $1000 item
_________________
What has been learnt if you make the same mistake?

[newguy]