EEVblog 1722 - Manufacturing Hardware War STORY

[EEVblog]

Re-telling of a manufacturing hardware war story talk I gave at a recent hardware meetup in Sydney.
https://luma.com/Sydney-hardware-meetup?k=c
How a simple manufacturing change cost a company millions of dollars.

00:00 - Sydney Hardware Meetup
01:16 - Seismic Oil Survey Exploration: How to find oil
03:29 - Seismic survey vessels are HUGE!
04:25 - Solid Towed Array Streamers
05:56 - On the back deck of the ships
06:33 - Leading edge low noise 24bit ADC's
07:23 - Hydrophone module construction
09:33 - The actual war story - Continuous Improvement...
14:12 - It CAN'T be that, it's impossible...
16:58 - Field repair wasn't easy...

[abeyer]

Took one look at the survey ship and was sure someone had found the front that fell off.

[cunningfellow]

"That's not very typical.  I'd like to make that point"

[Red Squirrel]

Reminds me of when I'm coding and wonder "why was it done this way?" and change something I then ask myself. "hope this doesn't cause an issue later..." 

[Jesse_G_EE]

Reminds me of this last month...Having high failure rate on an amplifier board and our internal SMT plant(we are a very large company) decided to allow the PCB fabricator to plate and fill all via's on the printed circuit board. Didn't consult us, or our customer(we are a CM), and we are experience 50% or more failure rate on an end of the line functional test(medical nebulizing device)...had them CT scanned which showed showed huge solder voids under the drain of a the driving transistor. I reflowed the dpak with a heat gun and problem went away. Had another EMS run my fabrication drawing with the and not a single one had VOID's or the issue that was occuring. I reached out and said hey remember when I asked why all the via's were filled and plated over and you said don't worry that's won't be an issue? Well it was a ****ing issue. 1-star on yelp to our internal "excellence"