View Single Post
Old 17th Apr 2021, 2:54 pm   #41
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: Which signal generator?

A lot of test equipment is simply weighed down by the sheer number of parts involved. Even if the parts were reliable examples of their type, the final calculation of mean time between failures would not come out well.

We have far better parts nowadays and that helps.

Back in the 80's I had a spell in the 'manufacturing engineering' department at HP. During that time I got the task of deciding what changes we needed to make to reduce the measured failure rate (we only got data for the warranty period) across all our products. The aim was a 'decade in a decade' the clever sounding words chosen to appeal to top management. Well we did it. We got much better than a tenfold reduction in failures in the field and we did it in well under ten years.

The most powerful tool was a database.

Each failure was broken down into the parts that got replaced in the repair. Repairers in the field weren't really up to the task of always investigating a root cause, they had the next box to fix, waiting. The database could be searched by part, PCB assembly, instrument or groups of instruments.

Each part had a preference code, used by the materials department to flag preferable parts and rogue parts. I could scan an instrument looking for low preference parts, for parts that just failed. I found patterns. Ostensibly good parts with a good fail rate across the whole corporation showed up bad in some instruments. The database gave component designators as well. I could dive into that assembly and go looking for under-rated components and plain design errors.

Several years later, those top managers decided the database IT costs were too high, so they shut it down.... oh, well.

So yes, test gear is unreliable. The reasons are understood and real action has been taken to improve it.

At the same time I also had the responsibility across the division's product line to do what changes were necessary to get it all through the new EMC regulations.

I stayed in manufacturing for about 7 years and learned a lot I'd never have come across if I'd stayed in R&D. Having someone with a design viewpoint and the depth to understand it all, made quite a difference to manufacturing engineering, and then I moved into R&D at the new RF/microwave instruments division being set up.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline