in

The Computer Monitor Testing Analogy

This entire testing situation reminds me of when I was a newly minted EE. The company I worked for was making a computer monitor that had a very high failure rate. The solution was to build an environmental chamber that heat/cool cycled the unit for 24 hours. It took a full time employee to run the chamber. The guy who ran the chamber liked to dress as Elvis which made the scene rather comical looking.

If a monitor failed in the chamber, the employee would take another memory board (which was almost always the problem) and plug it in. He’d take the failed board and set it aside — which he would then try in the next failed monitor. If a board failed three times, he’d look to see if it was actually defective but in most instances, a failed board would pass the test in a different monitor.

Being young, I thought “I wonder if there’s a design problem” which I found in about an hour (it was a violation of a setup time.) Testing never had any hope of fixing the “bug.” So what did the company do? NOTHING! They were so invested in testing that no one wanted to fix the bug and that was that.  This was one of those life lessons where a senior engineer would have had to admit that his design was fatally flawed and the company would have to admit that they’d been shipping a defective product for several years. It was a lot easier to tell the new guy to just leave it alone.

By the way, it was believed by management that the failures were “early life failures” which could be uncovered through temperature cycling. The technician who ran the chamber was well aware that this was not the case but that it was a “compatibility issue” in his words. Actually, the problem was that the timing of the memory chips had a rather wide variance  and running them too hot or too cold would shift them out of spec. This was actually OK if all of the chips on the board were out of spec in the same direction. The data sheet for the parts clearly said that they couldn’t be used in the way they were because they couldn’t guarantee they’d work.

That’s where we are today with the virus. We don’t know what we know. We also don’t know what we don’t know. And yet we think that testing will fix the problem even though it won’t. It will mask the problem to a degree by perhaps getting people to self-isolate with the hope that people that test “positive” do actually have the virus and that self-isolating actually does some good. Of course, since a virus isn’t alive and doesn’t die, all self-isolating might do is delay someone with the virus in their nose from spreading it around. Maybe that’s good enough since, just like the computer monitor, you do still get a lot of warranty repair but at least you’re still shipping product. By the time someone gets around to a fix, which we call a vaccine, the virus will probably have pretty much petered out. Or in the case of the monitor, the company stopped making them because their costs were too high. And, yes, that’s an actual picture of the monitor! The Internet is forever!

 
Mark Rosneck

Written by Mark Rosneck

Site owner and bilagáana

19 Comments

Leave a Reply

Leave a Reply

Loading…

Pelosi Includes Billions of Dollars for Illegals in New Stimulus Bill, Would Give Them $1,200 Each Like American Citizens

Happy Heretic Warhammer 40k Hobby

Welcome To The Happy Heretic