The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: "Michael A. Terrell"
Subject: Re: NEED A TRANSFORMER - WE MANUFACTURER
Date: Tue, 05 Nov 2002 01:03:05 -0500
Organization: Have you seen my bench? No, really! Where is it?
X-Mailer: Mozilla 4.61 [en]C-CCK-MCD (Win95; U)
References: <email@example.com> <firstname.lastname@example.org> <3DBF35E2.BC89DF0F@mfi.net> <email@example.com> <3DC60E49.DAA3E5CE@mfi.net>
Winfield Hill wrote:
> "Michael wrote...
> > THE MEETINGS! NOT THE MEETINGS, PLEASE! I PROMISE ANYTHING,
> > BUT NO MORE MEETINGS!
> > Sorry, I just had a flashback.
> > You also have endless quarterly audits with people who don't
> > understand your business, processes, or that every business isn't
> > just made from the same mold. Try to explain to a paper pushing
> > auditor why you need to use an alternate method to test a board
> > or module because the only specified piece of test equipment is
> > not available because three are gone for calibration, two more
> > were junked because you can't get parts, and the only working
> > unit is on loan to the engineering department and it is thousands
> > of miles away. Or that you are doing special test for engineering
> > because they are short handed, and they need to see if a new design
> > / modification / component / software rev works properly on the
> > production floor. They neither care, or understand. ...
> Actually, even tho I never subjected my company to the process,
> I think it's a good idea. Perhaps one key is the right kind of
> redundancy, extra people to make rough things go right, multiple
> copies of critical test fixtures, made by skilled technicians who
> work very lightly under the nominal supervision of the original
> creators, etc. With multiple back-up teams creating the permanent
> documentation for the creative types who usually do most of the
> best work in the technically-oriented firms. Along the lines of
> that great tome, "The Mythical Man Month." But this can't happen
> without the right management understanding and determination.
> - Win
Win, the company was started by two engineers who were still
consultants 30 years later. One was in charge of the test equipment
purchases. We had a room full of old test equipment left over from the
days when they built satellite TV receivers for the CATV & MATV
industry. He couldn't see buying a piece of equipment if there was any
way possible to cobble together a couple benches and carts full of older
junk to do a half assed job. It took several years to get the first
Another engineer insisted that no EE would ever need a computer. If
they couldn't design the project with a slide rule and a pad and paper
they should retire.
Management was no better. They didn't add anyone to handle the extra
work caused by the ISO 9001 certification. Everyone was expected to
"Pick up the slack and work harder!" For instance. Their 700 series
receiver AGC system took a tech about three to four hours to test and
align. I spent two hours to build an improved fixture with all the
calibration voltages preset to less than a 1 mv error. I was able to
turn the boards out in 18 minutes, on average. This included all minor
rework, and finding swapped parts. I got chewed out for building the
fixture. Management wouldn't spend 10 cents to save $1000. When they
closed the plant to move it to Pennsylvania the company president had
started as an intern in test 25 years earlier, and he had the same
mentality of the two original engineers. It was ok to spend an hour
doing select in test to save $1.00 on a better RF transistor.
Win, the biggest problem I had with it was management froze
everything after receiving ISO 9001 certification. Techs were not
allowed to keep marked schematics on their bench with any
troubleshooting notes, all notebooks were to be destroyed because you
could only use official documentation. That meant the only way to teach
a new tech was to dump a job on his bench and hope someone could
remember any quirks. I worked on 16 layer VME cards that had just been
released from engineering to manufacturing. Test fixtures didn't exist
for these. Engineering told me they didn't have time to build them, and
if I needed them I had to get permission to build my own. The
documentation was lousy and the "Junior Engineer" who wrote it refused
to correct anything because he didn't use it to test boards in
engineering. He also liked to make up words, and his command of english
was poor. What "fixtures" we did get from them were for very old
products and nearly useless. They were really engineering jigs someone
had thrown together to see if a design worked.
Some boards had the same test run at four different levels, and the
only one that mattered was when the full set of boards was in a card
cage. It was a pair of DSP based telemetry receivers with a combiner.
Several thousand parts in each unit, and they made a number of mistakes
in the early design. For instance: A 125 MHz oscillator for a DDS
(AD8950) was phase locked to a 10 MHz internal, or a 5 or 10 MHz
external source. You had to select resistors to set the bias on a
transistor in the loop to get the range they wanted. They insisted that
the circuit lock over a wide temperature range, and plus or minus 120
Hz from the reference at 10 MHz. This raised the phase noise in the
circuit, but they claimed the DDS would magically remove all the noise.
When I asked why they didn't use a single op amp and a trim pot to
eliminate the hour it took to get the circuit set up properly they gave
their stock answer, "You wouldn't understand, you aren't an engineer."
Somehow, I managed to get a lot of the newer test procedures cleaned up
by writing new ones myself. Then the test supervisor managed to get
permission to authorize the new documents in spite of engineering. Of
course, this happened after a couple bad audits, and missing some
I think one reason L3-Com closed and moved the plant was so they could
get rid of all the dead wood in management and engineering. They only
took the five engineers and a couple ETs involved in designing the
newest product, then they either retired or laid off everyone else in
I am sure ISO certification works well in some places but you have to
understand how it works and be willing to make changes to both the way
things are done, and to the ISO process. I was caught in a company that
decided to do neither.
Michael A. Terrell
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup