The ETPL - John Dunn, Consultant, Ambertec, P.E., P.C.

In the further avowed interest of cost savings at the company that had lost the services of The Sculptor, the purchasing department often bought components from one particular semiconductor company who always seemed to be the lowest cost supplier. The name of that semiconductor company I shall not state, but it is no longer in business, thank heavens. These purchases were made in spite of it being repeatedly pointed out that purchased items kept coming in non-compliant to specifications or just plain defective.

The purchasing department's impregnable defense to such input was "The vendor says they meet the specs." even though nothing could have been further from the truth. Thus, as each defective item caused this or that problem, each of those problems was duly entered into the official Engineering Technical Problem Log, the ETPL.

It fell to one unfortunate man in the engineering department, a physics major as it happened, to maintain the ETPL which grew and grew and grew and.... as equipment failures kept mounting, all of which, according to the department manager, the same one who had lost The Sculptor, was the Engineering Department's fault.

The ETPL's openly stated purpose was to ensure that Engineering didn't let any problems continue on by way of neglect.

These were some of the parts problems:

There were operational amplifiers that were supposed to be type µA741, but were mislabeled as "µA714". This was several years before Fairchild Semiconductor actually introduced their µA714 op-amp. Also, all of these mislabeled '741 op-amps oscillated at 1 MHz in any and all circuits to which they were applied.

There were 1N914 diodes that would fail open-circuit like fuses at currents as low as 5 mA. The ribbon whiskers inside these diodes were misplaced on the die and some could be seen to be just barely touching the die at a tiny little piece of the die's corner.

A diode problem of this kind was illustrated in a wall calendar photograph of a few years ago from Hi-Rel Laboratories in Spokane, WA:   

 

 

(Photo used here with the permission of Hi-Rel Laboratories.)

There were 2N2222 transistors with DC betas as low as five at room temperature.

Some years before all of this was happening, back when I worked at a really good company in New Hampshire instead of at the place I've been referring to, that same semiconductor maker presented a seminar to introduce their "improvement" over the then dominant µA709 op-amp. Every semiconductor maker wanted to dislodge the '709 in favor of their own product(s) and this bad-news semiconductor company was no exception.

There I was at the seminar, with the successor to the '709 in glorious display for all to gaze upon in wonder, awe and astonishment. Data sheets and literature were in abundant supply.

I was astonished all right! This op-amp needed external stabilization consisting of a 10 µF film capacitor that was nearly the size of a flashlight C-cell.

The next morning, I and my co-workers who had attended the seminar had some interesting discussions over morning coffee and the purchasing people at that good company never bought anything from that supplier.

An illness in the family forced me to relocate back to New York and so I had to leave the good company.

Little did I know what was to come about at my next place of employment.

 

Comments

[John Dunn]

It didn't occur to me at the time (almost forty years ago), but with so very many junk parts coming in, I now wonder if those incoming parts were counterfeit? I guess I'll never know.

[Tom Terlizzi 10-19-2011]

John,
thanks again for another Interesting article. "When I was a boy", I built a UA709 hybrid with the frequency compensation circuitry in a TO-8 package in the early 1970's for General Instrument. I never knew how much of a problem it was to tame the op-amp until reading your blog.

The problems you saw in the early 1960's with semiconducors were before the mil-specs, ISO9000, etc. were written and when "garage semiconductor operation" could advertise they had an equivalent part without an liability or they re-branded scrap material which was very common on transistors and diodes.

The counterfeiting today is more of an outrageous fraud as seen on "60 minutes" and is currently under congressional legislation
and certifications for distributors.........
I saw Senator Carl Levin & McCain grilling several government contractors and top DOD officials this week including
LTG General O'Reilly Director of the US Missile Defense Agency (MDA) see link below
http://levin.senate.gov/newsroom/press/release/background-memo-senate-armed-services-committee-hearing-on-counterfeit-electronic-parts-in-the-dod-supply-chain

http://www.defensenews.com/story.php?i=8183313c=AME

[anthony stewart]

Those guys in Ohio buy millions of $ in electronic spare parts every day in auctions for Military use with central purchasing in Ohio, ought to know better. But then there are so many shlockers in that biz, Certificates of Genuine Origin are printed on toilet paper then photoshopped onto original letterhead and printed . ha