On Tue, 28 Feb 2017 14:50:27 -0500, micky
wrote:
In sci.electronics.repair, on Mon, 27 Feb 2017 15:36:52 -0800, Jeff
Liebermann wrote:
(...)
I agree with you here.
Then something must be wrong. Nobody ever agrees with me.
The JEDEC would have
been better off just starting at 00001 and counting up in sequence. It
I don't agree they would be better off. The system was valuable for
many early tubes and I didn't expect it to work forever. so I ignored it
when it didn't work. Numbers in sequence would have meant nothing,
Exactly. The whole idea is that they are meaningless so that advances
in tube technology would not be lost in a jumble of letters and place
holders. It didn't take long for all the tube numbering systems to
fall apart. The first problem was when manufacturers started putting
two different tubes in one glass or metal envelope. For example, the
12EC8 is actually a triode and a pentode in one package:
http://www.radiomuseum.org/tubes/tube_12ec8.html
Even though the "8" was the number of element, there was no way to
decode that into a 3 and 5 element device.
The next to go was the Compactron with multiple devices inside the
envelope and more pins on the base. The habit of significant
numbering was maintained, so they too had numbers that were difficult
to decode. Some Compactrons ended in "11" for a triple triode where
two triodes shared a common cathode. No way to deduce the internal
structure with just the total number of elements:
https://en.wikipedia.org/wiki/Compactron
I think the final change to non-registered numbering was the Nuvistor:
https://en.wikipedia.org/wiki/Nuvistor
Some Nuvistors used the significant numbering scheme (6CW4, 6DS4, etc)
while others used non-significant 4 digit numbers (7586, 8056, etc)
depending on what the target market wanted.
The old adage "Those who don't remember the screwups of the past, are
condemned to repeat them" is quite true. However, the problem isn't
that they don't remember, it's that they never experienced the
problem. I've worked for or with 3 companies all of whom had
significant numbering systems fall apart on them. Renumbering a
companies entire product option system and internal part numbering
system, is not a trivial or painless exercise. Each company had it's
own "better" idea, which eventually failed (usually after the
perpetrators were long gone).
Oddly, some of these numbering failures creating an even bigger
problem, where individual departments each contrived their own
numbering scheme. This problem is fairly common in the consumer level
computer business. Marketing would have a clever name for the
product, engineering would have a code name, production would have
it's own product number, and sales would build a catalog of short and
simplified numbers that could be easily remembered by customers. While
immensely confusing to the typical customer, it does have some
organizational benefits for the manufacturer.
Unfortunately, I was partly responsible for creating a significant
product numbering scheme at a former employer. It took about 8 years
for the system to fall apart, which is about typical. Adding a few
extra digits required 2 years of rewriting every computer program that
the company was using. A few years later, the company went out of
business, just before they again ran out of significant digits.
--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558