On Nov 28, 9:01*pm, Michael Black wrote:
On Mon, 28 Nov 2011, K7ITM wrote:
With regard to the parts that places like DigiKey and Mouser and other
authorized distributors stock: *I seriously doubt that any homebrew
projects are going to have much effect on what they stock. *It's the
parts that are bought for production, by the reel, or at least by
hundreds for some of the more exotic parts, that get their attention.
If you see them with a stock of 10,000 or more of something, you can
bet they didn't get them because they got some orders from random
homebrewers.

But in the old days, it did matter. *When there were local parts stores,
and small mail order places. *Why were there all those 2N706s specified
for a long time, until 2N2222s became common?
....
Hey, I'm just guessing here, but I'd bet that the 2N706 ("High speed
logic switch," says the data sheet I brought up) was used by the
hundreds of thousands, maybe by the millions, by computer
manufacturers and other industrial/military electronics suppliers back
in the pre-integrated circuit days. That made them inexpensive and
readily available, and on the surplus market. Remember
"PolyPaks" (sp?)? I still have the 30MHz universal counter I built
out of PolyPaks parts, including Nixie tubes. I certainly had no
illusions about me and hobby people like me driving the market for
parts. I remember having a distinct impression at the time that the
hobbyist parts companies bought what they could get their hands on for
very low cost and sold what they could to us geeks. There's a supply-
side reason a lot of those companies were located in the Boston tech
corridor, Silicon Valley, or the L.A. aerospace zone. Yes, certainly
some of them bought stock of things they could move from distributors,
but I still don't see that the hobbyists drove what parts became the
standards.

The people designing ICs these days do it so they can sell large
quantities of them. That means getting them designed into reasonably
high volume products, and that's NOT the hobby or ham market. Heck,
the stuff I design professionally is never sold in high enough volume
to warrant design of any electronics parts 'specially for it, and it's
produced in higher volume than any but the highest volume ham gear--
orders of magnitude more volume than any homebrew projects.

Perhaps we're looking at this from completely different viewpoints.
I'm looking at it from the point of view of a company that makes and
sells, say, semiconductors. From that point of view, you don't ask
how many the hobby market will buy--you find markets that need
millions of parts. If you're successful, you don't mind that a tiny
fraction of those parts end up in hobby projects, but when the main
market dries up, you are unlikely to continue making wafers of that
part just to supply homebrewers.

That's not to say the home experimenter or low-volume industrial user
has no impact on the large (or smaller) semiconductor manufacturer, by
the way. I know that feedback from a ham trying to get low phase
noise, who happened to discover a problem with a particular chip, had
a significant effect on the manufacturer of that chip. I know that my
own feedback to a very large semiconductor manufacturer caused them to
discover a furnace problem and they shut down that line to fix it. I
know that my feedback to the semiconductor product planners that come
by occasionally has some effect on what they decide to do because they
respect that what we do is pretty much cutting-edge, but the ideas
would never see the light of day if they don't see a significant
market for the resulting products (and that market is never going to
be because of things I design, for sure).

Cheers,
Tom