On 3/8/16 8:27 AM, Richard (Rick) Karlquist wrote:
On 3/8/2016 3:18 AM, [email protected] wrote:
Good Morning,
technically you are correct, most buy what they find and live with a
compromise. But companies like mine, R&S, test equipment , need superior
performance and many parts which we need, we have made by foundries.
Numerically controlled oscillators belong to this and modern IQ
modulators and arbitrary wave form generators are the norm., much better
then many analog type designs. Most chips on the market are
compromises for power consumption and phase noise. We now have fraction
I worked for HP/Agilent for 35 years, retiring 2 years ago just before
the Keysight spinoff. Yes, they do have proprietary chips made by
internal and external foundries (for example my phase detector). Other
than that one case, I was never able to get any custom
chips made during my career because of the high NRE cost and
the need to have either very high volume, or an extreme value
proposition to cover NRE. There was a small group of designers
who designed a handful of fractional-N chips. The rest of
us were merely users of them. Newer designs have tended towards
COTS frac-N chips. Similarly, there was a small department of
designers of NCO's, AWG's, etc (I was in the same lab with these
guys), and the rest of the engineers were merely users of these
IC's.
So in terms of the book market, it would be limited to a small
fraction of the engineers in test and measurement. And that
small fraction probably has already gone way beyond any
technology that has made it into textbooks. A lot of this
really advanced work is based on trade secrets that of
course will only appear in internal documents.
Or is subject to export controls because of the specific application.
While you could probably "generalize" it to get out from under export
controls, that's a lot of work. Ulrich and Rick raise interesting points:
The people who really know about this stuff do it for a living, all the
time, and probably don't have a lot of free time. I've done a couple of
book chapters, and it's an enormous amount of work, and that's with
collaborators and editors to help. I suspect that for these people the
problem is not a lack of "funding" from the employer, but, rather, that
there is more work to do than people to do it.
And, often, the "state of the art" is either proprietary or subject to
other controls on distribution. Unless you are in a special situation
(e.g. you own the company, or it's a small company and the owner(s)
agree), I can see management not seeing the "value added proposition"
for letting your talented, knowledgable PLL guru work on getting into a
form suitable for publication: they'd rather you be making boxes.
_______________________________________________
time-nuts mailing list -- [email protected]
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
