[time-nuts] high rev isolation amps

Charles Steinmetz csteinmetz at yandex.com
Wed Mar 30 18:00:19 EDT 2016


Rick wrote:

>If only this were true.  Authors at NIST have consistently told me 
>that the conditions of working for the government preclude them from 
>mentioning the names of vendors.  Thus you get generic JEDEC 
>transistor numbers.  Along with the fact that JEDEC numbers are 
>useless concerning unspecified parameters, it tends to make NIST 
>designs non-reproducible.

It's often much worse than that.  NBS/NIST tends to require new 
designs to be compatible with existing equipment crates and frames, 
so designs are routinely compromised by legacy form factors, 
technologies, power supply availabilities, and other considerations 
that preclude building the best widget they know how to 
design.  (There are some other contributing factors, as well.)  You 
might think NIST wouldn't settle for anything less than the best they 
can do, but in fact "good enough" often rules the day.

>can you please let the group in on the actual vendors and part 
>numbers of the new era transistors accidentally having low flicker noise?

Zetex (Diodes, Inc) pioneered this new technology.  There was some 
discussion on the list last summer (see, for example, 
<https://www.febo.com/pipermail/time-nuts/2015-July/092985.html>). 
There are also some discussions in the volt-nuts list archives.

I (and others) discovered these parts back in the '90s and have been 
using them "off label" for ultra-low voltage noise designs for 20-odd 
years.  Then, the 3rd Edition of Horowitz and Hill's "The Art of 
Electronics" made it public knowledge.  (Speaking of which, Chapter 8 
is absolutely essential reading for anyone who even pretends to be 
designing low noise circuits.)

Best regards,

Charles




More information about the time-nuts mailing list