Damon Kelly wrote:

> A bit OT, I know, but I thought this forum would be a good one to ask...
> I have been designing analog circuits (for PCBs) for about 13 years now, so
> I class myself as "middle career", with some skill. However, lately when
> I've been browsing job advertisements, it seems everyone wants analog and
> mixed signal designers for ICs.
> Can anyone suggest some good books for self-study, with a view to moving
> into these fields? In particular:
> - analog IC
> - "mixed signal"
> - VHDL et al...

OHHHH, what a can of worms.  I have been doing the same.  My boss thought
he'd send me to talk to one of the experts in the field we are in (basic
research) for a week, and then I'd come back and start designing ICs.  HA HA!
I know this was totally rediculous, but I just didn't know HOW totally out of
touch such a thought was.  We got hooked up with an EE professor who has
been doing tricky analog IC designs for a few years, and looking over his
shoulder, while doing the system-level design, test boards, and control
hardware for the chip he is working on, I now have a feel for the level of
difficulty involved.  WOW!

VHDL is pretty easy, there are good books, but I can't recommend one in
particular.  It is not really that hard to learn.

But, the intricate aspects of carving each and every transistor, with respect
to width and length, to get just the right tradeoff between noise and gain, is

an enormous quagmire in larger systems.  And, there are issues of crosstalk
that are not even properly simulated by the simulator software, to give
you sleepless nights.  A small change on a PC board is something you can
try with a magnifier hood and an exacto knife, and know whether you are
onto something useful in 15 minutes.  Any change to a chip costs $8000 or
more, and you wait 12 weeks to see whether it helped or not!
The guy who is doing this with us also has some other projects he is working
of for other people.  One of their chips came back with a missing via on the
master clock input to the whole chip, due to an error in a program that
the Mentor data to Calma files for the masks, and so the entire run was a
waste of time and money!  We were lucky, all of our chips functioned at some
level, and we were able to make measurements of the performance and then
see what we wanted to change for a next try.

We are pushing the .6 uM CMOS process to the limits, in terms of noise and
detectability of low-level signals, so our problems may be a lot worse than
may run into.  We are just about to submit our 4th chip, this one has all the
control logic added to it, as we pretty much have the basic analog functions
now doing what we need them to do.

Jon (with no delusions I can do mixed signal IC design!)

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
* http://www.techservinc.com/protelusers/leave.html
* Contact the list manager:
* Forum Guidelines Rules:
* http://www.techservinc.com/protelusers/forumrules.html
* Browse or Search previous postings:
* http://www.mail-archive.com/proteledaforum@techservinc.com
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to