[sdiy] Stupid question- in a series of 4,000 (from me)

[Magnus Danielson]

From: Peter Grenader <petergrenader at mksound.com>
Subject: [sdiy] Stupid question- in a series of 4,000 (from me)
Date: Thu, 23 Jan 2003 20:48:19 -0700

> 
> Does a discrete component logic function, let's say an or gate made with
> diodes, pull less current than an or gate IC? I reckon yes, but I just
> wanted to put it past the group.
> 
> anybody?  anybody?

Me? Me?

If you take a quad 2-input NOR chip like the 4001 it has the quiescent supply
current of typical 10 nA (250nA max) when fed with +5V in a comfortable
temperature of 25 degrees Celsius. That naturally assumes you have no major
current the output must pull, but if it is a fellow CMOS gate, then you are
not to worry, since the input impedance is in the scale of 1TOhm and only
experience a tiny rush when shift of input voltage due to the capacitance.

Got a wrist-watch? Maybe a crystal-driven thingie? There is a tiny little
battery in there, it drives a pair of CMOS inverters at the frequency of
32,768 kHz and then 15 flip-flopps for division to toss out the 1 Hz to drive
the mechanics or to drive the second, min, hour counters. It's an application
made possible since CMOS eats virtually no current at all for low frequenceis,
and only a few parts see any frequency at all.

Got a Pentium-processor (or any modern processor, like the PowerPC range)?
Thats millions of CMOS gates running at high speed, many of them at full clock
speed. Just the clock-grid eats quite alot of the power-budget. It's possible
since CMOS gates allows scaling such that power-consumption shrinks along with
it in a suitable rate.

If you do M2L (Mickey Mouse Logic) you will need pull-up/pull-down resistors
and eventually some gain element anyway. Sure it works, but if the power-
consumption was your worry then I think you should turn to the 4000-series
anyway.

Cheers,
Magnus - now to dig into the Amazon packet that just came with UPS ;O)