>   I *think* the bug was that it used a single-inverter CMOS driver for
> the crystal, and most 32,768 Hz crystals won't operate reliably in
> parallel mode.  As a result, virtually *all* PCs but the very earliest
> used, including *all* those using the 82C505, and many still do, an
> external oscillator using a 4069 with two gates in the loop to excite
> the crystal in series mode.

I used a 6818T in a product some years ago. I think this is an SPI
version of the same thing. The problem I recall with it was that it
had software programmable gain but this defaulted to the value needed
for a 4MHz crystal. That meant that it would overdrive the 32kHz one
until corrected in software. This is not good for 32kHz crystals.
Just for entertainment value, a 32kHz crystal overdriven this much
makes many transitions. Several MHz worth. So it appears to be
working, sort of.

Steve.



======================================================
Steve Baldwin                Electronic Product Design
TLA Microsystems Ltd         Microcontroller Specialists
PO Box 15-680, New Lynn      http://www.tla.co.nz
Auckland, New Zealand        ph  +64 9 820-2221
email: steveb@tla.co.nz      fax +64 9 820-1929
======================================================