[GTALUG] mysterious restarts

[James Knott]

On 06/16/2016 11:00 AM, Lennart Sorensen wrote:
> On Wed, Jun 15, 2016 at 06:15:30PM -0400, James Knott via talk wrote:
>> > At work years ago, we had "no break power", where incoming AC ran a
>> > motor connected to an alternator and an 8 ton flywheel.  When the power
>> > failed, a clutch would kick in the diesel, with the flywheel maintaining
>> > the power, while starting the diesel.  One problem though was the output
>> > AC frequency was slightly low and threw off the real time clocks in the
>> > computers.
> Who built a computer that cared about AC frequency to drive the real
> time clock?  Never heard of anyone doing that.

Data General Nova & Eclipse computers had a choice of AC power or
various clock rates derived from a crystal.  The crystal wasn't all that
accurate, but normally AC is.
> Must have been a large machine given all microcomputers are fed DC from
> their power supply and have no clue nor care what the AC frequency was.

The Data General Nova predates microprocessors.  It was first build
around 1969, IIRC.  The Eclipse, while a later generation, used the same
basic I/O board, which included the RTC circuits, along with serial port
for the console and ports for the paper tape punch & reader.

Incidentally, one thing I did was modify those boards from 20 mA current
loop to RS-232 for the console and also replaced the fixed crystal
serial port clock (you had to change the crystal to change speeds) with
a baud rate generator chip that used a colour burst crystal to generate
a variety of baud rates.  With the 2 mods, those boards moved from 110
b/s Teletypes, to 9600 b/s CRT terminals for the console.