Re: More ZIP stuff...

[dempson@atlantis.actrix.gen.nz (David Empson)]

In article <4b7etl$aon@hearst.cac.psu.edu>,
Jason Jeffries <JJeffries@psu.edu> wrote:
> 
> The Zip seems to "round" the speeds of the oscillators I use. I have a
> 48 and a 50 (I've scavenged them--look on old PC video cards); both
> result in 12mhz. 
> My crystal for 14mhz operation is actually s 56.644 mhz part--strange
> but it works.

I assume you are looking at the number reported by the ZIP utilities
(the CDA, for example).  This number is by no means exact.

The ZIP measures its own clock speed by using an internal fixed-speed
clock (1 ms, I think) and counting clock cycles.  This is probably
rounded to the nearest integer before being displayed.

You will be getting 12 MHz with a 48 MHz crystal, and 12.5 MHz with
50 MHz.  Your 56.644 MHz crystal will be producing 14.161 MHz.

> Is there any way I can test cache RAM, other than doing many RAM
> intensive things and waiting for crashes? The TWGS, if I remember
> correctly, had a cache test built-in. Is there an similar application
> for the Zip?

No.  The ZIP has no ROM (the TransWarp does), so it has no built-in
code for doing any kind of self-test.  This sort of function should be
provided by the disk-based utilities, but I don't recall seeing one.


> Lastly, while working on my Zip, I noticed there were parts that seem
> to be "missing". What's missing, and can these parts do anything to
> enhcance my existance if I were to put 'em on?

There are a lot of options on the card, some of which are
mutually-exclusive.  The main one (involving Q2 and several other
components just to the left of the slot connector) is an adjustable
regulator which supplies the CPU with a higher voltage (usually about
5.5V instead of 5V) derived from the 12V rail.  My second ZIP had this
circuit installed.

The higher voltage allows the processor to run more reliably,
especially at higher clock speeds ("slew" times are reduced if the
processor is running at a higher voltage).

The voltage needs to be set carefully, because the rest of the
computer can be damaged if the CPU is outputting signals which exceed
5V by too much.


There are also other options for supporting options such as "split
cache" (which I know nothing about).
-- 
David Empson
dempson@actrix.gen.nz
Snail mail: P.O. Box 27-103, Wellington, New Zealand