Well . . . here we go . . . the fact that N* memory mapped their FDC was one
thing that clearly would fall in the MISTAKE category. What the reason for
the existence of the smaller TPA resulting from memory mapping anything is
of no relevance. It was a justification for SOME of us, me included, to
draw a line through their products whenever they appeared in a list. Of
course their price would have been another.
Nevertheless, I ALWAYS noticed their ads in BYTE or KILLABROAD. They looked
nice . . . kind-of like an ALTAIR or IMSAI, but without those annoying
switches.
Dick
-----Original Message-----
From: Allison J Parent <allisonp_at_world.std.com>
To: Discussion re-collecting of classic computers
<classiccmp_at_u.washington.edu>
Date: Sunday, October 31, 1999 12:50 PM
Subject: Re: Northstar Horizon
>
><well, too, but rather that the N* environment is so limited, particularly
i
><the Horizon with its 8K ROM space, that many compiled programs won't work
>
>There is only 2k of ram space (e800h to efffh) the upper 4k is usable.
>This space is carved out by the memory mapped disk controller not the cpu
>any other component.
>
><because the TPA is too small. What's more, the FDC isn't capable using of
><CP/M-standard (IBM-3740, SSSD 8") diskettes.
>
>Typical TPA using a NS is 56k, though if you put the bios in the top 4k
that
>58k.
>
>ALL if that is ONLY if the NS* controller is used as the rest of the system
>is not biased in any way by eprom/rom maps.
>
><They're OK as a curiosity, but back in the lat '70's and early '80's, they
><were not well received because of the TPA and FDC issues mentioned above,
><and I warn everyone off them due to their resulting limitations.
>
>They were popular and widely used, thats why they are common. The greatest
>featur of them at that time was they worked more so than most of the other
>s100 gear. No if you want MMU equiped CPU and DMA controllers there were
>few if any of those before the early 80s and the NS* is a 1977 machine.
>
>I have two, one running a MDS-A single density controller in the
>configuration you'd find on in back in 1977 save for I have 3 half height
>floppies where the 2 full height ones are and a half height (st225) hard
>disk using a teltek controller on with a 52k tpa (the hard disk driver
>eats 2k) and the second has a softsector controller of my design that is
>62k tpa, the z80 is been modified (different crystal) for 8mhz and supports
>a hard disk (also teltek controller). I consider them fine 4mhz z80 CPM
>systems and they run everything. NS* dos is also ok and the hard sectored
>(real NS* controller) runs UCSD Pascal P system as well (I have the
original
>NS* build I bought in 1978).
>
>As a collectors sytem or very exciting Z80 system these are not it. They
>were too common for collectable and are only classic. As a really fancy
>no holds z80 system again they were vanilla. They did get purchased in
>gobs around '78-80 as business turnkey boxes as they were known to work and
>most of the bad press they got was the shugat SA400 floppies (not rugged
>drive!) and the 16K NS* ram was not so good (plenty of other ram cards
>were substituted very successfully).
>
>You want a hot (s100) z80 system? Look at CCS, Compupro, Morrow, Vector
>or one few get to see a full front pannel Ithica Intersystems.
>
>You want fast... Teltek or SDS single board s100, these were Z80 4/6/8mhz,
>2 serial, parallel printer and FDC plus 128k ram on one card.
>
>The one system turned 21 this year! The other was a 1980 build.
>
>Rather than get into the S100 vs XXX bus... S100 was a really badly
>designed bus (fell in to it would be the most kind) but it worked if
>understood it's quirks. At early 80s the issue was cost and software
>not the bus used anyway.
>
>
>Allison
>
>
Received on Sun Oct 31 1999 - 20:53:04 GMT