Path: utzoo!attcan!uunet!husc6!mailrus!cornell!rochester!uhura.cc.rochester.edu!ur-valhalla!micropen!dave
From: dave@micropen (David F. Carlson)
Newsgroups: comp.sys.ibm.pc
Subject: Re: need 12Mhz controller that runs SCO Xenix
Summary: sounds familiar
Keywords: controller Xenix bus speed
Message-ID: <596@micropen>
Date: 22 Dec 88 15:07:59 GMT
References: <8812220135.AA04469@brahms.berkeley.edu>
Organization: Micropen Dirent Writing Systems, Pittsford, NY
Lines: 33

In article <8812220135.AA04469@brahms.berkeley.edu>, cotner@BRAHMS.BERKELEY.EDU (Carl Cotner) writes:
> 
> I have an AT clone that runs at 12MHz;  unfortunately, I think the bus
> speed also runs at that rate.  I've tried to install a Western Digital
> hard disk controller WD-1003-WA2 on my motherboard, but I'd get write errors 
> on my hard disk which end up corrupting the file I was trying to access.
>I also have an NCL controller which runs fine at 12MHz, however this controller
> Carl
> cotner@brahms.Berkeley.EDU

I have had a prominent clone maker give a fairly reasonable explanation for
the failure of certain disk controllers at high speed EVEN if the bus is run
at 8MHz compatibility mode.  Particular among the failures is the most common
of all AT boards:  WD1003-WA2!

It seems the bus precharge cycle is a function of bus length (reasonable given
capacitance over the bus, etc.)  When the bus is speed up, the precharge cycle
must remain relatively fixed (as capacitance would also remain fixed) however,
certain boards (WD1003 among them) calculate the precharge timing themselves
and access the bus at inappropriate times.  Ouch.  We thought it was us for
a long time and I got quite far into tech support before this little gem leaked
out.  (And yes, UNIX installs gave the most troublesome operation.)  This 
manufacturer claims that *EVERY* high speed AT bus computer will have the
problem because it is one of physics and incomplete specs.  That includes
12/16MHz 286, all 386 and even 386SX.  Certain cards that were made for high 
speed operation do work:  the new WD1006-V is a good bet if you can get one!
(1:1 interleave controllers are almost all better than 3:1 for high speed 
buses.  This is a gross generalization but essentially correct.)
-- 
David F. Carlson, Micropen, Inc.
micropen!dave@ee.rochester.edu

"The faster I go, the behinder I get." --Lewis Carroll