🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.sys.sun.hardware
Article #3049

Re: 3-chip 30-pin parity SIMM's in Sparc 1, 1+ and 2 machines

#3049
From: Mike Santangelo
Date: Tue, 02 Jun 1998 00:00
61 lines
3460 bytes
Craig Dewick <cdewick@lios.apana.org.au> writes:

Sorry, Sun4c systems cannot make use of 3 chip 30pin SIMM's, said SIMM's
have different and imcompatible refresh timing constraints than
the 9 chip versions.  They will pass the firmware memory tests but
once the OS is loaded you'll get all manner of strange behaviour and
crashes.  Its a shame yes but you have to get the somewhat more
expensive 9 chip version to make your Sun4c system happy.

>Hi folks,

>I've received a batch of 3-chip 30-pin parity SIMM's which I'm going to use
>to build up a couple of sun-4c machines to 64 meg of RAM (they currently
>have no RAM fitted). The SIMM's were sold to me as suitable for sun-4c
>machines by a regular advertiser in the newsgroups.

>However, I'm having troubles...

>The modules's are QSIMM QS944x4LP's, and are dated as being made on the 11th
>week of this year (1998), The RAM chips are surface mount, resulting in
>modules that are extremely thin (about 3.5mm from the back of the PCB to the
>top of the device encapsulation).

>I've noticed problems with these SIMM's. They pass the POST RAM test fine
>(usually), but when starting the kernel it fails with a watchdog reset.
>I've tried swapping banks around, swapping entire banks out of the
>machine, etc. all with mostly the same result. 

>Sometimes the machines will boot, but fail soon after. 

>I've noticed that the RAM chips (marked as made by Mitsubishi) are not the
>same speeds, which seems odd. The two 4-bit devices are 70 ns, while the
>single-bit device is 60 ns. I don't see this as being a problem in itself,
>since I'm suspecting the capacitive loading resulting from the SIMM's being
>3-chip instead of 9-chip is what's upsetting the DRAM controllers.

>3-chip SIMM's have traditionally been seen, afiak, as a bad idea for early
>Sparcstations. 

>I did try swapping out all the 3-chip SIMM's for some 9-chip ones I have
>spare from another system, and they worked without a hitch, so perhaps the
>old adage has come out of hiding to show it's still applicable?

>What are your opinions about this? I don't think the supplier of the SIMM's
>will object to them being returned if there is a genuine problem since I'm a
>regular customer. 8-)

>Regards,

>Craig.

>--
>            Craig Dewick. Send email to 'cdewick@lios.apana.org.au'
>  Point a web browser at 'http://lios.apana.org.au/~cdewick/sun_shack.html' to
>    access my archive of Sun information and links to other places. For info 
>      about Sun Ripened Kernels, go to 'http://www.zeta.org.au/~craig/srk'

--
-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
Mike F. Santangelo, Dept. Head-Computer & Network Systems, UMCES/CBL Solomons
Technical Details
Message-ID: <6kvl50$7lb$1@gamera.cbl.umces.edu>
Path: arc1-ca.pugleaf.net!unknown!not-for-mail
References: <Pine.GSO.3.96.980530071603.697A-100000@lios.apana.org.au>

Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: f4671,495e04127a63dcd0
X-Google-Attributes: gidf4671,public
X-Google-Thread: f88e1,495e04127a63dcd0
X-Google-Attributes: gidf88e1,public
X-Google-Thread: 10d053,495e04127a63dcd0
X-Google-Attributes: gid10d053,public
From: mike@cbl.umces.edu (Mike Santangelo)
Subject: Re: 3-chip 30-pin parity SIMM's in Sparc 1, 1+ and 2 machines
Date: 1998/06/02
Message-ID: <6kvl50$7lb$1@gamera.cbl.umces.edu>
X-Deja-AN: 358646070
References: <Pine.GSO.3.96.980530071603.697A-100000@lios.apana.org.au>
Organization: University of Maryland, Chesapeake Biological Lab.
Newsgroups: comp.sys.sun.hardware,aus.computers.sun,alt.sys.sun