Date: Sat, 21 Aug 1999 13:22:38 +0200 (CEST) From: Achim Patzner <ap@noses.com> To: freebsd-alpha@freebsd.org Subject: Re: Installing FreeBSD on a Multia Message-ID: <199908211122.NAA18179@noses.com>
next in thread | raw e-mail | index | archive | help
: From what I've recently learned, using a Multia is a Really Bad Idea.
: Yes, they are cheap, but that's the ONLY good thing about them. Both
: David O'Brien and Peter Wemm were able to burn their's out just by doing
: make worlds. They are underpowered, overheated, undersized, but very
: very cheap. Ever wonder why?
Mine died after I tried installing NetBSD (which doesn't really mean that
NetBSD is that bad an OS 8-) ). I paid about USD 60 for each hour it ran.
But as the NetBSD people (obviously - they started quite a bit earlier) had
lots of them they started repairing them (better: improving their health 8-)
).
Take a look in the NetBSD/alpha archives to find out more about un-frying
Multias.
Achim
Subject: more Multia repair part numbers and comments
To: None <port-alpha@NetBSD.ORG>
From: Ross Harvey <ross@teraflop.com>
List: port-alpha
Date: 04/28/1998 02:35:15
OK, so I borrowed a supposedly dead Multia. Murphy's law is in fine form,
though...when you _want_ a dead one, _specify_ a dead one, really can only
_use_ a dead one (for the purpose of doing a friend a favor and testing the
big multia fix) ...then...of course...it arrives and works just fine.
Now that I've seen the parts, the bad news is that they aren't likely
to be in any catalog...good thing for me no one took up my offer on that
bet. AFAICT it is going to be necessary to order them through one of the
manufacturer's franchised distributors, and they can be difficult when you
don't already have an account. Sometimes you can order samples through
a web page, and certainly you can locate the distributer that way.
The generic thing you want is: (qty 6) a 20-lead .220 (i.e., medium
body) soic 4.7K bussed (pullup) resistor network, with lead 20 as the
common. I think the Bournes parts are probably safe, I guess it was a
bad batch, or possibly that thick-film part didn't like actually being
_in_ the solderwave, which is what would have happened given that these
components are on the bottom of the board.
And, I can now see why the Jupiter guys replaced the logic IC: on Tim's
Multia the printing on the chip is starting to bake into a brownish
off-white right in the center, i.e., over the die. I think it needs to
go too, although using thermally conductive epoxy you could conceivably
heatsink it...not much room to work with though. The 5% is perfectly
fine as a tolerance, but if it is out of stock and the 2% or 1% is in
stock...
WhichOne Mfr Part Number
R-network Bournes 4820P-002-472
R-network Dale SOMC2001472J (5%)
R-network Dale SOMC2001472G (2%)
R-network Dale SOMC2001472F (1%)
...there are probably more...
Logic-IC TI SN74ABT623DW
Logic-IC Philips 74ABT623D
Logic-IC Pericom PI74FCT623TS
Logic-IC Pericom PI74FCT623ATS
Logic-IC Pericom PI74FCT623CTS
Logic-IC Pericom PI74FCT623DTS
Logic-IC IDT IDT74FCT623TSO
Logic-IC IDT IDT74FCT623ATSO
Logic-IC IDT IDT74FCT623CTSO
...there probably are not any more...
This is almost trivial with the right lab, but it definitely isn't the
easiest home project. Besides the difficulty of obtaining the components,
you have 140 leads to remove and resolder...and you don't want to lift
or destroy a single pad while doing it. At least it is a lot easier
than if they were those ugly old dips.
Ross Harvey
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-alpha" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199908211122.NAA18179>