War story: I remember when I worked a hardware engineer 20 years ago - power
supplies were an issue then too. It's not just the PSU, it's also the
distribution of that current from it, poorly laid out distribution can ruin
a good PSU.
I am not defending ABIT, it seems as though they may not have designed
sufficient tolerance into the board - as a consequence the variation we are
seeing may be due to production tolerances yielding boards which don't work
( I have seen this sort of thing before - even the prototype eventually
failed, even though it worked originally).
I may be teaching a few granny's here, but go fetch your eggs anyway...
Make sure the wires from the PSU are tied tightly together (this reduces the
inductance and so lowers the PSU impendance - a good thing!).
Make sure the current is evenly distributed amongst the devices that need it
(this reduces the average amount of current carried by a particular bunch of
wires - also a good thing!)
You will notice, it you look at the board that the area around the
processors and other high current demand areas is festooned with capacitors
(these increase the capacitance between the traces and so lower the AC
impedance of the supply seen by the processors - another good thing). Are
any of these broken? have a look, it WILL have a BIG effect if they are,
especially at higher frequencies!
Also, note that if you have introduced increased inductance by spreading out
the PSU wiring, this will combine with the capacitors capacitance to make a
supply which can resonate (known as ringing - a VERY BAD thing!).
Derrick.
----- Original Message -----
From: Mike Frisch <mfrisch@saturn.tlug.org>
To: Rob Hall <rwh@charter.net>
Cc: <linux-kernel@vger.rutgers.edu>
Sent: Sunday, December 19, 1999 6:39 AM
Subject: Re: [Fwd: Dual Celery on BP6]
> On Sun, 19 Dec 1999, Rob Hall wrote:
>
> > culprit)... Now, if we could get Abit to look into the problem... They
> > may have already tried this, but it's always good to know...
>
> I wish ABit were more involved in solving this problem. Being the
> designer and manufacturer of the motherboard, they most definitely would
> have the required test equipment to properly diagnose the problem at hand.
> If the problem ends up being a particular sensitivity to cooling and/or
> power supply, life is good and those with this board (myself included) can
> make provisions to fix the problem, otherwise we'll be wondering forever
> as to what the cause is.
>
> > token, other people with the exact same configurations aren't having
> > problems... That
>
> I am not entirely convinced of this. Some people who've reported their
> machine being 'clean' aren't even running 2.3.x kernels that log these
> APIC errors. I've 'caught' at least one person myself.
>
> > should look into it.. they've got a fine product here, I'd hate to see
> > it shot down by something as trivial as a regulator....
>
> I agree completely!
>
> Mike.
>
> ======================================================================
> Mike Frisch Email: mfrisch@saturn.tlug.org
> Northstar Technologies WWW: http://saturn.tlug.org/~mfrisch
> Newmarket, Ontario, CANADA
> ======================================================================
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.rutgers.edu
> Please read the FAQ at http://www.tux.org/lkml/
>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/