View Single Post
  #2  
Old September 12th 04, 05:19 AM
kony
external usenet poster
 
Posts: n/a
Default

On Sun, 12 Sep 2004 11:35:28 +1200, "~misfit~"
wrote:

Ok, I've built a strange system from parts. It's an Atrend 6310M mobo. It's
an AT, Slot 1 board. I was quite pleased to get it as I have a lot of AT
stuff (cases, PSUs etc) that I've been loathe to throw away. I welcomed the
chance to use some with this board.

I Googled the board and found a manual in .pdf format. It says that the
board supports up to a PII450Mhz and 128MB SDRAM in each of it's two slots.
Great thinks I, I can make a decent machine out of my AT stuff before I
finally biff it (The back room is half-full of AT stuff, including two boxes
of working PSUs).


Something is wrong there, any box that can run a PII450 will
accept at least 256MB (low-density) modules, 512MB or more
total capacity.



So I put it together with one 128MB module, 10/100 3Com NIC (I'm on a LAN
here), 8MB AGP card (S3 but it's adequate for it's purpose), ISA
Soundblaster 16, ISA Rockwell 56k modem and a two-port PCI-USB card. There
are pins on the board for a USB header but I don't have the header and there
are pins for a PS/2 header. I *do* have a header for a PS/2 mouse but it
just has four wires in a row and I've tried fitting it either way to the
pins on the mobo and it doesn't seem to work. I don't have a schematic for
the board or a PS/2 port to work out how to/if I can successfully connect
it.


Being AT, it's going to have the 5V line connected to PSU
main 5V rail, you could check continuity between any/all
pins to 5V, and same for ground, and have two of the 4
figured out. The other two, try one way and if it won't
work, swap the two's pin-positions and retry... just be sure
the 5V and ground are right, an ohm reading would be a good
followup to the initial continuity check.

snip

Ok, the problem. This machine runs Prime95 for 24 hours no problem. It runs
SETI CLI sweet *but*, I have the monitor set to turn off after 10 minutes,
no screen saver. If the machine has completed a SETI WU and it's got the
monitor turned off it locks up. No mouse movement or pounding of keyboard
will wake it. A hard re-set is needed. This has happened three times and
every time the first thing the PC does when it starts is complete the last
10 seconds of a SETI WU and then connect to get another one. It's connecting
through my LAN at the moment.


You might see if the NIC driver properties has different
settings for power management, and see if any bios settings
seem related and enable them... hard to say on an old
lower-end board like that.

Some boards with power management wakeup problems, benefit
from bios update.


At any other time I can just move the mouse slightly and the monitor springs
to life, it's only at the conclusion of a SETI WU that this problem occurs
*So far*.


You might try running SETI locally instead of across lan and
see if problem persists... I'd wondering if the NIC isn't
responding.


Ok, the person I'm gonna give it to isn't going to run SETI but
this could be indicitive of a larger problem. Any Ideas? I was wondering if
it's maybe happening as a consequence of the machine trying to access the
network while the monitor is off?


LOL, i should read ahead more often.

I don't know. The NIC is fine, (3Com) I
pulled it form a machine that it's been functioning in perfectly. In fact,
I'll probably pull the NIC when I part with the machine (although I like to
leave NICs in machines I give away/build for others so I can easilly connect
to me LAN if I need to troubleshoot at any stage).

Any thoughts folks? TIA.


I'd probably use/leave a cheaper NIC in it, presuming the
3COM 10/100 is PCI, not ISA, since it'd be better used for
NAS box than an old AT box. Something else I now wonder is
if the system has 3.3V for the NIC at all, you might
actually be better off with an older NIC. Since I have old
10Mb 3COM ISA NICs I can't reuse in modern boxes, that's
what i'd try & leave in it, for free at least, if NIC was
part of a system requirement then perhaps a better (faster)
NIC.

You might see if the board can be identified in a PCChips et
al flavor, or links followed produce the pinouts for similar
boards,
http://www.stud.fernuni-hagen.de/q3998142/pcchips/