[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: Present Signals
- To: Mailing List Recipients <pci-sig-request@znyx.com>
- Subject: RE: Present Signals
- From: Jeff Dahlin <JDahlin@appiantech.com>
- Date: Tue, 15 Apr 1997 17:46:00 -0700
- Encoding: 87 TEXT
- Resent-Date: Tue, 15 Apr 1997 17:46:00 -0700
- Resent-From: pci-sig-request@znyx.com
- Resent-Message-Id: <"WOdfH1.0.Ml.D82Lp"@dart>
- Resent-Sender: pci-sig-request@znyx.com
Quite a few have responded to this. This is a composite of the various
suggestions.
>----------
>From: John R Pierce[SMTP:pierce@hogranch.com]
>Sent: Tuesday, April 15, 1997 1:09 PM
>To: Jeff Dahlin
>Subject: Re: Present Signals
>
>More likely, your boards are drawing more current than the supply can deliver
OR they are presenting so much load on the PCI bus that it is not functioning
>correctly.
>That was my first guess. We have a P5-200 system with a 235W power supply
>and it successfully ran with 4 boards in it. Our customer has a P5-200 with
a 300W power supply and is having a problem. His system may be using
more power overall, but this does point away from a power supply that is
>too small (also see later regarding crow-barring).
>
>Have you tried using something like the Future Plus PCI bus analyzer in here?
>You should be able to watch ALL the POST (power on self test) cycles across
>the
>PCI bus and determine why its failing. Remember, the ISA bus is 'beyond' the
>PCI bus, so all ISA cycles (including the beeper) pass across the PCI bus.
I am taking a shot at fixing it over the phone. If I can't figure what
it is, I'll get the customer's system in and take a look at the signals
on the bus.
>----------
>----------
>From: Krantz, Joe[SMTP:JKRANTZ@COSMPDGATE.CO.SYMBIOS.COM]
>Sent: Tuesday, April 15, 1997 3:17 PM
>To: Jeff Dahlin
>Subject: RE: Present Signals
>
>
>Try using a different model/brand video board (assuming it is a PCI
>video card). I had a similar problem (except everything worked but
>video, and without video it was hard to know anything was working).
>If you have other PCI cards (network etc.) try removing them as well.
>
> Joe
Unfortunately, the four cards that we are providing in the system are
video cards. There is a fifth card in the system that doesn't make a
difference. With 3 boards it works with or without it. With 4 boards
it doesn't work with or without it.
>----------
>From: sbus!sparc5!root@UCSD.EDU[SMTP:sbus!sparc5!root@UCSD.EDU]
>Sent: Tuesday, April 15, 1997 3:09 PM
>To: Jeff Dahlin
>Subject: Re: Present Signals
>
>
>You might listen to see if the crowbar circuit on the power supply
>tripped - maybe you ARE drawing too much power. The fan is probably
>a 12v fan, but the 5v may be crowbarred off as well as disabled by
>the "intelligent" BIOS. Perhaps it's just a wimpy power supply...
>
> Gary
He checked the system voltage levels and they are boarder-line (4.8 V
with 3 boards and 4.75 with 4 boards), but the power supply didn't
crow-bar.
So where I am at is there is a possibility that the 4.75V from the power
supply is causing problem, but I expected (like Gary) that it would
crow-bar if too much current was drawn. The customer says that the
SuperMicro has 8 PCI slots and there is an option for providing more 5V
power to it. They are looking into doing that. I am also waiting to
hear back from SuperMicro in case they have a suggestion. Unless
someone has info about AMI doing something unusual, or unless it ends up
being a power supply problem, I think that I will end up having the
customer ship the system to me to look at in-house.
Thanks to everyone for input.
Jeff Dahlin
Appian Graphics
jdahlin@appiantech.com
www.appiantech.com
, ø ç