P2B-S and S370-133 Vcore changed on own?

  • Thread starter Thread starter difficult
  • Start date Start date
D

difficult

I've been running a PIII800 in an Asus S370-188 at 1.8 Vcore in a
P2B-S motherboard and Nexus PS for several years. Both the bios and
MBM5 always reported 1.79 Vcore. Recently the system has halted on
boot with a hardware warning that indicates the Vcore has risen to 2.0
in the bios and 2.02 in MBM5. I've pulled the slocket and cleaned the
jumpers. Still 2.0.
I even tried reducing the slocket jumpers from 1.8v to 1.6v but the
system wouldn't boot. The system seems to run fine but I am worried
that 2.0 is near the limit. Strangely the CPU temperatures reported
have not gone up at all. (30deg C at idle). Does anyone have any ideas
as to why this is happening or perhaps some insight as to how to
troubleshoot this with a Voltmeter? TIA

Mark
 
I've been running a PIII800 in an Asus S370-188 at 1.8 Vcore in a
P2B-S motherboard and Nexus PS for several years. Both the bios and
MBM5 always reported 1.79 Vcore. Recently the system has halted on
boot with a hardware warning that indicates the Vcore has risen to 2.0
in the bios and 2.02 in MBM5. I've pulled the slocket and cleaned the
jumpers. Still 2.0.
I even tried reducing the slocket jumpers from 1.8v to 1.6v but the
system wouldn't boot. The system seems to run fine but I am worried
that 2.0 is near the limit. Strangely the CPU temperatures reported
have not gone up at all. (30deg C at idle). Does anyone have any ideas
as to why this is happening or perhaps some insight as to how to
troubleshoot this with a Voltmeter? TIA

Mark
Click to expand...

VID4 VID3 VID2 VID1 VID0

0 0 1 0 1 1.80
0 0 0 0 1 2.00

It would seem VID2 is being grounded somehow.

The copy of the S370-133.pdf manual I've got (URL no longer valid)
shows a 3x5 array of header pins. The center five pins would be
the VID signals, as sent to the motherboard. Putting jumpers between
the left column of pins and the center column, connects the processor
output pins to the VID signals. The column of pins on the right,
are ground signals. Now, the third row of pins down, should have
no jumper (as that is VID2). This leaves the VID2 pin floating.
There should be a pullup on the regulator, to define a logic 1 when
no jumper is installed on a particular row.

The pinout for a PIII module can be seen on page 75:

ftp://download.intel.com/design/PentiumIII/datashts/24445209.pdf

It is easier to see visually, on page 6 of the SC242 doc:

http://www.intel.com/design/PentiumIII/designgd/24518103.pdf

Pin A119 VID2 is next to Pin A118 GND, so perhaps there is a short
between A119 and A118 on the module. Or, maybe there is an alignment
problem, causing A118 and A119 to touch at the socket.

HTH,
Paul
 
I've been running a PIII800 in an Asus S370-188 at 1.8 Vcore in a
P2B-S motherboard and Nexus PS for several years. Both the bios and
MBM5 always reported 1.79 Vcore. Recently the system has halted on
boot with a hardware warning that indicates the Vcore has risen to 2.0
in the bios and 2.02 in MBM5. I've pulled the slocket and cleaned the
jumpers. Still 2.0.
I even tried reducing the slocket jumpers from 1.8v to 1.6v but the
system wouldn't boot. The system seems to run fine but I am worried
that 2.0 is near the limit. Strangely the CPU temperatures reported
have not gone up at all. (30deg C at idle). Does anyone have any ideas
as to why this is happening or perhaps some insight as to how to
troubleshoot this with a Voltmeter? TIA

Mark
Click to expand...

Now that's a bizarre one!

The voltage request from the CPU to the onboard regulator uses 5 signals
called VID[0-4]. Sounds like your VID2 signal has changed state due to a
failure on either the slocket or motherboard:

VID[0-4] 1.8v = 1 0 1 0 0
VID[0-4] 2.0v = 1 0 0 0 0

The regulator reads a logic 0 when the corresponding pin is connected to
ground on the CPU, and a logic 1 when the pin is left disconnected on
the CPU and a pullup resistor on the mainboard pulls it up to at least 2.0v

I suggest first checking to see if the CPU is really being supplied with
2.0v. The easiest way to check that is to connect the negative lead of
your voltmeter to ground (black wire) on a molex connector, and the
positive lead to pin AA37 on the back of the slocket (viewed from the
front - i.e. memory is closer than CPU - AA37 is the 7th pin on the top
row, counting from the right).

If your meter reads 2.0v Vcore on pin AA37, the problem is confirmed and
you now want to find out what the voltage regulator sees on VID2. The
voltage regulator is a 20-pin chip adjacent to the CPU fan connector,
labelled HIP6004xxx, and VID2 is pin 6 (6th pin from the left on the
side closest to the CPU slot).

I'm guessing your VID2 pullup resistor may have gone open circuit, but
the results of the above tests should tell the story.

P2B
 
On Tue, 03 Aug 2004 19:00:25 -0700, (e-mail address removed) wrote:

Thanks guys. I will get it on "the bench" sometime this week and
report anything interesting.

I very much appreciate the time you take to help us all out..
 
Back
Top