ICH5R chipset burn on a P4P800-E Deluxe

  • Thread starter Thread starter Ian
  • Start date Start date
I

Ian

First: Excuse me for my bad English

Second: The problem

Here's the spec of my homebuild pc

Case Tsunami VA3400BWA (with 400 watt power supply)
Asus P4P800-E Deluxe motherboard
Pentium IV 3.2C
2 X 512 Mo Ram DDR400 in dual channel config (Kingston)
HDD: Western Digital Raptor (WD740GD)
PowerColor ATI Radeon 9800Pro 128 Mo
DVD Burner 16X (LG)
Floppy
Windows XP Home Edition

I've done all the setup yesterday. Everything's going fine.

After the first boot I gone in the BIOS to check if everything detected
corretly and all seems to be OK

I've create 3 partitions on the drive and format the first in NTFS. After
that I've install windows without any problems. So the computer work well
for approximatively 1 hour (maybe a little less)

Then the system boot up in windows and the little presentation of windows XP
begin. Just after that (maybe 20 seconds), I smell that something is burning
in the computer. I unplug it immediately and check in the computer and Oh
no!!! The ICH5R chipset is burned (melted). I re-check in the book to see if
I have done something wrong but I could't find nothing.

The mobo is really dead. I've tried to put power back, but when I plug the
power without pressing the power button, the ICH5R chipset become to be hot
again. I've also tried to press the power button, but nothing occurs.

I've bring back my computer to the store to see with them if they view
something wrong with my setup. They tell me they don't see nothing and that
they will replace the motherboard and check the HDD.

My question is: Can some bad connection in the computer can do that and if
yes, which one (USB, Firewire) ? Can that be the power supply and if yes,
how can I see that ?

I will try to rebuilt my pc myself. But before I do that, I want to know if
this
trouble have something to do with my installation, incompatible
hardware or a default in the motherboard.

Thanks for your precious help.

Ian
 
JK says there must have been a AMD64 in the near proximity, the
chipset burned out of jealousy. ;^)

possible mainboard solder point coming into contact with the case, or
an overly exuberant power supply could cause it.
 
I'll check with my neighbors for the AMD64 ;-)

Seriously, I think that what you say is possible but I really don't know
where the contact came from.

For my next try, do you have some advices to give to me (beside check if
there's some contact with the board). If I go to the BIOS and check the
voltage, is this a good check or not ?

Thanks.
 
Thanks for your answer.

Do you think some other component have suffered this problem (I hope not) ?

I'll give you the result of my next try (I really hope everything's going
right).

Thanks.
 
Hi All,

I've re-check and re-re-check what I could have done wrong with my
installation since I still waiting for my new mobo.

The only thing I could see is maybe a bad connection of the front audio
connectors and since the problems begin just after I tried to connect a
headphone set to my computer, I have some doubts that the problems come from
there.

My questions is: Is it possible that a bad connection of the front audio
connector (or the front IEEE1394 connector) on the motherboard create a
short-circuit that can fried the motherboard.

Thanks again for your answer.
 
oh yes and the 1394 connectors are varied in there pinouts compared
to some
peripherals pinouts. I had a system where if you plugged the speakers
in after a boot and the system is running, it would shutdown or
reboot.
 
Hi all,

I've bought a new board (Foxconn A865A01-PE-6EKRS) and do all the setup
again.

Now Everything seems to work. I plugged all the component, install windows
and now it work for about an hour.

Next, I Install all the drivers and the monitoring software.

Now I check the voltage and I see a problem. The -12V rail is not monitoring
in the BIOS so I did't see the problem earlier. But now, with the monitoring
software, I can see that the -12V give -14.10 to -14.30 value.

I check the value with 2 software (the one come with mobo and Motherboard
Monitor 5) and they give me the same value. The -5V (not monitored in the
first software) give a -6V value in Motherboard Monitor 5.

I don't have a voltmeter to check the voltage (maybe tomorrow I will have
one) directly from the ATX connector but is this enough to be the source of
my first problem (the ICH5R that fried on my first mobo).

Second question: To check the voltage with a voltmeter, do I just have to
plug the "-" on a ground pin of the ATX connector and the "+" to the -12V
connector ? Do I have to disconnect all the component in the computer ?

Thanks for your answers.
 
I've read in other post that since the -12V rail does't be use a lot,
sometime the powersupply have some difficulty to "regulate" the voltage. So
I've done the test with and without a serial mouse connect (since according
to my reading, the serial port are the only component that use the -12V
rail)

I've test with a voltmeter
-13.28 with mouse
-13.59 without mouse

I've also test the -5V with the voltmeter and it give my -5.2 to -5.3. So I
think it's a little high but more acceptable than the -6V I report earlier

Is this acceptable or should I bring back my powersupply to the store ?

Thanks.
 
Ian said:
I've read in other post that since the -12V rail does't be use a lot,
sometime the powersupply have some difficulty to "regulate" the voltage. So
I've done the test with and without a serial mouse connect (since according
to my reading, the serial port are the only component that use the -12V
rail)

I've test with a voltmeter
-13.28 with mouse
-13.59 without mouse

I've also test the -5V with the voltmeter and it give my -5.2 to -5.3. So I
think it's a little high but more acceptable than the -6V I report earlier

Is this acceptable or should I bring back my powersupply to the store ?

Thanks.

It doesn't matter.

Nothing at all uses -5V and, even if it did, 5.2 is within spec (ATX spec
+-10% for 4.5V to 5.5V).

The ATX spec for -12V is +- 10%, which would make 13.2 the lower limit so,
technically it's '.67%' (probably less than the tolerance on your meter)
over spec but -12 volt is only used by serial devices and the '-V' for
serial can be anything from -3 to -25 for the rs-232 standard. Point is, it
doesn't matter.

What matters are the +5V, +3.3V, and +12V rails (all +-5%)
 
Back
Top