Archived from groups: alt.comp.periphs.mainboard.asus (
More info?)
On Sun, 15 Aug 2004 05:57:28 -0400, nospam@needed.com (Paul) wrote:
>In article <hnfth05c0cvdhhnb675ij6336gno6g4uiv@4ax.com>, JPeterman
><biteme@thelibrary> wrote:
>
>> On Sat, 14 Aug 2004 19:55:28 -0400, nospam@needed.com (Paul) wrote:
>>
>> >In article <rs2sh0h9kquu63o29a0ajhrno0dbob4plu@4ax.com>, awtwtw wrote:
>> >
>>
>> >
>> >As Dave said, the power supply is always a likely source of
>> >failure to boot.
>> >
>> >But, one of your other symptoms bothers me a bit. The red LED
>> >on the motherboard is the "AGP_Warn" circuit. It is supposed
>> >to check for an illegal 3.3V only video card, and prevent the
>> >computer from starting. If an illegal (improperly keyed) video
>> >card is present, the PS_ON# signal is prevented from reaching
>> >the power supply, by AGP_Warn.
>> >
>> >The AGP_Warn has two ingredients. One I have traced down, so
>> >I know it is there. The TYPEDET# signal from pin A2 on the
>> >video card is connected to one of the transistors. Apparently,
>> >if TYPEDET# is grounded, everything is OK. If TYPEDET# floats,
>> >then that will cause the red LED to light up.
>> >
>> >A second ingredient of the check, is the circuit can only
>> >stop the power, if an AGP card is inserted. It would be
>> >inappropriate for the circuit to light the LED, if no video
>> >card is present. So, to do its job, the logic has to be
>> >AGP_Present AND TYPEDET#_floating. I don't know what pin is
>> >checked to verify the card is present.
>> >
>> >Now, your symptom of the LED flashing for a moment is weird.
>> >The inputs to AGP_Warn are supposed to be static and unchanging.
>> >Yet, your AGP_Warn lights for a moment and then stops, and
>> >I cannot think of a mechanism for that to happen, unless +5VSB
>> >is going out. And, you said the green LED on the motherboard
>> >stays lit and the fan runs on the back of the case. That means
>> >+5VSB is present and +12V is there to run the fan.
>> >
>> >If you have another video card compatible with the board,
>> >try swapping that in. It could be the video card, but as I
>> >say, I cannot really think of a failure mechanism that would
>> >cause the LED to light (indicating trouble) and then
>> >extinguish, unless all power was being lost at the same time.
>> >
>> > Paul
>>
>> Firstly let me thank you for your detailed response. Most of this was
>> over my head but I understand the gist of it.
>>
>> The power button on the front of the case will not turn the power off.
>> I have to hit the switch on the back.
>> I just tried it again and the red LED on the motherboard will not
>> flash at all.
>>
>> I forgot that there is a voice "post reporter" at boot that may tell
>> me something. I didn't have the speakers on when this happened so I'm
>> going to try it.
>>
>> So the power supply may be suspect even though it powers the other two
>> fans?
>> I don't have another video card but I have been leaning towards
>> getting another one.
>> Thanks again for your patience and time.
>
>Your observation that the switch on the back is needed for control
>changes things a bit.
>
>The PSU has several output voltages. Perhaps only one of them is
>not working, and the others are. If you own a voltmeter, you can
>take a disk drive power connector in hand, and check for +5V and
>+12V. If the PSU has a six pin Aux power connector, there will
>be a +3.3V pin on it. Those are the major voltages, and I don't
>know if the motherboard depends on -5V and -12V in any meaningful
>way or not. On my P4C800-E for example, -12V is used by
>the serial port, so if -12V dies, my serial port will die, but
>the rest of the board will still run. It could be your PSU has
>lost +3.3V or +5V.
>
>When using your voltmeter, connect the black lead to a lug on
>one of the connectors on the back of the computer. I'm assuming
>you have some alligator clips or another clever way to connect it.
>The reason is, you don't want to accidently touch the two voltmeter
>leads together while working on the power supply. I find with one
>red lead in my hand, things are much safer. It is like safety rules
>for jumpering car batteries - there is a lot of available current
>there if you make a mistake.
>
>I think as Dave said, I'd try a PSU swap first. If the result is
>that the red LED stays on, borrow another video card, to see
>if it will go off again. If the red LED stays on even with
>another video card, it is time to RMA the motherboard. The
>AGP_Warn circuit has failed on some motherboards out of the box,
>but I've never heard of one dying on a system that was working
>for a while.
>
>Your AGP_Warn is just another mystery for the books
🙂
>I haven't traced the whole circuit, and as those SMT components
>are so hard to identify, I probably never will finish it.
>
>HTH,
> Paul
Since I don't have a voltmeter or another video card I think I''m
going to buy another power supply. The one I have is a Duro
400 watt supply so I probably won't buy one of these. If that fails
to correct the situation I'' buy a new video card.
Thanks for all of your help. I've got a lot more insight for the
future.
Facebook
Twitter
Instagram