Re: Boot process not starting
- From: "Jan Alter" <bearpuf@xxxxxxxxxxx>
- Date: Sat, 15 Aug 2009 16:18:18 -0400
--
Jan Alter
bearpuf@xxxxxxxxxxx
"Lee Antony" <leemailx-leigh@xxxxxxxxxxxxxx> wrote in message
news:U4KdncehzaX3bxvXnZ2dnUVZ8iOdnZ2d@xxxxxxxxx
To Paul or anyone still following this thread:I attempted to go to Asus to get the system requirements by downloading your
"Paul" <nospam@xxxxxxxxxx> wrote in message
news:h668om$ons$1@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Lee Antony wrote:
"Paul" <nospam@xxxxxxxxxx> wrote in message
news:h653as$atl$1@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Lee Antony wrote:
I've built PCs before and they always worked but maybe I was just
lucky because all I'm trying to do now is update this XP Pro desktop
with a new motherboard, cpu and video adapter and I can't get it to
work.
When I push the power button all the fans come to life and spin for 14
seconds, then they stop for a few seconds and then start up again for
14 seconds. This repeats for as long as the power is on.
There are no beeps (chassis speaker tested and working) and the
monitor does not register a video signal (monitor tested and working)
so I don't know how to start troubleshooting this.
My first thought was overheating but I have checked the cpu cooler and
thermal paste - all seems good. Also, the power supply outputs seem
to have the the right numbers.
The new items that I'm trying to install are:
motherboard Asrock P45XE
cpu Intel Core 2 Duo E7500
video Asus Radeon EAH4850.
I know I've read something somewhere about this stop/start symptom but
danged if I can remember where I saw it. If you have any suggestions
at all please fire them at me because right now I'm stuck.
Regards Lee Antony.
Have you double checked there are no bent contacts in the LGA775 socket
?
Sometimes, a motherboard gets damaged in shipping, somewhere along the
line, and the contacts can be upset.
Is the video card in the green slot ? I'd try it in the green
slot first.
http://www.asrock.com/mb/overview.asp?Model=P45XE&s=
ftp://europe.asrock.com/manual/P45XE.pdf
You can do beep tests, as a means to determine whether the processor
is executing code. Connect the computer case speaker ("chassis speaker
header"
in the manual). With all power removed (switch off at the back of the
computer),
pull the RAM sticks out of the board. On the next POST attempt, the
BIOS, if it
is running, will beep the speaker, indicating it has detected bad RAM.
That
proves some of your system is running.
If there are no beeps at all, then check that the processor is
getting power. There is a 2x4 ATX12V power connector, and if a
supply only has a 2x2, you can still plug that in and use it.
According to the manual, it goes into the lower holes. Some
motherboard companies stick a label across the unused pins,
to make it easier to figure out where a 2x2 connector goes.
There is a picture on page 33 of the downloadable manual,
which shows two black wires on the left, and two yellow
wires on the right, of the ATX12V 2x2 connector. The staff
at Asrock cannot count, which is why the two figures have
the number "6" in the right hand corner, instead of "5".
HTH,
Paul
No I haven't looked at the socket contacts carefully because I used the
last of my Arctic Silver thermal paste for re-assembling after checking
the cooler fan pushpins but I will have to have another look at this.
Good idea to try and force a beep by pulling the RAM. I'll let you know
how it turns out. Do you think that the 14 secs it will run for is
enough time to trigger the warning?
Is there an easy way to check if the processor is getting power? The
PSU is definitely sending power to the motherboard connector and the
pre-applied thermal paste was part melted when I removed the cooler.
The ATX 12V output is on a 4-pin plug but I think I have it connected
right. The printed manual also shows pin 5 with the number 6!
Thanks for taking the time to help.
Regards Lee Antony.
It should look for RAM pretty soon after it starts. Some of the
BIOS code runs "in registers only", until the Northbridge is programmed
and the memory controller is set up. Once the RAM is running, then the
BIOS programming can have all the normal amenities. If the BIOS code
is really running, 14 seconds should be enough time for some beeping.
*******
I don't understand the significance of 14 seconds. The number doesn't
ring any bells.
Something you should know about recent Intel chipsets, is they have
a microcontroller inside them. The microcontroller is capable of
running firmware, and is a miniature world unto itself.
That feature is not fully enabled on all chipsets. It is fully
enabled on the Q series chipsets. They're used in large businesses,
and the purpose of the microcontroller, is to support remote
management. One thing you can do on a Q series chipset, is
reach through the Ethernet interface and cause the computer
to reboot. So if a computer is not responding for example,
because it crashed, the IT department could force a reboot
at 3AM and push out patches or the like. Servers have had this
capability for some time, as provided by a small controller
card and special socket on the motherboard. But Intel decided it
would be cool, for computers used in large businesses, with
big IT departments, to have remote control of desktops via
the network.
On regular desktop motherboards, with non-Q series chipsets, only
one third of that is turned on (the safe stuff). There is apparently
a fan controller in there for QST. The 14 second time constant could
be related to that fan controller, unless you notice that all power
is being cut and not just a fan speed modulation. If all the
lights stay in the same state, but the fans vary, that could be
related to QST.
This is the ICH10 datasheet. On PDF page 240 I found
http://www.intel.com/assets/pdf/datasheet/319973.pdf
"The internal PWM control register is programmed to a non-zero
value by the Intel QST firmware."
which means it could be under control of that microcontroller.
So perhaps the 14 second time constant has something to do
with how the QST is controlled. PWM is normally used for
just the CPU fan. The motherboard would need a circuit to
convert from PWM to voltage, to control any other three pin
fans.
At one time, if you had a single stick of RAM in the system,
it had to go in a particular slot, to please the microcontroller.
The microcontroller had a small RAM allocation, from system DRAM.
It'll be interesting now, when you do your "RAM removal" test,
to see how the fan responds. If there is a microcontroller, it
won't have any place to store stuff :-) So be prepared for
different fan symptoms. Maybe the 14 second time constant will
change, when the RAM is missing.
The only details I've seen for the rest of what that microcontroller
can do, was presented in a Powerpoint slide set. I haven't seen
any detail in a datasheet. I've also not seen any evidence,
in the way of BIOS release notes on a motherboard manufacturer site,
that any part of the BIOS is reserved for that microcontroller.
I've never seen a note acknowledging that something is needed to
control QST. So I can never be sure that the fan controller,
is controlled by the microcontroller, and not by the main
processor and the regular BIOS code.
Anyway, on with your testing... :-)
Paul
I've looked at the cpu socket pins with a magnifying glass and nothing
seems amiss.
I pulled the RAM sticks as suggested and nothing changed at power on. The
missing memory produced no warning beeps. The on/off time stayed the
same. (I've referred to this as 14 seconds but it would be more accurate
to say it is 14.0 +/- 0.2 seconds by my stopwatch.)
I can't be sure if all power is being cut but the power led certainly goes
on and off. The board doesn't seem to have any leds on it so no clues
there.
My worry is that I'm overlooking something silly and obvious otherwise I
would be RMAing the motherboard. Other than taking it to a professional I
can't think what else to do.
Thanks for taking the time to help.
Regards Lee Antony.
graphic's card's manual. Page wouldn't load as Asus says the site is
overloaded. However, have you checked the power requirements for this card?
I would bet you would need at least a 400 - 500 watt PS, and even that might
not be enough if the PS is not decent quality. If underpowered that would
explain the cycling.
--
Jan Alter
bearpuf@xxxxxxxxxxx
.
- Follow-Ups:
- Re: Boot process not starting
- From: Lee Antony
- Re: Boot process not starting
- From: Paul
- Re: Boot process not starting
- References:
- Boot process not starting
- From: Lee Antony
- Re: Boot process not starting
- From: Paul
- Re: Boot process not starting
- From: Lee Antony
- Re: Boot process not starting
- From: Paul
- Re: Boot process not starting
- From: Lee Antony
- Boot process not starting
- Prev by Date: Re: Mobo PS fault problem
- Next by Date: Re: Boot process not starting
- Previous by thread: Re: Boot process not starting
- Next by thread: Re: Boot process not starting
- Index(es):
Relevant Pages
|
