Motherboard Forums


Reply
Thread Tools Display Modes

Auxiliary/TMPIN2 temperature too high?

 
 
danielffford@gmail.com
Guest
Posts: n/a
 
      01-10-2013, 01:19 PM
I have a Tyan S5101 Trinity i875P motherboard (yeah, I know it's old, but it's been working well till recently!). I run Windows 2000 Pro SP4, though I don't think that's very relevant to my problem.

In recent weeks it has started rebooting at random times, especially in hotweather (it's summer now in Australia). The CPU and chassis fans are all running. I've cleaned and re-applied heatsink compound to the CPU heatsink, which brought its temperature down a few degrees, but the reboots still occur.

SiSoftware's Sandra Lite reports 'Auxiliary temperature' as about 72C (CPUID's Hardware Monitor calls it TMPIN2). All the other temps are below about62C, including the CPU.

My question is, what is this Auxiliary/TMPIN2 temperature? Could it cause reboots at the >70C level? How can I lower this temp?

Any help would be much appreciated, as I'm not ready to retire this beast just yet. :-)
 
Reply With Quote
 
 
 
 
Paul
Guest
Posts: n/a
 
      01-10-2013, 06:36 PM
(E-Mail Removed) wrote:
> I have a Tyan S5101 Trinity i875P motherboard (yeah, I know it's old, but it's been working well till recently!). I run Windows 2000 Pro SP4, though I don't think that's very relevant to my problem.
>
> In recent weeks it has started rebooting at random times, especially in hot weather (it's summer now in Australia). The CPU and chassis fans are all running. I've cleaned and re-applied heatsink compound to the CPU heatsink, which brought its temperature down a few degrees, but the reboots still occur.
>
> SiSoftware's Sandra Lite reports 'Auxiliary temperature' as about 72C (CPUID's Hardware Monitor calls it TMPIN2). All the other temps are below about 62C, including the CPU.
>
> My question is, what is this Auxiliary/TMPIN2 temperature? Could it cause reboots at the >70C level? How can I lower this temp?
>
> Any help would be much appreciated, as I'm not ready to retire this beast just yet. :-)


This is from the user manual.

http://www.tyan.com/manuals/m_s5101_101.pdf

Note: The onboard Winbond 83627HF hardware monitoring ASIC automatically
detects the system, motherboard and CPU temperature. It detects the
CPU and chassis fan speeds in RPM. The hardware monitor ASIC also
detects the voltage output through the voltage regulators.

That implies three sensors. One could be CPU diode. I can't really be sure
how to interpret the others. On some motherboards, there can be a thermistor
placed to detect case air temperature. The chips (875P/ICH5 family), I
don't see a thermal diode listed for those.

On my 875P motherboard by another manufacturer, only two of the three
sensor channels are used. The third is placed on an optional two pin
header, so you can connect your own sensor. It suggests that *something*
on the motherboard can have a "free" sensor, which they can connect up.
I just haven't figured out yet, what that free one could be. So when
they say "system, motherboard", I suspect one of them is part of some
chip, and the other one involves paying for a thermistor or transistor
(whichever is cheaper).

*******

Also, rather than concentrate on the temperature reading,
I'd be inspecting the tops of the capacitors for leakage.
There are a total of around eleven capacitors in the upper
right hand corner of this photo. They're part of VCore
for the CPU. If the stamped "K" on the top breaks open,
and fluid leaks out, the VCore voltage can become unstable,
and random reboots or crashes result. (Of the eleven, four
could be in the +12V input side, and the other seven run at
VCore voltage. They put seven in parallel, to withstand
VCore ripple current.)

ftp://ftp.tyan.com/img_mobo/i_s5101.tif

There are other caps on the motherboard, so you should
really inspect them all. Sometimes, its a cap near
the video card slot that is oozing, and then the
video card seems to be unstable.

While it would be fun to trace down the sensor, there
are other things you can be looking at, for evidence
it's in need of repair. Boards can be "re-capped", if
you spot leaking or bulging caps. It's just time consuming
work with a soldering iron and vacuum desoldering station.
(Caps can be very hard to get out, even with good equipment.
It all depends on "hole diameter" versus "lead diameter",
as to how hard it is. My first employer used generous
hole sizes, and the caps would almost fall out. Not many
companies will design them to work that easily.)

*******

One other thing. Your Northbridge uses the "four hook,
one spring wire" retention scheme. The problem with that
is, the hooks soldered into the motherboard, are typically
made of the wrong materials. They use something that looks
a lot like stainless steel, which does not solder well.
It should have been a fat copper wire with tin/lead on the outside.

The solder joint on the bottom of the board, which holds
the four hooks, is a "cold joint", and then, the hook easily
pulls out of the motherboard. When that happens, the
Northbridge heatsink can fall off. Inspect the blue colored
heatsink with the word "Tyan" printed on it. It won't be
completely motionless. Even if the spring is still secure,
you may be able to rotate the heatsink a few degrees. If they
used thermal tape between heatsink and chip, it might remain
stuck in place. In any case, if a Northbridge overheats,
that would be the reason - a hook pulls out.

On some Dell, the hooks are actually checked by a motherboard
circuit. They run a continuity check through the hooks. So a
small amount of current flows through the spring wire that
holds the heatsink. If a hook lets go, the circuit is broken
and... the motherboard won't start. The user is expected to
figure that out for themselves, that the hook has let go.
Now, the interesting thing about that, is the design engineer
*knows* those hooks are made of the wrong material, and put
a sensing circuit in place, waiting for the inevitable. And on
hearing about the approach, all I could think is "how stupid is
it, to know the mechanical details aren't secure, and put in
an electrical feature to check for it". I'd much sooner use another
kind of retainer. Push-pins aren't very nice, but they're
another solution to mechanically securing the thing.

Fortunately, the ICH5/ICH5R runs cool enough, it doesn't
need a heatsink. Your ICH5 family Southbridge never had
a heatsink, and should never get more than warm to the touch.
That's the square just above the two blue IDE connectors
in the photo.

Paul
 
Reply With Quote
 
 
 
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
WiMAX in 2010: Too little, too late? How fast, how far, how much? Obaid Embedded 0 10-19-2009 10:42 AM
Intel's new offerings, too little, too late michaelanderson4@sympatico.ca Intel 6 01-25-2006 12:10 AM
what too use too make a backup of 2 sata raid disks??????????????? mike6 Asus 1 03-03-2004 08:05 PM
NF7-S rev. 2.0 RT1 Temperature Thermistor too high?? user Abit 3 01-08-2004 12:33 AM
Too high (and too LOW) thermal resistances??? Da Loaf AMD Overclocking 0 08-05-2003 04:23 AM


All times are GMT. The time now is 07:07 AM.


Welcome!
Welcome to Motherboard Point
 

Advertisment