The PLE133T board must be reading from the on-die diode, it has no
other sensor. IIRC, an earlier version of the board did but not the
later, perhaps last version, which I have.
sometimes they use just a small SMD element inside socket & you can
not tell/find easily/see which one is a resistor, a NTC probe or
capacitor & sometimes there is no marks on them or are so tiny, that
you need a powerful magnifying glass to see the difference (this is
case with my MoBo temp sensor also!)
I'm not sure I can agree that an on-die sensor would typically read
47C temps... even the 1.2GHz-on-694X Tualatin here, produces very
little heat.
for that one in mine case would be few °C less, but also my cooler @
idle does not get hot, barelly warm .. also mine does not produce much
heat @ idle ...
It wouldn't even be running in the 30s (celsius) if it
had a case with better ventilation or if the fan were producing more
normal airflow rate... the constraints of a poor case and desire for
low noise were exactly why I used a Tualatin at the time. It was a
similar situation with the other Tualatin @ 1.47GHz, I had it in
slightly modified IBM Netvista case with horrendous airflow and
heatsink height limitations.
well my old CPU sensor fixed now inside socket says now 38° & the
temperature of the bottom of a HS is 36°C & these 2 temps if fans are
running show greater differences with diminishing/rising room temps
than the one from the core /I made an experiment varying a week ago
after fixing properly my gadget inside case the temps - leaving the PC
in a cold room @ 10°C & later heating it up with a hair dryer in a
catoon box to 35°C -simulating hot summer for more than an hour/, so
the "amb" temps were -10 to +15 °C than a normal 20° my room one. I
tried with an idle & full load & generally so with TL 25°C of
difference of Amb temp, the core temp difference was less (-5max to
+10max, TL max of 15°C difference), while the other sensors followed
amb temps changes quite well ... interesting findings came out (have
to update soon my website). The same behavior shows also Coppermine
Celly600@900 but with much greater Idle/full load difference &
naturally lower temps in generally, since its HLT works properly with
Rain even if the gadget calibrated & made for Tuallie should show for
the Cu-Mine minor differences in temps; the Tuallie shows from onDie
only few °C of difference there in my setup....
There is no doubt, that my temps are false, since I am quite capable
in electronics & made so many experiments because of my curiosity & I
was also surprised by mine own findings. I checked, rechecked &
recalibrated my sensors so many times with different CPUs, that I am
really sick of experimenting with temps ... ;-)
I don't know about Miracle, perhaps just proper BIOS support.
Could be!
The
Intel Celeron Datasheet for the 1.2GHz Tualatin does make same mention
of HLT states as did the prior Coppermine Celeron 'sheets.
I know that, so I was disapointed of seeing higher temps than should
be & I first thought that this is happening because of the mod. :-(
but it looks it is not the case ...
From
what I vaguely recall, my sister's Dell Inspiron laptop has what's
essentially a Tualatin mPGA CPU, which I suspect also HLT-idles, on a
??? brand 815 chipset mainboard.
Yes, but mobile CPU have a bit different core schematics than a
desktop ones IMHO, not just the package & better power management!
Could be that Intel intentually cripled desktop Tualatins HLT to
prevent using them on cheap Laptops by some manufacturers, who knows,
Intel does not always tell the truth, we know that ... & replaced them
with a crap of p4 based Cellys ...
Some AMDs do though. I've had a couple Gigabyte boards that did, were
Via KT333 chipset (forget model name). Since most of the Athlon
boards can HLT idle the CPU after a couple of chipset registers are
changed, it would be simple enough for the motherboard manufacturer to
just set those registers automatically or with a BIOS option, but
since it reduces performance and possibly PCI issues, most disabled
it.
I know that ... but some SoftCoolers work like a charm with them &
IMHO some problems using them with WinXP were already fixed till now..
..... & I would like also to know some hardware register tweak like
those for AMDs for my Tuallie to cool it down better ... but it looks
no joy till now :-(
Another interesting experiment I made time ago with a program
Hmonitor: you can set it to full throttle the Cpu thru Bios directly
starting even with lower temps than ambient ones: oh, that cools than
properly my Tuallie really well like HLT shoud, but ... the machine
becomes sluggish like a turtle
))))) ... a feel of 90´ ....
))
-- Regards, MERRY HOLIDAYS & HAPPY NEW YEAR, SPAJKY ®
& visit my site @
http://www.spajky.vze.com
Celly-III OC-ed,"Tualatin on BX-Slot1-MoBo!"
E-mail AntiSpam: remove ##
