[lm-sensors] ECS 945GCT-D (Fintek F71862FG SuperIO)

[lm-sensors] ECS 945GCT-D (Fintek F71862FG SuperIO)

[mskala]

I have an Elitegroup (ECS) model 945GCT-D motherboard.  This is an Intel
Atom 230 fanless mini-DTX board with a Fintek F71862FG SuperIO chip; it
seems to be similar to the Jetway NC92-330-LF, which uses the same SuperIO
chip, and using the configuration for that board from the Wiki, I get this
output:

f71862fg-isa-0a00
Adapter: ISA adapter
Vcc3V:       +3.22 V
Vcore:       +1.21 V
   NB:       +1.91 V
+ 5V:        +0.00 V
+12V:       +10.74 V
5VSB:        +3.91 V
VDIMM:       +1.04 V
VSB3V:       +3.42 V
Vbat:        +3.33 V
CPUFAN:        0 RPM  ALARM
SYSFAN1:       0 RPM  ALARM
SYSFAN2:       0 RPM  ALARM
?CPU Temp:   +52.0C  (high = +85.0C, hyst = +81.0C)
                      (crit = +255.0C, hyst = +251.0C)  sensor = transistor
?Sys Temp:     FAULT  (high = +85.0C, hyst = +81.0C)
                      (crit = +100.0C, hyst = +96.0C)  sensor = transistor
???? Temp:  +127.0C  (high = +70.0C, hyst = +68.0C)  ALARM
                      (crit = +85.0C, hyst = +83.0C)  ALARM  sensor = transistor

The BIOS only reports three numbers:  VCore 1.210 V, VDIMM 1.912 V, and
CPU Temp 53.0C.  From this I conclude that the VCore and CPU temperature
numbers are being reported correctly by sensors, and the voltage labelled
"NB" is actually VDIMM.  The other two temperatures seem to be garbage,
and they don't change with system load, so I think the board probably does
not have those channels of the SuperIO chip hooked up to anything useful.

The board has three-pin headers for two fans (called "CPU" and "PWR") but
I have nothing hooked up to those.

So:  it's not difficult to create a configuration file stanza based on the
one from the NC92-330-LF with the non-useful parts set to ignore.  I
wonder about those other voltages, though.  Most of them seem to be
measuring something, but 10.74 sounds low for the +12V supply, so I wonder
if the formula for that is inapplicable to this board.  But then again, I
am running this from an 84W power supply which maybe doesn't have enough
oomph to keep that bus at the full 12V?  The fact that the BIOS doesn't
report any other voltages than VCore and VDIMM is a little odd, and the
BIOS's VDIMM seems to be on a different channel from the one that was
called VDIMM on the Jetway board.  I don't know if any of those voltage
channels can be trusted.
-- 
Matthew Skala
mskala@ansuz.sooke.bc.ca                    Embrace and defend.
http://ansuz.sooke.bc.ca/

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

Re: [lm-sensors] ECS 945GCT-D (Fintek F71862FG SuperIO)

[Hans de Goede]

Hi,

On 07/04/2009 04:06 AM, mskala@ansuz.sooke.bc.ca wrote:
> I have an Elitegroup (ECS) model 945GCT-D motherboard.  This is an Intel
> Atom 230 fanless mini-DTX board with a Fintek F71862FG SuperIO chip; it
> seems to be similar to the Jetway NC92-330-LF, which uses the same SuperIO
> chip, and using the configuration for that board from the Wiki, I get this
> output:
>
> f71862fg-isa-0a00
> Adapter: ISA adapter
> Vcc3V:       +3.22 V
> Vcore:       +1.21 V
>     NB:       +1.91 V
> + 5V:        +0.00 V
> +12V:       +10.74 V
> 5VSB:        +3.91 V
> VDIMM:       +1.04 V
> VSB3V:       +3.42 V
> Vbat:        +3.33 V
> CPUFAN:        0 RPM  ALARM
> SYSFAN1:       0 RPM  ALARM
> SYSFAN2:       0 RPM  ALARM
> ?CPU Temp:   +52.0C  (high = +85.0C, hyst = +81.0C)
>                        (crit = +255.0C, hyst = +251.0C)  sensor = transistor
> ?Sys Temp:     FAULT  (high = +85.0C, hyst = +81.0C)
>                        (crit = +100.0C, hyst = +96.0C)  sensor = transistor
> ???? Temp:  +127.0C  (high = +70.0C, hyst = +68.0C)  ALARM
>                        (crit = +85.0C, hyst = +83.0C)  ALARM  sensor = transistor
>
> The BIOS only reports three numbers:  VCore 1.210 V, VDIMM 1.912 V, and
> CPU Temp 53.0C.  From this I conclude that the VCore and CPU temperature
> numbers are being reported correctly by sensors, and the voltage labelled
> "NB" is actually VDIMM.  The other two temperatures seem to be garbage,
> and they don't change with system load, so I think the board probably does
> not have those channels of the SuperIO chip hooked up to anything useful.
>
> The board has three-pin headers for two fans (called "CPU" and "PWR") but
> I have nothing hooked up to those.
>
> So:  it's not difficult to create a configuration file stanza based on the
> one from the NC92-330-LF with the non-useful parts set to ignore.

I agree with all of the above :)

> I
> wonder about those other voltages, though.  Most of them seem to be
> measuring something, but 10.74 sounds low for the +12V supply, so I wonder
> if the formula for that is inapplicable to this board.  But then again, I
> am running this from an 84W power supply which maybe doesn't have enough
> oomph to keep that bus at the full 12V?

No, the 10.74 looks like it might actually be measuring 12V because of the
scaling done in sensors.conf. The best way to create a configfile for a new
motherboard is to first make sensors.conf completely empty, then you will see
the actual voltages as measured by the chip (which might be scaled down from
the actual voltages by external resistors.

> The fact that the BIOS doesn't
> report any other voltages than VCore and VDIMM is a little odd, and the
> BIOS's VDIMM seems to be on a different channel from the one that was
> called VDIMM on the Jetway board.  I don't know if any of those voltage
> channels can be trusted.

An other possible source of information for this is the windows software
if any. If that measures more boltages there are probably more voltages hooked
up if not, well then probably not. Or try asking Jetway.

Regards,

Hans

_______________________________________________
lm-sensors mailing list
lm-sensors@lm-sensors.org
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors