Hallo,
I noticed that sensors reports some strange numbers regarding my PSU:
sensors
k8temp-pci-00c3
Adapter: PCI adapter
Core0 Temp: +52.0°C
w83627thf-isa-0290
Adapter: ISA adapter
VCore: +1.41 V (min = +0.70 V, max = +1.87 V)
+12V: +12.52 V (min = +7.24 V, max = +11.49 V) ALARM
+3.3V: +3.17 V (min = +0.66 V, max = +3.15 V) ALARM
+5V: +5.01 V (min = +0.64 V, max = +4.35 V) ALARM
-12V: -12.36 V (min = -11.04 V, max = -14.58 V) ALARM
V5SB: +5.05 V (min = +2.37 V, max = +0.99 V) ALARM
VBat: +2.69 V (min = +2.32 V, max = +1.94 V) ALARM
fan1: 0 RPM (min = 784 RPM, div = 8) ALARM
CPU Fan: 1418 RPM (min = 860 RPM, div = 8)
fan3: 0 RPM (min = 5672 RPM, div = 2) ALARM
M/B Temp: -15.0°C (high = +46.0°C, hyst = +2.0°C) sensor = thermistor
CPU Temp: +46.5°C (high = +40.0°C, hyst = +39.0°C) ALARM sensor = diode
temp3: +42.0°C (high = +80.0°C, hyst = +75.0°C) sensor = thermistor
I have not change any sensors configuration for 3 years now (from the day that I makde my PC), and I have not seen those numbers before.
So, why are the maximum levels so low?
Why is the VBat and V5SB readings so messed up?
Can I do something to correct it? If not, is it a problem of the PSU?
V5SB is 5 volt stand-by. a +5.05v reading seems ok. Why is it messed up?
Also, doesn’t VBat refer to your CMOS battery? If so, it’s reasonable, as this is usually a 3 voltCR2032 flat round battery that powers your “bios” memory.
On 06/18/2011 07:06 AM, tpe wrote:
>
> Why is the VBat and V5SB readings so messed up?
to me it looks like your config is scrambled…
now, how or why that happened i have no idea, but ask:
have you checked your drive health lately?
i think if it were my machine i’d boot from a live CD and run fsck
against each partition…without writing to the partition (check ‘man
fsck’ closely, i think you want the -N switch)…
if you find lots of errors DO NOT try to boot it up or mount the drive
read/write…instead, use the CD to mount READ ONLY and get your data to
a safe place…and then try to fix it
of course, if you already have a good pre-problem backup . . .
On 06/18/2011 09:14 AM, DenverD wrote:
> to me it looks like your config is scrambled…
WAIT! i don’t know where the minimum and maximum values come from, do you?
now that i think about it, those are probably embedded in the PSU
circuitry…
so, maybe you are seeing NOT hard drive scramble, but rather PSU memory
scramble/read out problems/or what??
as @brunomcl mentioned the actual values delivered from the PSU appear
ok…it is only the supplied min/max that is corrupted…
this may be a precursor to slow PSU decline and failure, i do NOT
know…but, i’d guess you are near or have already exceeded its
warranty period…those with a three year warranty are NOT
engineer/built them to last 10 years!
> On 06/18/2011 09:14 AM, DenverD wrote:
>> to me it looks like your config is scrambled…
>
> WAIT! i don’t know where the minimum and maximum values come from, do you?
>
> now that i think about it, those are probably embedded in the PSU
> circuitry…
>
<snip>
tpe;
These values are normally contained in /etc/sensors3.conf with an entry
similar to the following, the actual values vary by chip set:
If /etc/sensors3.conf does not exist then /etc/sensors is used, more details
are given by man sensors.conf. You should also check: http://www.lm-sensors.org/wiki/Configurations for board specific files.
I’m guessing your intx_min and intx_max are not configured correctly. You
can edit /etc/sensors3.conf as needed. You will need to edit as root with
either:
kdesu kwrite /etc/sensors3.conf
or
gnomesu gedit /etc/sensors3.conf
–
P. V.
“We’re all in this together, I’m pulling for you.” Red Green
Well, I have not forgot my maths, and usually with positive numbers min<max, not the other way arround! So, under that perspective, V5SB and VBat values are messed up.
OK, it seems that for the w83627thf- chips, the min/max values are based on calculations:
Taken from /etc/sensors3.conf
# Rather than an internal inverting op amp, the 627thf uses standard positive
# inputs and the negative voltages are level shifted by a 3.6V reference
# (same as 82d/83s).
# The math is convoluted, so we hope that your motherboard
# uses the recommended resistor values.
# Note that in1 (+12V) is the usual in4, and in4 (-12V) is the usual in5.
# Data sheet is obviously wrong for in4, the usual formula should work.
# No in5 nor in6.
>
> OK, it seems that for the w83627thf- chips, the min/max values are based
> on calculations:
> Taken from /etc/sensors3.conf
>
> Code:
> --------------------
>
>
> # Rather than an internal inverting op amp, the 627thf uses standard
positive
> # inputs and the negative voltages are level shifted by a 3.6V reference
> # (same as 82d/83s).
> # The math is convoluted, so we hope that your motherboard
> # uses the recommended resistor values.
> # Note that in1 (+12V) is the usual in4, and in4 (-12V) is the usual in5.
> # Data sheet is obviously wrong for in4, the usual formula should work.
> # No in5 nor in6.
>
> --------------------
>
tpe;
Are you sure that is for the min/max values and not just the current reading?
Look at the assignment for your chip. You should still be able to override
calculated values with the static entries as shown in the sensors3.conf
segment I posted. You could also check for an updated version for your
mother board at the address posted earlier.
–
P. V.
“We’re all in this together, I’m pulling for you.” Red Green
I’m old, but not too old to forget this very basic mathematical premise… I (wrongly) assumed you knew the limits were configurable - I’ve learned this years ago by simply reading the sensors README - and that perhaps you where concerned about the battery, mistakenly thinking it referred to a laptop battery, perhaps. There’s no call for irony, really.
The min/max values are configurable, they are not part of the sensor hardware, as other pointed in the previous posts. You can change/correct them if you will or, as you state they were right before, you may want to log a bug with the sensors package developers.