5

I am building a small NAS server, the motherboards is a Intel D2700MUD (fanless) and the PSU is "Ace Outburst 480W". After installing Ubuntu Server, I installed lm-sensors. The output:

coretemp-isa-0000
Adapter: ISA adapter
Core 0:       +38.0 C  (high = +80.0 C, crit = +100.0 C)
Core 1:       +33.0 C  (high = +80.0 C, crit = +100.0 C)

w83627dhg-isa-0290
Adapter: ISA adapter
Vcore:        +1.22 V  (min =  +0.00 V, max =  +1.74 V)
in1:          +1.02 V  (min =  +0.58 V, max =  +1.06 V)
AVCC:         +3.38 V  (min =  +2.78 V, max =  +0.64 V)  ALARM
+3.3V:        +3.38 V  (min =  +0.82 V, max =  +0.59 V)  ALARM
in4:          +1.03 V  (min =  +1.25 V, max =  +0.42 V)  ALARM
in5:          +1.52 V  (min =  +1.78 V, max =  +0.51 V)  ALARM
in6:          +1.02 V  (min =  +0.16 V, max =  +0.10 V)  ALARM
3VSB:         +3.34 V  (min =  +1.65 V, max =  +1.15 V)  ALARM
Vbat:         +3.22 V  (min =  +2.34 V, max =  +1.02 V)  ALARM
fan1:           0 RPM  (min = 2109 RPM, div = 128)  ALARM
fan2:           0 RPM  (min =  458 RPM, div = 128)  ALARM
fan3:           0 RPM  (min =  234 RPM, div = 128)  ALARM
fan4:           0 RPM  (min =  659 RPM, div = 128)  ALARM
fan5:           0 RPM  (min = 5273 RPM, div = 128)  ALARM
temp1:        +44.0 C  (high = +35.0 C, hyst = +32.0 C)  ALARM  sensor = diode
temp2:       +110.5 C  (high = +80.0 C, hyst = +75.0 C)  ALARM  sensor = diode
temp3:        +30.5 C  (high = +80.0 C, hyst = +75.0 C)  sensor = diode
cpu0_vid:    +2.050 V

Do I need to be worried about the voltage alarm on in4,in5,in6.. or temp1/temp2?

Improve this question

3 Answers 3

Reset to default
4

The limits aren't configured correctly, they're probably defaults or from some other board. For example:

+3.3V: +3.38 V (min = +0.82 V, max = +0.59 V)

That looks reasonable for a 3.3V supply, but the limits are completely out, and the max is lower than the min!

temp2 looks a bit worrying, but that may also be a misconfigured sensor.

Improve this answer
2
  • Do you have any tips on finding/calculating reasonable limits? The closest I have found is the motherboard spec but it doesn't say anything about how much voltage diff that's tolerable. downloadmirror.intel.com/20716/eng/D2700MUD_TechProdSpec01.pdf
    – pojo
    Apr 16, 2012 at 10:56
  • Afraid not .. there might be something in the BIOS, or you could just log values for 24 hours, average them and round to nearest decimal place, then set a 5% threshold on that.
    – pjc50
    Apr 16, 2012 at 13:05
2

Voltages are rarely reported by drivers correctly, because a scaling factor needs to be applied to the limited range of the internal sensors.

See this page for a description and tutorial of how to fix the scaling factors for your particular system setup:

http://www.lm-sensors.org/wiki/VoltageLabelsAndScaling

Once you know which sensors are reporting which voltages, and the numbers are scaled correctly, you can setup or max and min values appropriately. The page above gives an example of this.

Improve this answer
0

Voltages should normally be with 5% in most cases and should be never off by more then 10%. Consult the device specs for exact figures. Ignore voltage tolerance errors only if you have good backups and can afford downtime.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .