motherboard sensors broken?

jon_k · 11-04-2004, 04:21 PM

Here is the output of the 'sensors' command:

Code:

w83627hf-isa-0290
Adapter: ISA adapter
VCore 1:   +1.64 V  (min =  +1.76 V, max =  +1.94 V)
VCore 2:   +2.45 V  (min =  +1.76 V, max =  +1.94 V)
+3.3V:     +3.33 V  (min =  +3.14 V, max =  +3.48 V)
+5V:       +4.98 V  (min =  +4.77 V, max =  +5.25 V)
+12V:     +11.81 V  (min = +10.84 V, max = +13.21 V)
-12V:     -13.08 V  (min = -13.16 V, max = -10.77 V)
-5V:       -5.18 V  (min =  -5.23 V, max =  -4.73 V)
V5SB:      +5.47 V  (min =  +4.77 V, max =  +5.25 V)
VBat:      +3.37 V  (min =  +2.40 V, max =  +3.60 V)
fan1:        0 RPM  (min = 2657 RPM, div = 2)
fan2:     3229 RPM  (min = 2657 RPM, div = 2)
fan3:        0 RPM  (min = 1328 RPM, div = 4)
temp1:       +39°C  (high =   -21°C, hyst =   -14°C)   sensor = thermistor

temp2:     +45.0°C  (high =   +75°C, hyst =   +70°C)   sensor = diode
temp3:     -48.0°C  (high =   +75°C, hyst =   +70°C)   sensor = thermistor

vid:      +1.850 V  (VRM Version 8.2)
alarms:
beep_enable:
          Sound alarm disabled

eeprom-i2c-0-53
Adapter: SMBus I801 adapter at 5000
Memory type:            DRDRAM RIMM
Memory size (MB):       128

eeprom-i2c-0-52
Adapter: SMBus I801 adapter at 5000
Memory type:            DRDRAM RIMM
Memory size (MB):       128

eeprom-i2c-0-51
Adapter: SMBus I801 adapter at 5000
Memory type:            DRDRAM RIMM
Memory size (MB):       128

eeprom-i2c-0-50
Adapter: SMBus I801 adapter at 5000
Memory type:            DRDRAM RIMM
Memory size (MB):       128

Relevant lines include
temp1: +39°C (high = -21°C, hyst = -14°C) sensor = thermistor

temp2: +45.0°C (high = +75°C, hyst = +70°C) sensor = diode
temp3: -48.0°C (high = +75°C, hyst = +70°C) sensor = thermistor

Temp2 = CPU temp
Temp1 = Ambient Case Temp

How was the highest temp for the case temp -21 C when it is currently +39 C (breaking the high record of -21 C)? I'm in Texas anyways, and it's not possible it was ever that cold.

Whats the 'hyst' mean?

Why is temp3 -48 degree's (it always seems to be around that temp) again I'm in Texas and it's not possible for that temp unless I had some liquid nitrogen on it.

Any ideas?

Electro · 11-04-2004, 04:48 PM

Re-edit your lmsensors script to whatever formula or value it should be. Temperature sensors especially thermistor are not accurate at all. Thermistor have an log graph, so you will need a very long formula to make it close to linear. Linear means a straight line rising or failling for every point on the graph. Diodes and transistors make execellent temperature sensors because they have a predictable tempature charactistic that relates well with resistance.

Baddox · 11-04-2004, 04:48 PM

I may be wrong, but I've noticed with the utilities that come with motherboards, 3rd party utils, and in BIOS's, temperature sensors never seen to be reliable. For me the temperatures are either stuck at on stop and never change, or seen to be completely unresonably high or low, like in your case. In your case it looks like there's some sort of bug that just isn't recording new high-low records. But if you want really reliable measurements I'd recommend purchasing temperature probes that interface to 5.25" drive bay led displays. Sorry that I can't really help you with your problem though.

EDIT: sorry for the redundancy with the prev post, coincidentally we submitted at nearly the exact same time.

ror · 11-04-2004, 07:06 PM

just installed lm_sensors today

temp 3 (the crazy -21 one) will just be a sensor that's not actually used.