Not entirely a "silence" question

Got a shopping cart of parts that you want opinions on? Get advice from members on your planned or existing system (or upgrade).

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
rbsteffes
Posts: 114
Joined: Mon Sep 13, 2004 7:35 pm

Not entirely a "silence" question

Post by rbsteffes » Thu Oct 07, 2004 6:40 am

Ok I just got my new CPU in the mail so I eagerly put everything together last night and ran into some...oddness.

System:

AMD mobile 2400+ (Stock HS from an old retail package put on with radioshack white thermal compound)
SOLTEK 75MIV2-L SOCKETA mATX
512 MB Mushkin RAM
Seagate ST3200822A
Old PSU
Some old DVD player
Happauge WinTV-Radio


The PSU I have is the one I mentioned in another thread that I don't entirely trust, but still I didn't expect numbers like I'm getting, as I'm about to discuss. I'm not sure how familiar most of you are will Linux, but there's a tool to read diodes similiar to MBM called lm_sensors. Like MBM it can be configured to read different chips of chips and will try to autodetect the chips you have but it can get them wrong. The readings I'm getting off of it are so screwy that either my PSU is completely shot (but the system seems to RUN fine) or the sensors have got to be mislabeled. Here's the output I'm getting:

Code: Select all

[root@mythtv mythtv]# sensors
it87-isa-0290
Adapter: ISA adapter
VCore 1:   +1.31 V  (min =  +1.42 V, max =  +1.57 V)   ALARM
VCore 2:   +2.59 V  (min =  +2.40 V, max =  +2.61 V)
+3.3V:     +6.46 V  (min =  +3.14 V, max =  +3.46 V)   ALARM
+5V:       +5.08 V  (min =  +4.76 V, max =  +5.24 V)
+12V:     +11.90 V  (min = +11.39 V, max = +12.61 V)
-12V:     -19.38 V  (min = -12.63 V, max = -11.41 V)   ALARM
-5V:       -2.62 V  (min =  -5.26 V, max =  -4.77 V)   ALARM
Stdby:     +5.00 V  (min =  +4.76 V, max =  +5.24 V)
VBat:      +3.06 V
fan1:     4115 RPM  (min =    0 RPM, div = 8)
fan2:        0 RPM  (min = 3013 RPM, div = 8)          ALARM
fan3:        0 RPM  (min = 3000 RPM, div = 2)
M/B Temp:    +40°C  (low  =   +15°C, high =   +40°C)   sensor = thermistor
CPU Temp:    +44°C  (low  =   +15°C, high =   +45°C)   sensor = thermistor
Temp3:       +22°C  (low  =   +15°C, high =   +45°C)   sensor = thermistor

eeprom-i2c-0-50
Adapter: SMBus Via Pro adapter at 0400
Memory type:            DDR SDRAM DIMM
Memory size (MB):       512

lm90-i2c-0-4c
Adapter: SMBus Via Pro adapter at 0400
M/B Temp:    +40°C  (low  =    +0°C, high =   +70°C)
CPU Temp:  +46.6°C  (low  =  +0.0°C, high = +70.0°C)
M/B Crit:    +85°C  (hyst =   +75°C)
CPU Crit:    +85°C  (hyst =   +75°C)
Now the fans are okay, they are L1As. But the voltages? Holy crap! Anyone want to take a crack at guessing whether the numbers are mislabeled, wrong, or whether I need to turn off my machine RIGHT NOW before smoke comes out?

Also, I'm a little disappointed that the CPU is so hot running at 1.35 VCore, even with stock fan. Think that's an indication I need to reseat the heatsink?

I'm going to dig into the bios when I get home tonight and see what THEY say about the voltages, there should be a place to get that but I can't check them at the moment.


The other weird thing is that my RAM seems to buzz or hum when it's going through the selfcheck at startup. I've never heard that before, and was just a little weirded out by it. It's fairly loud, and stops as soon as the selfcheck is done. Clues?

Mati
Posts: 18
Joined: Wed Feb 12, 2003 11:14 am
Location: Estonia

Re: Not entirely a "silence" question

Post by Mati » Thu Oct 07, 2004 8:01 am

your PSU is probably fine, the sensors on your motherboard could be misreporting them. my motherboard (epox 4pea800) reports all voltages to be 0.2-0.4v lower than they are meant to be. I'm using linux and lm_sensors, too.

just get yourself a good digital multimeter and check the voltages out yourself. you can measure 5 ja 12 volts (red and yellow wire, black is ground) from a spare molex connector and 3.3 volts from the AUX connector (orange wire) if your PSU has one.
rbsteffes wrote:The other weird thing is that my RAM seems to buzz or hum when it's going through the selfcheck at startup. I've never heard that before, and was just a little weirded out by it. It's fairly loud, and stops as soon as the selfcheck is done. Clues?
this is probably the memory test sound some motherboards generate. you should be able to turn it off in bios.

DonP
Posts: 433
Joined: Mon Jun 09, 2003 5:47 pm
Location: London, UK

Re: Not entirely a "silence" question

Post by DonP » Thu Oct 07, 2004 5:39 pm

rbsteffes wrote:The PSU I have is the one I mentioned in another thread that I don't entirely trust, but still I didn't expect numbers like I'm getting, as I'm about to discuss. I'm not sure how familiar most of you are will Linux, but there's a tool to read diodes similiar to MBM called lm_sensors. Like MBM it can be configured to read different chips of chips and will try to autodetect the chips you have but it can get them wrong. The readings I'm getting off of it are so screwy that either my PSU is completely shot (but the system seems to RUN fine) or the sensors have got to be mislabeled.
...
Now the fans are okay, they are L1As. But the voltages? Holy crap! Anyone want to take a crack at guessing whether the numbers are mislabeled, wrong, or whether I need to turn off my machine RIGHT NOW before smoke comes out?
I reckon it's common for voltages to be mis-read by software - I get weird values for at least one of the rails on all machines where I run MBM and/or lm_sensors.
For comparison this is the output I get from my 75MIV2-L, XP-M 2400+ and Nexus NX-3000 PSU:

Code: Select all

root@mythtv:/etc/rc.d# sensors
it87-isa-0290
Adapter: ISA adapter
VCore 1:   +1.34 V  (min =  +1.42 V, max =  +1.57 V)   ALARM
VCore 2:   +2.61 V  (min =  +2.40 V, max =  +2.61 V)   ALARM
+3.3V:     +6.62 V  (min =  +3.14 V, max =  +3.46 V)   ALARM
+5V:       +5.07 V  (min =  +4.74 V, max =  +5.24 V)
+12V:     +11.84 V  (min = +11.40 V, max = +12.60 V)
-12V:     -19.54 V  (min = -12.63 V, max = -11.41 V)   ALARM
-5V:       -2.12 V  (min =  -5.24 V, max =  -4.76 V)   ALARM
Stdby:     +5.06 V  (min =  +4.74 V, max =  +5.24 V)
VBat:      +2.03 V
fan1:     3125 RPM  (min =    0 RPM, div = 8)
fan2:        0 RPM  (min = 3013 RPM, div = 8)          ALARM
fan3:        0 RPM  (min = 3000 RPM, div = 2)
M/B Temp:    +35°C  (low  =   +15°C, high =   +40°C)   sensor = thermistor
CPU Temp:    +39°C  (low  =   +15°C, high =   +45°C)   sensor = thermistor
Temp3:       +61°C  (low  =   +15°C, high =   +45°C)   sensor = thermistor

root@mythtv:/etc/rc.d#
I _think_ the Temp3 value is that reported by the onchip thermistor - but currently that value is a bit out of whack - it usually tracks the "CPU Temp" to within a few degrees but sometimes it goes nuts - sometimes drops to below room temp, sometimes temps way above the "CPU Temp" like now.

My uptime is at nearly two months now... and the system's been rock.
rbsteffes wrote:Also, I'm a little disappointed that the CPU is so hot running at 1.35 VCore, even with stock fan. Think that's an indication I need to reseat the heatsink?
Are these load temps? You can see my load temps above - that's a 2400+ XP-M with a DP-102 with a (temporary) crappy 60mm fan. Before I installed lm_sensors I used a thermocouple at the base of the DP-102 and got results in the same range (+- 5oC or so). I dunno about you but I reckon MY temps OWN for a CPU churning out the equivalent 2.4GHz of lovin F@H goodness.
If you are worried about the seating of your HSF and these are idle temps then try loading the CPU - if temps do go nuts then you probably do have poor HSF contact.
rbsteffes wrote:I'm going to dig into the bios when I get home tonight and see what THEY say about the voltages, there should be a place to get that but I can't check them at the moment.
Good idea - they should be accurate. I don't think BIOSes have an idle loop so in the BIOS it's kind like the CPU was pretty well loaded (at least this USED to be the case AFAIK) - so expect slightly higher temps than you see in an unloaded Linux (which has the idle/sleep states).
rbsteffes wrote:The other weird thing is that my RAM seems to buzz or hum when it's going through the selfcheck at startup. I've never heard that before, and was just a little weirded out by it. It's fairly loud, and stops as soon as the selfcheck is done. Clues?
I tend to agree with Mati on this - I've never had these noises because my case doesn't actually have a PC speaker.
btw.. you could also disable the memory check completely if you like - I think I've done this to minimise boot times.

rbsteffes
Posts: 114
Joined: Mon Sep 13, 2004 7:35 pm

Post by rbsteffes » Fri Oct 08, 2004 6:53 am

Thanks for the feedback. I'm highly paranoid about CPU temperatures, mostly because I can't afford to replace anything I destroy. I adjusted the config file to fix the +3v number, and I'm just ignoring the -5 and the -12 because I'm guessing they aren't reporting valid numbers and probably don't matter anyway.


I've got it full load now and am hitting 51 C which isn't terrible. Now I just need to quiet the thing down a little bit and I'll be set. The case it's in has crap for airflow. It has NO exhaust in the back and one intake in the front with a not so great underneath vent that's in the carpet. I cut out a small vent hole in the front to help pull in some air, and have an L1A going full blast in the PSU. I need to swap out the heatsink and maybe turn down the PSU fan and it'll be better, I think.

frosty
Posts: 636
Joined: Fri Jun 06, 2003 9:40 am
Location: USA

Post by frosty » Mon Oct 11, 2004 8:08 am

Dremmel or tin snip a 80 mm hole for exhaust get of any fan grills if safety allows, pets/kids? Get good airflow, how about cables, tied up neatly or using that automotive wrap or even black tape helps reduce size.

Post Reply