Fan voltages problem

[ICIT2LOL]

Am fixing a friends older Compaq some idea of the specs in the Speccy snip/s.
Problem is the CPU is running rather hot average 35C at idle and up to 60C on scans and net searches.

I have cleaned everything redone the compound with Arctic Silver and reseated all the plugs etc etc

Now the problem is the case and CPU fans are not running fast enough - the volts at the 24 pin are fine - jut on 5.06v each red. Now the volts at the fan headers both the CPU and back case fan are 4.6v and I know they can run faster because I have had them out and un on my test PSU.

So any idea on why the volts are dropping from the 24 pin to the headers??

CPU fan runs full blast on power up then slows to a steady - what I can only describe as what one would exect on idle.
Temps can be seen in the pics and the discrepsancy or so it seems with the volts whetaer tis is Specy or the machine I don't know.

 

[fireberd]

Maybe power supply problem. Many factory PC's have marginal power supplies and over time they can deteriorated and unable to handle the power load. One area to look at. If you have another power supply you can use in the PC, for testing, that would be one thing to try.

 

[ICIT2LOL]

Thanks Jack I didn't get a mail to let me know you had replied mate.
The PSU checked out fine although the watts are a bit on the pathetic side at 228.

Still since then I have installed the rear fan to a Molex connector and it is running as fast as it should and dragging a little more air albeit it hot through the case which is sadly lacking in a lot of inlet venting.

I also installed CPUID HWMonitor and the temps now are as in the pic the only thing is the two cores are quite different temps I expected some but 10C has got me wondering if the CPU is on the way out?
The machine is a tiny bit noisier but as long as the temps are down eh?
I think Speccy could do with bit of work on it to be honest as it is still showing 35C!

But thanks again for chipping in - appeciate it

 

[fireberd]

Power supplies, such as the one you have are marginally acceptable when new, but given some age it may have deteriorated somewhat. I would still try it with a known good power supply and see how it acts and how the fan connector voltages are on the motherboard.

Given there may be other component issues (aging) that may also be contributing to the overall picture.

Thanks for the rep, too

Jack

 

[ICIT2LOL]

Power supplies, such as the one you have are marginally acceptable when new, but given some age it may have deteriorated somewhat. I would still try it with a known good power supply and see how it acts and how the fan connector voltages are on the motherboard.

Given there may be other component issues (aging) that may also be contributing to the overall picture.

Thanks for the rep, too

Jack

No probs mate I suspect you are on the money there too with the aging mobo

 

[madcratebuilder]

Have you checked the BIOS and all fan control software that may be installed?

If all fans spin up to full rpm at POST that should indicate adequate voltage. If the fans slow down after POST then it's a BIOS setting or software setting.

 

[ICIT2LOL]

Yes I checked to see if there was any fan control in the BIOS but nothing that I could see and the fan surge on power up and then reverting back to an idle speed is something I don't understand . Perhaps there is fan control in the BIOS but where I don't know.
The CPU fan header is a three pin one with the fan having red, black, and blue leads which again is odd coding to me but at least the machine going reasonably well .
my fried has taken it back and if he brings it back at any time I shall have to test that header pin that the blue lead goes to to find out just what volts are there.

Anyhow I fixed extra fans in and it has brought the temps down to 30C on one core and 40C on the other which again i think is rather odd but at least it isn't hitting 50 - 66C now even in the very hot conditions we have right now.

 

[Britton30]

If you work on it again, there should be AMD Cool'nQuiet. That's the AMD fan control, you can disable it.

 

[ICIT2LOL]

If you work on it again, there should be AMD Cool'nQuiet. That's the AMD fan control, you can disable it.

Thanks Gary I probably will be seeing it again especially as the weather is like it is at the moment.

 

[Britton30]

OK, with Cool'n'Quiet enabled with a 3-pin connector, fan speed is controlled by varying the voltages. In fact voltage is used on all speed controlled fans with 3 connectors.

They are +V, -V, and RPM sensor.

 

[ICIT2LOL]

OK, with Cool'n'Quiet enabled with a 3-pin connector, fan speed is controlled by varying the voltages. In fact voltage is used on all speed controlled fans with 3 connectors.

They are +V, -V, and RPM sensor.

Ok Gary now I find that in the BIOS I take it or is a program I download? as I haven't had chance to look into it yet.

 

[Britton30]

It's part of the BIOS for AMD system boards. It will be listed under some sub-menus, or a sub under one of them.

 

[ICIT2LOL]

It's part of the BIOS for AMD system boards. It will be listed under some sub-menus, or a sub under one of them.

Ok I will have a look as I think this board was an ECI?? one obviously made I suppose for Compaq.

 

[Britton30]

It can be any brand board, I meant a board that uses an AMD CPU.

 

[ICIT2LOL]

It can be any brand board, I meant a board that uses an AMD CPU.

Gotcha

 

[Mellon Head]

Something that may explain the temperature difference on your cores. That's a Brisbane core processor and AMD always had trouble with reading the temps off of a Brisbane CPU. They're never right, or if they are they never agree. I have a Brisbane 5400+ in my HTPC and the two cores aren't even remotely close in temps. Sometimes they differ as much as 20 degrees.

It's a known bug and it's nothing to worry about unless you're overclocking.

 

[ICIT2LOL]

Something that may explain the temperature difference on your cores. That's a Brisbane core processor and AMD always had trouble with reading the temps off of a Brisbane CPU. They're never right, or if they are they never agree. I have a Brisbane 5400+ in my HTPC and the two cores aren't even remotely close in temps. Sometimes they differ as much as 20 degrees.

It's a known bug and it's nothing to worry about unless you're overclocking.

Actually Melon if I remember right it was a Brisbane as I did look that CPU on CPU World and was surprise at what a low temp range it could only run to.
My mate did say it was touching base with 65C at times and the lettering was red when he opened it up and had a room pedestal fan blown in . The difference then as now is 10C and from the few bits I did like the cleaning, compound, and adding that extra fan at the rear it was running 28C and 37C when he picked it last night (Ambient temp 32C as we are in a heatwave at the moment) so I am guessing those steps are not too bad. My last alternative is to add yet another fan to the vented side (no fixings so will have to improvise) but as of now things are not to bad considering it is going to get hotter they tell us

 

[Layback Bear]

From the Speccy picture in post #1 your 3.3V is only 1.76V.
That is way to low.
Should be around 3.290V or a little higher with the system up and running.

 

[ICIT2LOL]

From the Speccy picture in post #1 your 3.3V is only 1.76V.
That is way to low.
Should be around 3.290V or a little higher with the system up and running.

Ok Jack I will get it back and flip in a spare PSU as that one you will have seen is a rather sad 228 watt model.