A7V temps

[cleever]

I'm a proud owner of an A7V =) with the help of an article from Virtual Hideout I made myself a thermistor to use with the extra lead (JTPWR).when I pluged the thermistor it gave the same readings as the mobo's sensor,but with an plain old Hg thermometer I found that it was 6 degrees up from what it should be.does anyone know where the mobo's sensor is located,what temp it is for (case or chipset)???? could this mean that the other temps are not as they should be ...??

 

[Bignate603]

i've seen all different numbers cause of these thermsistors. That hg thermometer probably can't measure the exact temp cause it's not on the core. As soon as the heat hits the heatsink it starts to disapate. The most accurate measurement i've heard of is a laser probe shined in between the heatsink and chip.

 

[Mikewarrior2]

the HG thermal probe attached to the jtpower connector is infinitely more accurate than the thermistor that is one mb.

The on-mb thermistor is underneath the cpu(in the socket). If your'e lucky, it contacts the back/cold side of the cpu. Which is one of many reasons why it is inaccurate.

While your jtpower probe still isnt' 100% accurate, it is about as accurate as you'll get socket A temp measurement(low cost wise, that is). Just keep in mind that it still isn't super accurate.


Mike

 

[cleever]

well the thing is tha the thermistor is too thik to fit under the HFS so I didn't use it to mesure the CPU temp,instead I measured the HD temp and then I've plased a small Hg thermometer to measure the right HD temp.The diference was the 6 degrees I told u.Then I've measured the chipset of an old PCI MATROX Mystique and I hade the same 6 degrees diference.so,using the MBM as a monitoring prog,I've corected the diference (-6) and the readings I get are about the same as the Hg litle thermometer.the question is this..... are the other 2 sensors also inacurate ...?? should I lower them too by 6...???

 

[Mikewarrior2]

The problem is, adding a correction to the cpu temp still won't fix things, because you're still measuring the temp from a different spot.


Mike

 

[cleever]

My prob is that I have a 45 C CPU temp(Duron 600@900 1.65 Volt) and I don't know if it's right and what should I do about it and if I should compare it with the other temps from other users.

 

[Mikewarrior2]

First off, you can absolutely NOT compare temps with other socket A users. The method of reading temps is too unreliable to produce numbers with low enough error to compare correctly.

Secondly, depending on what heatsink you're using, your cpu will be running anywhere from 20C to 30C over ambient. If you are using a higher end socket A heatsink(taisol, globalwin, alpha), and are reading temps from a 1004 A7V bios, then your temps are great.

Thirdly, what is your ambient system temp?

And finally, again, do not compare socket A temps. There are many people on every bbs who claim their 1.1ghz T-bird 1.85V runs at 34C or 35C. And that is ridiculously impossible, regardless of heatsink. Go by if the system is stable, and the reading is within a reasonable number, then you're fine.


Mike

 

[Davegod75]

my jptwt reading and the reading from the on board monitor never differ more than 1 degree C. The onboard one is usually hotter by 1 c. Most of the time they are the same though

 

[Mikewarrior2]

DaveGod75,

Are you running a later a7v bios?

If you are running an older one, your jtpwr, if properly placed, will read higher temps(also more accurate temps).


Mike

 

[cleever]

my readings were the same befor the correction I made,(+ -) 1 from the mobo's sensor.that is what made me think it might be posible the other readings are wrong.