Wierd Temps for a dualie

I just got done with setting up my dual water cooled setup and I am getting some wierd temps for my cpu's. I have my cpu's in a series where pump>rad>cpu2>cpu1>pump. I m getting like 40c for cpu2 and only 30c for cpu1. Now to me this does not make much sense becasue seems like cpu2 should be cooler than cpu1 because it is getting cooled water fresh from the rad. I don't know what's up. help please

It doesnt make sense. There are a number of possible reasons why you are getting these figures. Before you water cooled were both cpus temps similar/same? Perhaps cpu1's block has not been mounted properly.

What are your systems specs?
What are you using to measure temperatures?

This rather proves the tough time they have with waterblock testing. There's so many factors involved to repeat the same results everytime.
I'd doubt you'll get both seeing the same temps, most don't.
There's differences in thermal goop thicknesses, mounting pressures, tubing pushing against a solid mounting, and my all time (least) favorite--thermal sensing on the motherboard.

You can try and get closer temps by seeing to the above named things, remount your waterblocks with a thin film of thermal goop, make sure your tubing isn't pushing against the waterblock etc. There's little you can do about the thermal sensing on the motherboard, but if it's using the little blue thermisters in the socket centers, try bending them upwards slightly and put a dab of your thermal goop on them. Remounting the cpu's should press them back down and give firm contact on the back of the cpu's.
Making sure you have the latest bios update can do wonders for your cpu temps too

Diggrr said:

This rather proves the tough time they have with waterblock testing. There's so many factors involved to repeat the same results everytime.
I'd doubt you'll get both seeing the same temps, most don't.
There's differences in thermal goop thicknesses, mounting pressures, tubing pushing against a solid mounting, and my all time (least) favorite--thermal sensing on the motherboard.

Click to expand...

Thats true but I doubt those differences will cause a 10C difference... I would think 2-3 maybe..

edit - unless the first block is mounted wrong...

Well assuming you have a decent flowrate, you can be pretty sure that the temperature of the both the water blocks is essentially the same.

So you either have a bad thermal contact on the hot cpu or incorrect temperature measurements. 10c (18f) is a huge difference, i'd think too big. I'd guess it was a measurement error.

mat6704AMD, did you take your readings from the bios or from a program like sandra? I remember back when I first got my old KT7Raid, sandra would report the system temp as that of my cpu

my temps are from a program called harware sensors monitor.

I have checked in the bios and they are correct. I am running twho AMD 1.2mp oc to 1.4 and I have already burn't out two of the same processors in similar ways so i am kinda desperate to find an answer. I will try the suggestions as soon as my dvd is done burning and get back to you. thanks guys

First, check your motherboard and make sure you know which CPU is actually number one and number two. On my Tyan Tiger S2466-4M board, the CPU I would have guessed was CPU1 is actually CPU2.

The next thing I would do is to test the effect of airflow on the thermistor. On my Tyan board, I can affect the temperature reading of a CPU by 10C simply by placing a cheap 80mm fan close by. If your board is susceptible to this, then the CPU that is nearer to the power supply, or some other fan, may indicate a lower temperature. Just hold a running fan close to a CPU and check the temp.

Next, I would switch the water blocks to see if the temperatures are affected. Here’s why.

I have two SK6 heat sinks with the same Delta fan model on each. However, one is consistently 7-8c degrees warmer than the other when switched on the same computer. One had a nice shiny and smooth surface. That was the hot one. The other didn't have such a nice looking surface. That was the cooler one. a precision straight edge showed that the heat sink with the shiny surface also had a convex surface. Someone had tried to lap the heat sink. I guess that when the temps were bad, he returned the heat sink claiming it was no good. The company then sold it to me.

So switching the water blocks will tell you if the blocks are causing the problem.

That’s what I’d do for starters.