VERY Strange temps

cougem · 07-25-2004, 01:17 PM

I was getting around 34 degrees idle on my 3000+ (C0 Newcastle, water cooled) when I got up in the morning (It was around 35/36 as soon as I mounted, before the AS5 had set) and I'd verified this was a typical temp as it got remounted twice.

It's running on a Chaintech VNF3-250, using the latest beta bios.

I noticed my cg Claw 3200+ arrived, and within an hour I had is AS5ed up. I was a pretty sweet spreading; an incredible thin layer, and yet the temperature reading read 53 degrees in bios. In windows, using the same program as before, it was reading 50 idle.

An extra 15 degrees on my idle temp upgrading from a C0 3000+ Newcastle, to a 3200+ cg Claw? That can't be right I thought. I took my water block off (MCW-6002), wiped all the AS5 off both surfaces, and remounted it with new AS5. Exactly the same as before. And I remounted it once again today, still no joy.

I'm pretty sure it's not my pump as my GPU temps haven't risen at all, staying at around 30 degrees idle.

One strange point is that when I try to stress it, say by running the 8M super-pi, it jumps from 50 degrees to 56 in 1 second, and that isn't an exaggeration. It then stays there for around 10 seconds, then goes down to 53 (?!?) while the test is still running, then it's jumping around 53-56 for the next few minutes, seemingly randomly, one second it could be 56, 5 seconds later, 53. Then as soon as I stop it, as in 3 seconds later, it's back down to 50 degrees.

What the hell's going on here do you reckon? Doesn't make sense to me.

I've got a pretty nice setup, MCP600, MCW-6002, 2 x single heatercores with 120mm fans, MAZE4 GPU, with lots of airflow.

Any help would be appreciated

jlrii · 07-25-2004, 01:28 PM

What version MCP600? If it's the old one the impeller could be stsrting to lock up. How do the rads feel are the warmed than usual? Could you have some air in the rads that breaks free now & again? :shrug:

cougem · 07-25-2004, 01:38 PM

Nah it's not air in the rads because before I installed the new CPU it was fine and I hadn't touched the rads since. I doubt it's the pump either because my GPU is fine, but I can't be 100% sure.

How do I tell which version the pump is?

cougem · 07-25-2004, 01:39 PM

By the way, another point is that however warm the room is, either really hot or a huge draft, the temperature never changes from around 49-51 degrees, even if the room changes much more, or even if I've changed from my old case with virtually no air flow, to my new one with a big 120mm sucking air over it.

jlrii · 07-25-2004, 01:43 PM

Not sure...I think the new ones model or serial# ends in -s. I've got the old version...still running after many months (jr crosses fingers)

BTW if it is the pump contact the seller for the new impeller assy. Try running a graphics bench to see if temps vary at all, much less heat put out from your gpu so it might not be as noticable. Also try the old CPU again & see if it does the same thing too. That would narrow it down quite a bit. Strange problem.

cougem · 07-25-2004, 02:21 PM

Here's a diagram of the flow.

The blue arrows show the water flow (Pump -> Rad -> Maze 4 GPU -> MCW-6002 -> Rad -> Pump)

The red arrows show air. There's holes in the bottom of the case, and in the stand the right rad is sitting on, below the PSU, which has a fan in the bottom to suck air through the rad as well as the fan on it.

Chew_Toy · 07-25-2004, 02:35 PM

I think it could be the on-chip diode reading differently from the older cpu.

cougem · 07-25-2004, 02:36 PM

It's very very different though, 15 degrees different, just for an extra 512k cache.

That pretty common then?

Chew_Toy · 07-25-2004, 02:56 PM

I dont know how common, I am just saying that if it is the only thing that changed that is the 1st place to look. My experience with P4s is that they all read about the same, but have heard other report variances with different chips on same setups. I have 3 P4s that can do 3500~3600 @ 1.6v on my setup and as far as I can tell they all read within 1°c of each other on my MB.

Also according to the instructions for AS5 on a cpu with an IHS you should put a small amount (grain of rice size) on the center and while mounting your block give it a couple twists to spread and continue mounting.

SlaterSpeed · 07-25-2004, 03:39 PM

Its just down to being a totaly differant cpu which is going to have differant heat output and also a totaly differant diode. The diode could be falty which is causing it to read wrong

One thing is for sure the 2 differant temps are totaly useless for comparison.

jlrii · 07-25-2004, 06:29 PM

Like I said try the old one & see if the problem is still there. If it is not you can chaulk it up to the new chip/ diode or opssibly the new chip is not seated properly.

07-25-2004, 01:17 PM	#1
cougem Cooling Savant Join Date: May 2004 Location: Cambridge Uni Posts: 176	VERY Strange temps I was getting around 34 degrees idle on my 3000+ (C0 Newcastle, water cooled) when I got up in the morning (It was around 35/36 as soon as I mounted, before the AS5 had set) and I'd verified this was a typical temp as it got remounted twice. It's running on a Chaintech VNF3-250, using the latest beta bios. I noticed my cg Claw 3200+ arrived, and within an hour I had is AS5ed up. I was a pretty sweet spreading; an incredible thin layer, and yet the temperature reading read 53 degrees in bios. In windows, using the same program as before, it was reading 50 idle. An extra 15 degrees on my idle temp upgrading from a C0 3000+ Newcastle, to a 3200+ cg Claw? That can't be right I thought. I took my water block off (MCW-6002), wiped all the AS5 off both surfaces, and remounted it with new AS5. Exactly the same as before. And I remounted it once again today, still no joy. I'm pretty sure it's not my pump as my GPU temps haven't risen at all, staying at around 30 degrees idle. One strange point is that when I try to stress it, say by running the 8M super-pi, it jumps from 50 degrees to 56 in 1 second, and that isn't an exaggeration. It then stays there for around 10 seconds, then goes down to 53 (?!?) while the test is still running, then it's jumping around 53-56 for the next few minutes, seemingly randomly, one second it could be 56, 5 seconds later, 53. Then as soon as I stop it, as in 3 seconds later, it's back down to 50 degrees. What the hell's going on here do you reckon? Doesn't make sense to me. I've got a pretty nice setup, MCP600, MCW-6002, 2 x single heatercores with 120mm fans, MAZE4 GPU, with lots of airflow. Any help would be appreciated

07-25-2004, 01:43 PM	#5
jlrii Cooling Savant Join Date: Feb 2004 Location: Massachusetts, USA Posts: 158	Not sure...I think the new ones model or serial# ends in -s. I've got the old version...still running after many months (jr crosses fingers) BTW if it is the pump contact the seller for the new impeller assy. Try running a graphics bench to see if temps vary at all, much less heat put out from your gpu so it might not be as noticable. Also try the old CPU again & see if it does the same thing too. That would narrow it down quite a bit. Strange problem. Last edited by jlrii; 07-25-2004 at 01:49 PM.

07-25-2004, 03:39 PM	#10
SlaterSpeed Cooling Savant Join Date: Nov 2003 Location: Suffolk, UK Posts: 234	Its just down to being a totaly differant cpu which is going to have differant heat output and also a totaly differant diode. The diode could be falty which is causing it to read wrong One thing is for sure the 2 differant temps are totaly useless for comparison. __________________ aka. slater3333uk - The self proclaimed 'Middle Player' 'Liquified'

07-25-2004, 01:28 PM	#2
jlrii Cooling Savant Join Date: Feb 2004 Location: Massachusetts, USA Posts: 158	What version MCP600? If it's the old one the impeller could be stsrting to lock up. How do the rads feel are the warmed than usual? Could you have some air in the rads that breaks free now & again? :shrug:

07-25-2004, 01:38 PM	#3
cougem Cooling Savant Join Date: May 2004 Location: Cambridge Uni Posts: 176	Nah it's not air in the rads because before I installed the new CPU it was fine and I hadn't touched the rads since. I doubt it's the pump either because my GPU is fine, but I can't be 100% sure. How do I tell which version the pump is?

07-25-2004, 01:39 PM	#4
cougem Cooling Savant Join Date: May 2004 Location: Cambridge Uni Posts: 176	By the way, another point is that however warm the room is, either really hot or a huge draft, the temperature never changes from around 49-51 degrees, even if the room changes much more, or even if I've changed from my old case with virtually no air flow, to my new one with a big 120mm sucking air over it.

07-25-2004, 02:21 PM	#6
cougem Cooling Savant Join Date: May 2004 Location: Cambridge Uni Posts: 176	Here's a diagram of the flow. The blue arrows show the water flow (Pump -> Rad -> Maze 4 GPU -> MCW-6002 -> Rad -> Pump) The red arrows show air. There's holes in the bottom of the case, and in the stand the right rad is sitting on, below the PSU, which has a fan in the bottom to suck air through the rad as well as the fan on it.

07-25-2004, 02:35 PM	#7
Chew_Toy Cooling Savant Join Date: Apr 2004 Location: Wisconsin Posts: 120	I think it could be the on-chip diode reading differently from the older cpu.

07-25-2004, 02:36 PM	#8
cougem Cooling Savant Join Date: May 2004 Location: Cambridge Uni Posts: 176	It's very very different though, 15 degrees different, just for an extra 512k cache. That pretty common then?

07-25-2004, 02:56 PM	#9
Chew_Toy Cooling Savant Join Date: Apr 2004 Location: Wisconsin Posts: 120	I dont know how common, I am just saying that if it is the only thing that changed that is the 1st place to look. My experience with P4s is that they all read about the same, but have heard other report variances with different chips on same setups. I have 3 P4s that can do 3500~3600 @ 1.6v on my setup and as far as I can tell they all read within 1°c of each other on my MB. Also according to the instructions for AS5 on a cpu with an IHS you should put a small amount (grain of rice size) on the center and while mounting your block give it a couple twists to spread and continue mounting.

07-25-2004, 06:29 PM	#11
jlrii Cooling Savant Join Date: Feb 2004 Location: Massachusetts, USA Posts: 158	Like I said try the old one & see if the problem is still there. If it is not you can chaulk it up to the new chip/ diode or opssibly the new chip is not seated properly.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)