Fanless XP-120 with my CPU?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Fanless XP-120 with my CPU?
I plan on buying one of these later with new case and mobo (Supports this cooler). I have a Pentium4 2.26Ghz (533FSB Northwood) and want to know if its ok to run it without a fan for now. Other fans in the case will be psu fan, 120mm rear fan and Arctic Cooling Silencer 1 on my Radeon 9800Pro
So is it ok to run it like that until i can afford to get a fan for the heatsink?
If its not ok then i can put the case fan on the heatsink and my current 92mm fan on the rear.
Just want to make sure that i wont fry the CPU or something.
Note that its currently on a heatsink but theres a duct going to the rear fan (Dell case etc)
Many Thanks
Ben
So is it ok to run it like that until i can afford to get a fan for the heatsink?
If its not ok then i can put the case fan on the heatsink and my current 92mm fan on the rear.
Just want to make sure that i wont fry the CPU or something.
Note that its currently on a heatsink but theres a duct going to the rear fan (Dell case etc)
Many Thanks
Ben
-
- *Lifetime Patron*
- Posts: 51
- Joined: Sat Apr 09, 2005 10:35 am
- Location: Berkeley, CA
- Contact:
Here are a couple of free options:
Motherboard Monitor, no longer supported, may not know about your mobo.
http://mbm.livewiredev.com/
Everest Home Edition
http://www.lavalys.com/products/overvie ... =1&lang=en
Motherboard Monitor, no longer supported, may not know about your mobo.
http://mbm.livewiredev.com/
Everest Home Edition
http://www.lavalys.com/products/overvie ... =1&lang=en
I replaced MBM5 with Speedfan because the author of MBM5 has stopped working on it, and it can't read temps of my SATA drives which Speedfan can. However, Speedfan can only show one temp in the tool tray. Unless I'm missing something (only started using it two days ago - tips, anyone?)
The P4 should throttle if the core temp gets too high. I think your P4 2.26 should do fine.
Depending in your case layout, you may want to consider fashioning a shroud around the PSU fan (if it is a 120mm fan PSU), the exhaust fan and the XP-120 to redirect more air over the heat sink.
BIOS may not give you better temps compared to Windows. With my Winchester 3200+ with XP-120 (with a very low speed fan), the temperature in BIOS goes up to the 40s, but in Windows it drops to the 30s (without C&Q). I'm not sure if it is due to Windows executing HALT in the idle loop, or to the way the BIOS is written (keeps polling for keypress thus keeping CPU at 100%), or just flaky sensor reading either in Windows or BIOS.
When I enable CnQ temp drops to the mid 20s (ambient temp in the low 20s). Before you get too impressed with the temps, this is an Asus MB and the temps are probably higher than reported. Still, dropping 10 C using CnQ (and dropping almost 20W on the AC outlet) has me really pumped.
Edit : just noticed - CJ4, Welcome to SPCR Forums!
The P4 should throttle if the core temp gets too high. I think your P4 2.26 should do fine.
Depending in your case layout, you may want to consider fashioning a shroud around the PSU fan (if it is a 120mm fan PSU), the exhaust fan and the XP-120 to redirect more air over the heat sink.
BIOS may not give you better temps compared to Windows. With my Winchester 3200+ with XP-120 (with a very low speed fan), the temperature in BIOS goes up to the 40s, but in Windows it drops to the 30s (without C&Q). I'm not sure if it is due to Windows executing HALT in the idle loop, or to the way the BIOS is written (keeps polling for keypress thus keeping CPU at 100%), or just flaky sensor reading either in Windows or BIOS.
When I enable CnQ temp drops to the mid 20s (ambient temp in the low 20s). Before you get too impressed with the temps, this is an Asus MB and the temps are probably higher than reported. Still, dropping 10 C using CnQ (and dropping almost 20W on the AC outlet) has me really pumped.
Edit : just noticed - CJ4, Welcome to SPCR Forums!
-
- Posts: 968
- Joined: Fri Jan 07, 2005 7:13 pm
- Location: Bristol, UK
- Contact:
hi guys, i'm planning on buying a celeron 2.8GHZ to use with xp-120.
do you think this would run passively without a fan to go with xp-120 ? the case will probably be sonata with a nexus 120mm case fan.
or would a sempron 3300+ run cooler than the celeron? i couldnt find the power consumptions on each on google :/ any ideas?
Thanks.
do you think this would run passively without a fan to go with xp-120 ? the case will probably be sonata with a nexus 120mm case fan.
or would a sempron 3300+ run cooler than the celeron? i couldnt find the power consumptions on each on google :/ any ideas?
Thanks.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
Some of these points have been made in these forums many times before...
1) There is little to be gained in running a CPU fanless unless there are no fans elsewhere. One less fan is not usually audible if all the fans are running very slowly and are quiet fans.
2) There is a huge cooling benefit to having a fan on the CPU cooler. Unless you have a case fan almost next to it, which means it's hardly fanless, the difference between minimal airflow -- say <10cfm -- and only convection is in the order of >20C with almost any HS and CPU. Even CPUs that produce much less heat than the ones discussed in this thread.
3) There is also the need to cool the voltage regulator module (VRM) on the motherboard. This is essentially a mini-power supply that takes the 12V from the AUX12V (2x12V) input and converts it to the voltages that the CPU uses. It's the coils, caps, etc around the CPU socket.
The efficiency of VRMs varies from as low as <75% to perhaps 90% on top server boards, and averaging ~80% on good quality desktop boards. This means that to deliver 89W to a 2.8G Prescott (even a Celeron) at full tilt, an 80% efficient VRM needs 112W input from the PSU. The heat that the VRM generates is 32W!!!
Intel and AMD tech docs actually specify a CPU fan as a required component of VRM / motherboard cooling. The air blown around the CPU by the HS fan has this additional function of helping to cool the VRM components. AMD & Intel assume the CPU fan is going to be there, as do the motherboard manufacturers. So removing the CPU fan means you run the risk of less stability and shorter component life.
1) There is little to be gained in running a CPU fanless unless there are no fans elsewhere. One less fan is not usually audible if all the fans are running very slowly and are quiet fans.
2) There is a huge cooling benefit to having a fan on the CPU cooler. Unless you have a case fan almost next to it, which means it's hardly fanless, the difference between minimal airflow -- say <10cfm -- and only convection is in the order of >20C with almost any HS and CPU. Even CPUs that produce much less heat than the ones discussed in this thread.
3) There is also the need to cool the voltage regulator module (VRM) on the motherboard. This is essentially a mini-power supply that takes the 12V from the AUX12V (2x12V) input and converts it to the voltages that the CPU uses. It's the coils, caps, etc around the CPU socket.
The efficiency of VRMs varies from as low as <75% to perhaps 90% on top server boards, and averaging ~80% on good quality desktop boards. This means that to deliver 89W to a 2.8G Prescott (even a Celeron) at full tilt, an 80% efficient VRM needs 112W input from the PSU. The heat that the VRM generates is 32W!!!
Intel and AMD tech docs actually specify a CPU fan as a required component of VRM / motherboard cooling. The air blown around the CPU by the HS fan has this additional function of helping to cool the VRM components. AMD & Intel assume the CPU fan is going to be there, as do the motherboard manufacturers. So removing the CPU fan means you run the risk of less stability and shorter component life.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
Dell uses a ducted fan setup that puts the HS/CPU socket very close to the fan with a fair amount of directed airflow. I would not call this fanless. It's just that the HS does not have a fan mounted directly on it.BenW wrote:Thanks for that detailed post Mike!
I'm only planning on running it fanless for a while and that due to finances, not noise reduction
Why is it that people like dell have no CPU fan if its basically a requirment?
Ah right!
So my new setup should look something like this (Same case), bar the cpu fan and i will have XP-120 instead of 92
http://www.hostdub.com/albums/Audiophiliac/DSC00310.jpg
Should that case fan be close enough to the cpu heatsink?
So my new setup should look something like this (Same case), bar the cpu fan and i will have XP-120 instead of 92
http://www.hostdub.com/albums/Audiophiliac/DSC00310.jpg
Should that case fan be close enough to the cpu heatsink?
-
- Friend of SPCR
- Posts: 502
- Joined: Sat Oct 23, 2004 2:02 pm
That is still one less fan to deal with through. Are there any companies that make ducts like Dell's for generic cases?MikeC wrote:Dell uses a ducted fan setup that puts the HS/CPU socket very close to the fan with a fair amount of directed airflow. I would not call this fanless. It's just that the HS does not have a fan mounted directly on it.BenW wrote:Thanks for that detailed post Mike!
I'm only planning on running it fanless for a while and that due to finances, not noise reduction
Why is it that people like dell have no CPU fan if its basically a requirment?
See page 64 of ftp://download.intel.com/design/celeron ... 235304.pdf. The TDP of a Celeron D 335 (2.8GHz) is 73W. This is the *idle* power (see footnote 1: "The TDP is not the maximum power that the processor can dissipate").perplex wrote:hi guys, i'm planning on buying a celeron 2.8GHZ to use with xp-120.
do you think this would run passively without a fan to go with xp-120 ? the case will probably be sonata with a nexus 120mm case fan.
or would a sempron 3300+ run cooler than the celeron? i couldnt find the power consumptions on each on google :/ any ideas?
Thanks.
On the other hand, the TDP of a Sempron 3300+ is 62W and AMD measures MAXIMUM TDP. See page 13 of http://www.amd.com/us-en/assets/content ... 30430.pdf) and read footnote 1 ("Thermal Design Power (TDP) is measured under the conditions of TCASE Max, IDD Max, and VDD=VID_VDD..."). Granted, this is an older document, but I don't believe AMD has changed their calculation method for TDP.
Even if AMD was reporting idle TDP instead of MAX TDP, the Sempron still runs 11W cooler than the Celeron D and is faster. Case closed.
-Derek