P180 + external HDs = A64 X2 system 100% fanless at idle
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Devonavar
P180 + external HDs = A64 X2 system 100% fanless at idle
The P180 is amazing, but it has the problem that the lower chamber relies on a fan (either in the PSU or mounted in the middle of the chamber) if you have hard drives generating heat down there. So I removed my hard drives from the case entirely, and voila - my system survives running entirely fanless at idle.
My components are:
Antec P180
Antec Phantom 350
Asus A8N-SLI Premium
AMD Athlon64 X2 4200+ (Manchester) w/ Scythe Ninja
Gainward GeForce 6200 TurboCache w/ stock fanless heatsink
NEC DVD-RW ND-3540A
2x Samsung SP2504C in Nexus Drive-a-way outside case
I have a single Nexus 120mm as exhaust at the back, which at 12V is enough to cool the entire system comfortably even under extreme load. I'll probably let it stay running at 5V even at idle, just to be safe - I can't hear it at 5V from where I'm sitting anyway.
My components are:
Antec P180
Antec Phantom 350
Asus A8N-SLI Premium
AMD Athlon64 X2 4200+ (Manchester) w/ Scythe Ninja
Gainward GeForce 6200 TurboCache w/ stock fanless heatsink
NEC DVD-RW ND-3540A
2x Samsung SP2504C in Nexus Drive-a-way outside case
I have a single Nexus 120mm as exhaust at the back, which at 12V is enough to cool the entire system comfortably even under extreme load. I'll probably let it stay running at 5V even at idle, just to be safe - I can't hear it at 5V from where I'm sitting anyway.
Oh, heh, yeah, I should've realized you guys wanted that
The numbers when running fanless and idle I remember very clearly, because I was monitoring them very carefully . The CPU got up to 60C, the GPU 72C, the MB 52C, and my own sensor that was stuck to the metal block of the chipset heatpipe was reading 82.5C. They sound extremely high, especially the last one, but they're all well below safe operating temperatures.
The heatpipe is over 65C at all times, and never below 75C at load, even with much more airflow. I worried a lot about that at first, but stopped when Asus told me the safe operating temperature of that chipset is around 100C.
I'm in Linux right now, from where I haven't figured out how to adjust the fan speed, nor how to load the GPU, but I can tell you that right now the fan is running at 9.2V, the CPU is at 38C, the MB at 46C, and the GPU at 60C.
The numbers when running fanless and idle I remember very clearly, because I was monitoring them very carefully . The CPU got up to 60C, the GPU 72C, the MB 52C, and my own sensor that was stuck to the metal block of the chipset heatpipe was reading 82.5C. They sound extremely high, especially the last one, but they're all well below safe operating temperatures.
The heatpipe is over 65C at all times, and never below 75C at load, even with much more airflow. I worried a lot about that at first, but stopped when Asus told me the safe operating temperature of that chipset is around 100C.
I'm in Linux right now, from where I haven't figured out how to adjust the fan speed, nor how to load the GPU, but I can tell you that right now the fan is running at 9.2V, the CPU is at 38C, the MB at 46C, and the GPU at 60C.
That's something I've been wondering about. Why do components have a listed max operating temperature if it's a really bad idea to run it even close to that hot? Has there been any definitive testing on how temperature affects component lifespan in modern desktop computers?hyperq wrote:Running all the components at high temperature might shorten their lifespan.
My GPU, for instance, has drivers which will throttle the speed if its temperature ever reaches 145C (!). This is over twice the temperature you're saying is "up there", and it must mean something. If Nvidia thought it would be harmful for the GPU to exceed, say, 80C, why didn't they set the throttle-limit at that temperature instead?
I will tell you a story that happened years ago, which I still remember vividly.
I once had a SoundBlaster AWE32 card, which I paid $300 for. After using it for about 3 years, my case fan became noisy. I really didn't like the noise, so I unplugged it and left the PC running without the door panel. I thought it would be fine.
Within hours, my sound card died! The case fan was directly blowing air to the bottom of the case, where the PCI slots are. Without any air, the sound card overheated and failed. It was a hot summer day. But nobody ever told me that this sound card has to have some airflow around it.
That $300 lesson taught me it is better to be safe than sorry. From then on, I prefer to keep all the components as cool as possible.
I once had a SoundBlaster AWE32 card, which I paid $300 for. After using it for about 3 years, my case fan became noisy. I really didn't like the noise, so I unplugged it and left the PC running without the door panel. I thought it would be fine.
Within hours, my sound card died! The case fan was directly blowing air to the bottom of the case, where the PCI slots are. Without any air, the sound card overheated and failed. It was a hot summer day. But nobody ever told me that this sound card has to have some airflow around it.
That $300 lesson taught me it is better to be safe than sorry. From then on, I prefer to keep all the components as cool as possible.
But... but... this is SilentPCReview, not CoolPCReview!hyperq wrote:From then on, I prefer to keep all the components as cool as possible.
I guess the trick is finding a good balance... since I really can't hear my fan at anything below 6V, I'll run it at that and watch the temps for a bit. I may add a second fan - if one at 6V is inaudible, two at 5V should be pretty silent as well.
I am in search for silence too, but not at the cost of burning my components.
You should at least have fans running at 5v. But at times, especially during gaming, GPU can become really hot. So you will then need a fan controller to turn up the speed during those times.
Recently I found SpeedFan to be a handy tool to auto-control the GPU fan. If you want to know how I programmed my GPU fan, click the link below.
http://forums.silentpcreview.com/viewtopic.php?t=25023
You should at least have fans running at 5v. But at times, especially during gaming, GPU can become really hot. So you will then need a fan controller to turn up the speed during those times.
Recently I found SpeedFan to be a handy tool to auto-control the GPU fan. If you want to know how I programmed my GPU fan, click the link below.
http://forums.silentpcreview.com/viewtopic.php?t=25023
-
- Posts: 194
- Joined: Sun Jun 19, 2005 3:10 pm
Yes, it's an nVidia card, but be advised, it's a 6200 turbocache. That means it farms out a substantial part of its workload to CPU cycles and system RAM. I wouldn't mess with the clock speed, even if you can, which I believe you can't.SoopahMan wrote:If you've got an nVidia card, the nVidia drivers can manipulate the clock speed and fan speeds for you automatically.
You can modify what fan speeds the drivers select with an additional tool, "ExperTool".
JonV, when you read a spec sheet or hear from a vendor rep about a component's maximum operating temperature, be careful not to translate that as meaning the acceptable normal operating temperature. If a component's max operating temp is 100C, that means it can survive hitting that temp briefly without an instant meltdown. That does NOT mean it will live a decent life at that temp.
The first set of temps you posted were definitely in the realm of "life-shortening" for such components. They may seem to run perfectly well for as long as half a year that way, but they are dying inside.
All applied power is converted to heat sooner or later, and that heat must be conveyed away from the components to prevent what is called a "destructive positive feedback loop", whereby the ambient temperature raises the component temp, which raises the ambient temp...until POOF. Everything cools down to room temp because it's not running any longer. Trying to operate a system fanless invites this behavior unless another substantial cooling method is at work (massive heatpipes, liquid cooling, etc).
Silence is a goal, but never at the risk of component life. In a very low-power system, the air volume inside the case may just be sufficient to dissipate the heat created by the components without needing added cooling support. Anything stronger than that should have airflow.