Thermally-controlled CPU cooling solutions
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Thermally-controlled CPU cooling solutions
What thermally-controlled CPU cooling solutions are out there? I am looking into a way to quiet a Vantec Stealth that I have on my Volcano 9 heatsink. I am too scared to put anything lower in airflow than that fan on my CPU and I am looking for a way to add thermistor control to it.
Maybe just no one in the last 10 hours felt like discussing it....give it time.Rory B. wrote:I guess there aren't any? That's a shame.
I have been trying the NMT-2 sold at SPCR-sponsor Silicon Acoustics for the last 1.5 weeks. So far so good. I use it with my Alpha 8942 with a Panaflo L1A pulling air off the heatsink, so I'm able to put the thermistor in the CPU's exhaust airflow.
Before NMT-2: (case/CPU temp) w/ambient 22-23°C
Fan @ 5V Idle = 34/35°C Load* = 36/58°C
Fan @ 12V Idle = 32/32°C Load* = 35/48°C
*Load test was Prime 95 for ~1hr.
I know that Prime95 is not the perfect stress test for a CPU, so I was a bit concerned that if I left my fan at 5V, to keep things quiet, that CPU temps could creep over 60°C, especially on a warm day if something was pushing the system.
That prompted me to investigate thermally-controlled solutions. I like the sound of my L1A @ 5V and when I looked at the thermal control graph (available as a 270k pdf from the manufacturer) for the NMT-2, I found that my idle and normal system use temps would keep the fan near 6V, so I ordered the NMT-2 (too fragile for USPS first class shipping from SA, btw).
With NMT-2:
Idle = 35/36°C Load = 35/48°C
Still very quiet. If I listened carefully, I could hear the fan was running higher if the central AC and refrigerator (next room over) were off and I was pretty near the PC (1-2m). But in normal use, my system is so quiet, my monitor is what I hear most when I sit at my desk-and that's not usually louder than the buzz of a 60W light bulb-so I have a pretty low noise level to begin with.
I also have the fan connected to my Zalman ZM-MFC1 and turned all the way up on the controller, so my voltage is probably about 11.5V if the fan maxs out. And it probably drops ~.5V off the voltage the NMT-2 is sending to the fan at the lower temps, too--thus adjusting the performance of the fan/NMT-2 in my system down ~.5V from what the manufacturer's doc shows. I have not had any problems with the fan starting up since the NMT-2 gives a 12V (11.5 in my case) burst at boot. I guess I could try lowering the fan speed even more with a tweak of the dial on the Zalman controller, but I have no need with this fan.
Other thoughts on NMT-2: If the idle temps of a computer are higher than mine, a NMT-2 equipped fan will run faster/noisier. But for my setup, I've been happy with it so far.
I have no idea how you'd place the sensor on a HSF like the SLKs, but it works great for the Alpha setup.
I also took a quick look at using an Enermax thermally-controlled fan for my CPU fan (IIRC it was based on ez2remember's recommendation). The Enermax fan would have been cheaper than the NMT-2, especially if part of another order (saving shipping), but I wanted something that would work with my Panaflo. I may yet try the NMT-2 with a Papst that I have sitting around, but I doubt I'd prefer it to the lack-of-noise my L1A makes.
Thermal control in the motherboard
I have an ASUS P4C800 motherboard, with a a technology that ASUS calls Q-Fan. You connect the fan cable to a special connector on the motherboard, and the motherboard regulates the fan speed according to how hot the CPU is (or even turns off the fan). As the temperature measured is the CPU temp, not the air outflow temp, I believe this is a quite safe and efficient solution. I have only used it for about a week, so I have not measured any temps, but it seems to work well. With a standard Intel HSF, the fan is switched off with light load.
However, I think ASUS is currently the only manufacturer with this kind of technology, so you are limited to theirs (expensive) motherboards.
However, I think ASUS is currently the only manufacturer with this kind of technology, so you are limited to theirs (expensive) motherboards.
The NMT-2 sounds like a good option for using a fan on my Volcano 9 heatsink. Now all I need to do is review the responses on my silent fan post, and I can put together a shopping list. I want to try a Noiseblocker S1 or S2 if there is a US supplier.
(edit) oops, I spoke too soon. The NMT-2 is not only out-of-stock, but the Alpha PAL8045 doesn't fit on my motherboard (an Asus A7V266) so I can't use that solution. I am really looking for a probe that can fit in between the fins on my heatsink, if one is available.
(edit) oops, I spoke too soon. The NMT-2 is not only out-of-stock, but the Alpha PAL8045 doesn't fit on my motherboard (an Asus A7V266) so I can't use that solution. I am really looking for a probe that can fit in between the fins on my heatsink, if one is available.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
Noiseblocker USA is one of the sponsors of the current SPCR Summer Promo Giveaway. Check link on that page.
Ok, I tried a test. I took the Vantec Stealth on my volcano 9 and flipped it over. The Volcano 9 heatsink features a shroud that kind of goes over the top half of the heatsink and lets the fan blow air over the very outside of the heatsink. This shroud kind of lets the air I switched the fan from blow to suck and it only raised my temps 3 degrees C, and it seems slightly quieter. The temperature rockets to 57 degrees C when the case is closed, because there is nearly no airflow with the current configuration. But I have an Asus board, and Asus boards are notorious for being inaccurate in their temperature readings. What is the transfer function of the Asus thermal probe? (That is, if Tc = actual CPU temperature and Tp = the temperature returned by the probe, and Tc = Tp + d where d = the correction factor, what does d equal?)
Now, when I run the CPU with its usual load with the fan sucking, (where Windows Media Player is playing 192kb/s files, Microsoft Outlook is open, and there are at least two Internet Explorer windows open, the CPU's temperature only gets up to 59 degrees C, which is only two degrees C above what it would get to when the fan was blowing down onto the heatsink, yet when the computer is not under load, the the temperature is 55 when the fan is sucking away and 52 when the fan is blowing normally.
Now, when I run the CPU with its usual load with the fan sucking, (where Windows Media Player is playing 192kb/s files, Microsoft Outlook is open, and there are at least two Internet Explorer windows open, the CPU's temperature only gets up to 59 degrees C, which is only two degrees C above what it would get to when the fan was blowing down onto the heatsink, yet when the computer is not under load, the the temperature is 55 when the fan is sucking away and 52 when the fan is blowing normally.
Last edited by Rory B. on Wed Jul 30, 2003 4:39 pm, edited 1 time in total.
-
- -- Vendor --
- Posts: 68
- Joined: Wed Jun 11, 2003 6:44 am
- Contact:
We are the US subsidiary of Blacknoise / Noiseblocker. Please contact us at [email protected] if you have any questions.Rory B. wrote: I want to try a Noiseblocker S1 or S2 if there is a US supplier.
You will also see a review of our products soon
-
- -- Vendor --
- Posts: 68
- Joined: Wed Jun 11, 2003 6:44 am
- Contact:
Foreseeing a sidetracking on to a Noiseblocker topic...let's try to keep this thread on thermal cooling.DryFire wrote:do you have an eta on when the site will be finished.
You may want to announce your presence/availability in the Deals & Vendors forum, too. Company reps have been pretty well received here and SPCR members really appreciate having someone to bounce product questions off of.Noiseblocker wrote:I'll say within the next month. We will make sure we keep you guys posted.
Speaking of a question...
Hey, Noiseblocker, do you suppose you could make a S2 available with a thermal probe attached for insertion between the fins of any CPU heatsink (thus providing the thermal solution I'm looking for?) I know I and many others would really appreciate it.
(edit) In fact, a model of the S2 with the thermal probe sticking out of the fan hub into the air stream would be a good case airflow solution, IMO.
(edit) In fact, a model of the S2 with the thermal probe sticking out of the fan hub into the air stream would be a good case airflow solution, IMO.
-
- -- Vendor --
- Posts: 68
- Joined: Wed Jun 11, 2003 6:44 am
- Contact:
Re: Speaking of a question...
We will take that into account for our future projects. Thanks for the input!Rory B. wrote:Hey, Noiseblocker, do you suppose you could make a S2 available with a thermal probe attached for insertion between the fins of any CPU heatsink (thus providing the thermal solution I'm looking for?) I know I and many others would really appreciate it.
(edit) In fact, a model of the S2 with the thermal probe sticking out of the fan hub into the air stream would be a good case airflow solution, IMO.
What specific model and configuration will you like?