Fanless GeForce4: victory at last!

[mcvan]

Nice going, Leo V!

The CALM, by the way, is not all it's cracked up to be. I have cooling issues with an XP1700+ in a Calm2 case under review...

 

[mcvan]

<DIV>Another quick thought about fans. Have you tried running the Panaflo at 7 or 5V? At 12V, it's not bad, but it still makes a fair amount of noise, especially high freq. At 7V, if it is inside the case, it is close to inaudible. At 5V, even as an exaust or intake fan, it's pretty much inaudible.</DIV>

<DIV></DIV>

<DIV>At 12V, the 80mm Panaflo "L" is supposed to make 21 dBA of noise. I think it is louder. But at 7V, I bet it's 15 dBA, and at 5V, it might be something like 12 dBA. Two fans at 12 dBA = 15 dBA. 2 fans at 15 dBA = 18 dBA. You get the picture? 2 undervolted fans are quieter than 1 at 12V. Two fans lets you direct the airflow more precisely. You probably even get more airflow with 2 undervolted fans...

My 2 cents...

Mike C.
www.silentpcreview.com</DIV>

 

[Leo V]

Hi Mike,
My Panaflo came from a QuietPC 230watt PSU and it's temperature-throttled; I guess that's why it's so quiet. Typically I found it much quieter than a "16dB" Adda fan. (I'm actually using a QuietPC 300watt PSU, but temp-adjust is still functional.) I'd be very interested in reading your CALM review!

5V and 7V fans definitely sound like a great idea! I think I'd do better with 2-3 super-slow fans instead of one semi-slow one. If CALM really sucks, 5V fans will just have to take its place

 

[mcvan]

Originally posted by: Leo V
Hi Mike, My Panaflo came from a QuietPC 230watt PSU and it's temperature-throttled; (I'm actually using a QuietPC 300watt PSU, but temp-adjust is still functional.)

By temperature-throttled, what do you mean? That it is still connected to the PSU? OK, and the fan speed is automatically controlled via a thermistor? Does that mean it goes up & down in speed? Or is it that the fan always runs at lower than 12V? If it doesn't change speed, then the latter is probably true. Have you tried measuring the voltage across the fan? What do you mean by "temp-adjust is still functional"?

The fact you can do cool the PC with a single low flow fan is quite amazing... unless you have something like a VIA C3 CPU?

 

[Leo V]

Does that mean it goes up & down in speed?

Precisely, so I figure it ups and downs the voltage. During a cold boot the fan is practically silent. A heat sensor secured onto the PSU heatsink is used to adjust the speed (and voltage I assume).

My CPU is an honest to god 1.466ghz AthlonXP running at 1.266mhz @1.45vcore.

 

[Phatty106]

Cheers!
Not being afraid to mod and not being complacent with your computer components is what makes being an enthusiast so fun! Don't let the man get you down!

ph

 

[Bakwetu]

Leo

Have you checked this placeno fans dedicated site?

 

[Auric]

Nice effort. But since you are not restricted to such a small fan why are you using it? A larger one will push more air with less noise. Since mounting it like that you may as well be using a 120mm. Also, was lowering the video voltage really necessary? Did you measure the temperature using the original cooling for comparision with the replacement heatsink? I would prefer using a switched auxilliary fan for gaming if necessary rather than hobbling the card.

For the record, my system has been cooled by a single quiet fan for ages -also thermal variable in the PSU but a 92mm side mount directly next to the CPU heatsink. Upgrading to a Radeon 8500 seemed to put the kibosh on that but I soon replaced its tiny HSF with an old "586" heatsink I had laying around and it has been fine at 285/330. Finger testing shows about the same heat. Original HSF's are usually so pathetic and poorly installed that I find it hard to believe that your replacement is not far outperforming and would allow overclocking if anything.

 

[Leo V]

Auric, believe it or not, the GF4 HSF is almost as good as a 486 HSF, and came well attached with thermal compound. After my mod, and before I lowered voltage, the card would exceed 80C just idling in Windows. To my knowledge, Radeons use much less power than GF4 cards. Mounting a 120mm fan will be very difficult in the same position. My noise right now is much lower than airflow in the air conditioning duct, which is itself pretty soft.

 

[mcvan]

For the record, I don't think you'll find a 120mm fan that's anywhere near the same level of quiet as your current 80mm.

 

[Yobbo]

Man, that takes some balls....I salute your machoness

 

[wfbberzerker]

should put some small heatsinks on the memory, thatd probably let you ramp up the memory clock some more.

 

[McCarthy]

Quick and quiet, very cool!



Heck I'm one of those "5 panaflos going full blast" guys more or less and my GPU fan is definately the loudest of the bunch. Well, that and the HDs.

Just too bad about PCI and AGP cards being upside down. I always preferred the ISA chips up style, never have come across an explanation of why they got flipped. In a desktop case there's no real change, in a tower it's a downer. Only upside I guess is dust doesn't settle on the chips and sinks, but when there's a fan there it ruins that. Wonder though how much easier it would be to go passive with chips up, especially on the video. Hmm, wonder if there's any reason a company couldn't do that since there's nothing to run into above the AGP card in 99.999% of systems.

Thinking aloud in text again...

--Mc

 

[Odoacer]

thats hardcore man, congrats

high five

 

[nortexoid]

looks good and everything, but why the hell did u leave the resistor leads so long?

i'd be afraid that, while bent, they touch other traces on the back of the cardd.

 

[Charles]

That's is really coo, Leo!

 

[Leo V]

nortexoid, the leads are suspended well above the board, so they don't touch anything. This was the least painful I could make the soldering process Another advantage is the resistor is harder to break off by accident, because its body rests on the board.

 

[Auric]

A larger equivalent design fan (Panaflo, Pabst, &c.) will move more air at the same dB output. True, constant speed models which run at the lower levels at default voltage may not be available but it is a simple matter of reducing their voltage to acheive it. Doing so would produce "free" (no extra noise) additional airflow. You can either get the the thermal variable models or preferably either those with RPM sensors and control them with SpeedFan or connect them through a manual variable resistor (rheostat or potentiometer) to acheive the optimal balance of airflow and noise.

Comparing the Panaflo models the 120mm produces roughly 50% more airflow for the noise than the 92mm which in turn bests the 80mm by 25%. Bigger is better. If anyone is a wiz at graphs it would be cool to plot their efficiencies for an accurate comparison. The specs can be found here.

-
P.S. if you are already using a thermal variable sensor in the PSU then obviously you do not need to get any extra stuff to use the larger fan. Being able to set the specific temperature you want to maintain with SpeedFan is optimum though good quality fans with RPM sensors are not as widely available.

 

[kami]

Very impressive. I don't mind a BIT of noise, in fact I somewhat like it. It helps me sleep at night for some reason. If the PC is off, I can't fall asleep. I made my GF2 GTS passive cooling because it was so noisy, but I don't know if I will be up to doing that for my Ti4200 which is on the way. At the very least, I will try to find another fan to replace the one on there.

 

[McCarthy]

kami, how big of a sink did you need to get your GF2 passive? I've got a GF2ti something or the other and would like to go passive on it. (won't overclock anyway, might as well quiet it down).

Leo V, you've inspired me! But, uh, I'm not doing a 6035

--Mc

 

[mcvan]

Originally posted by: Auric
A larger equivalent design fan (Panaflo, Pabst, &c.) will move more air at the same dB output. Comparing the Panaflo models the 120mm produces roughly 50% more airflow for the noise than the 92mm which in turn bests the 80mm by 25%. Bigger is better.

This is only true when looking at the raw numbers & if you need / want a higher level of airflow. The dB numbers don't always correlate to what you hear, the manufacturers don't all measure the same way, and some are overoptimistic. If what you seek is quietest first + then some airflow, the 80 -92 mm are the best. A quiet 80mm fan is inaudible at 5V. A 120mm is always audible, except when it is not running. The larger motors "hum" even at minimal speeds. So do the 80mm fan motors, but so much less. I have actually played with over 30 different fans, 60 -120mm size, trying to find the quietest ones that still effect some cooling. I use only 80-92mm fans in my systems. Haven't heard the Papst 80-120s, though I hear reports of inconsistent quality. Among 90mm fans, I think the Silencer by PC Power & Cooling at ~8V or less is silent. bigger range of choices among 80mm fans - Panaflo, Mechatronics, Silencer, etc. One final note: several 80-90mm fans at 5V will usually be quieter than one or two noiser fans, and give you about the same total overall airflow.

 

[sharkeeper]

I suppose one could exchange several thousand liters of air per minute through their case utilising 160 kHz nebulosis. Of course this is far too avant garde for atypical personal computer cooling! A NH3 absorbtion HX'r would be more appropriate! Both achieve excellent results with absolute silence!

Cheers!