KRONOS Fan Information

[curvebender]

Aribo,

After Maphill mounted his Silenx, he only measured a few degrees of difference in temperature on the panel, compared to when the original fan was in there.

How can we explain this?..

[popmann]

Mine has definitely gotten louder than it was when I purchased it a month ago. Wtf? This might explain the variance. If mines gotten noticeably louder can only mean the fan they used is really not a precision piece.

[aribo61]

After Maphill mounted his Silenx, he only measured a few degrees of difference in temperature on the panel, compared to when the original fan was in there.

How can we explain this?..

Well, I don't know.
Perhaps it's K73 vs. K61?
Also, maphill has a ixp-34-**16**
Perhaps he didn't stress the machine like I did.
And if not stressing you have to use the Kronos several hours

I now mounted the 80mm fan at the lowest speed I could get with the fan mate 2 and it's definitely cooler than the Silenx at 12V.
It's the same with the noiseblocker fan.

[curvebender]

Aribo,

I believe Maphill has the 34-12. (I have the 34-16.)

But yes, maybe the casing of the 73 is cooler than the 61 as it's bigger.

The quest continues!!

[aribo61]

An infrared thermometer is a awesome tool. About $50 US and can be used to do car work, check for fevers, etc.

...

Here is a page that shows a simple and cheap way to measure airflow. Perhaps we should try this for verifying airflow instead of how it feels?

I generally like that you want to get hard facts instead of a "feeling".

I also thought about buying an infrared thermometer.

But for those who don't have such a device it won't help to describe it that way. So I think we need both.

In the case of the fan throughput, especially of the Silenx, it's so obvious, that I don't really need an exact measurement.

But I understand, that you want a better measurement, which makes the values comparable and reproducible.

I personally test the fan (and it's mounting!) by installing it into my Kronos, that is in real live. I have the fan mate 2 outside of the case, so I then I look how high I have to adjust the speed for a working cooling.

In the end it's an optimization problem.
You have to weight the noise against the resulting heat.
And the noise is a subjective criteria anyway.
So I think everybody has to find it's own best solution.

[aribo61]

Here is a page that shows a simple and cheap way to measure airflow. Perhaps we should try this for verifying airflow instead of how it feels?

http://www.bit-tech.net/hardware/coolin ... case-fan/1

ok, I measured all values with the suggested method, though I don't know how much volume my "black sack" (which is blue in my case ) has. I guess it has about 60-100 litre (it's the biggest one I could find at home).

I measured the following values:

Code: Select all

Fan                    12V      9V (motherboard)

Kronos-Standard        23s      60s
Silenx IXP-34-12     >240s      --
  about half filled  >120s    >180s
Noiseblocker XR2       50s      50s
Arctic F8 (80mm)       15s      20s

The Silenx wasn't able to fill the sack completely, so I used a value taken when it was "looking" full. I also took another value (in a different run) at a time where the sack was looking half filled (which should be very inaccurate).

Also note this paragraph taken from maphill's link which tells us something about this problem:

"Also, even though we used a black sack made from lightweight plastic, it was still too heavy to fill completely for the fans that have a particularly low airflow or low static air pressure, so we were unable to measure their airflow accurately. The lowest-performance fan that completely filled the sack was the Revoltec RL037, which has a calculated airflow of 7cfm; therefore, any fan that couldn’t fill the bag provided a level of airflow that was less than this."

As you see, the values are very very disappointing.

When measuring this way you get the feeling the Silenx wouldn't blow any air into the sack. With the standard Kronos fan you see good progress from the beginning.
Also, you can see that the difference between 9V and 12V can be very different for each model. So the specified CFM-value doesn't really matter. Instead we should look at the 9V value (which unfortunately isn't specified).

As a conclusion the Silenx seems to be much worse in effect than I measured by "feel". I think that's because I really wanted to use this fan, because it is so silent.

Now I'm waiting for maphills measurements...
(ok, you don't need to measure, if you do believe my values)

[curvebender]

Aribo, you convinced me to remove the Silenx from my Kronos and put back the stock fan. Your findings are alarming to say the least!

Anyway, for now, I am hoping that mounting the original fan on rubber mountings will reduce the hum a little bit at least..

[maphill]

The "16" version of the SilenX claims 18, over the 14 of the "12". We may not believe the claims of the manufacturer, but lets say the ratio is at least correct for argument's sake.

Then the 16 is 28% more than the 12. According to Aribo's measurements, this would still put it at 64% of the stock fan's air flow, which seems too low, but it would depend on the safety margin factored in to the stock fan.

Hmm, perhaps I'll try the stock fan with the new "closed-chassis" rubber mounts I have.

Another possibility would be 2 SilenX "16" fans mounted in series.

Aribo, do you have a picture of how you've mounted the 80mm fan? I assume it is at an angle a bit?

Thanks,
Mark

[aribo61]

Anyway, for now, I am hoping that mounting the original fan on rubber mountings will reduce the hum a little bit at least..

before doing this you might better buy the Noiseblocker fan, which has enough throughput at 9V and is much more silent than the standard fan.
I tested it some time ago, and found it to be good, but I wanted to have an even more silent result and the Silenx seemed to fulfill this aim.

[aribo61]

first I edited the first line of my table, it should contain these values:

Kronos-Standard 12V: 23s 9V: 60s

The "16" version of the SilenX claims 18, over the 14 of the "12". We may not believe the claims of the manufacturer, but lets say the ratio is at least correct for argument's sake.

I agree

Then the 16 is 28% more than the 12

16/12 = 1.333, so it should be 33%

According to Aribo's measurements, this would still put it at 64% of the stock fan's air flow

hmm, how did you calculate this? even with my wrong values I can't see how.

[edit: ok, I understand, you took the 7 cfm from the citation, right? which is 50% of 14cfm of the stock fan and 50% * 1.28 = 64% -- but the value from the citation only marks the upper end, the Silenx has a much lower cfm, if the stock fan has the right value, the Silenx should have 17% * 14cfm = 2.4cfm which is very low]

With 60s at 9V for the stock fan and 240s for the Silenx at 12V under consideration of the relation for "half full" 12V/9V -> 120s/180s, I get

240s / 120s * 180s = 360s for 9V

(I took the "half full" values to be able to calculate this)

Now 60s / 360s is about 17%

and 60s / 360s * 1.33 (33% more) is about 22%

or using another way to calculate:

360s / 16cfm * 12cfm = 270s for the ixp-34-16 at 9V

then 60s / 270s = 22% (same value as above)

22% *is* way to low, isn't it?

Another possibility would be 2 SilenX "16" fans mounted in series.

I think this doesn't change much. According e.g. to this posting:
http://www.overclock.net/air-cooling/64 ... ost8172647
It wouldn't change the CFM but only static pressure (whatever that means, but I think we need more CFM).

more details can be found here:

http://www.tomshardware.co.uk/forum/162 ... s-parallel
http://www.comairrotron.com/engineering_notes_02.asp

Aribo, do you have a picture of how you've mounted the 80mm fan? I assume it is at an angle a bit?

no, sorry, I dismounted it already, because it didn't cool enough at the lowest speed and at higher speeds it was a bit too loud for my taste.
But this one (arctic F8) was cheap, I think there could be better 80mm fans for this purpose.

No, it wasn't mounted at an angle. I mounted it with my cable tie method at the position of the stock fan, but with more distance to the metal plate.

Btw. I took a closer look at the way the cooling may work.
I mounted it to blow cool air to the motherboard (stock fan blows the other way). I think this was a bit suboptimal. Some of the warm air will be pushed out above the keyboard and through the switches on the top cover. I assume, if the fan is directed the other way, cool air from outside will be drawn through the holes directly to the motherboard cooler.
I think this probably has a better cooling effect.

[Dniss]

Didn't hear the fan until I read this thread...



Now it's the only thing I hear.

[maphill]

16/12 = 1.333, so it should be 33%

The exp-34-12 is rated for 14 cfm.
The exp-34-16 is rated for 18 cfm.

18/14 = 1.28 = 28% more.

But then I took the wrong number from your graph for my further math, so it is irrelevant.

Mark

[aribo61]

From my current experiments, the 80mm Arctic F8 fan is very silent with an open bottom cover. But if I add the cover, it gets a bit loud (hum, I think only for my taste).

So it depends very much on the kind of mounting. The mounting with the cable ties I talked about in my last posts was suboptimal, because I only had two kinds of them, very thin and very thick ones. As the thin ones were too flimsy, I used the thick ones, which didn't decouple enough of the vibration.

So, I think I have to find a better size of cable ties or a different mounting solution.

One method coming to mind is using spring wire in a suitable form.
Or just some kind of springs.

Btw. the 80mm fan makes a nice low "windy" sound.

[aribo61]

16/12 = 1.333, so it should be 33%

The exp-34-12 is rated for 14 cfm.
The exp-34-16 is rated for 18 cfm.

ok, I'm an idiot

[curvebender]

Well, I put the original fan back, but with rubber mountings.

Hopefully, one day we'll find the ultimate ratio between noise and performance in a fan.

Have to say, I'm tempted to drill vent holes in the K!