MadShrimps 120mm Fan comparison May2009 (30 models)

Control: management of fans, temp/rpm monitoring via soft/hardware

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
tropical_nut
Posts: 27
Joined: Sat May 02, 2009 12:04 am
Location: SE Asia

MadShrimps 120mm Fan comparison May2009 (30 models)

Post by tropical_nut » Tue May 19, 2009 7:20 pm

I noticed a new fan comparison today. Have a look here:
http://www.madshrimps.be/?action=getarticle&articID=936

In this comparison a Gentle Typhoon model, two Panaflo models, and a Sanyo Denki San Ace model seemed to fair quite well.

The big loser seemed to be a Silverstone model made by Everflow.

Meijer
Posts: 9
Joined: Thu May 14, 2009 7:33 am
Location: Netherlands

Post by Meijer » Tue May 19, 2009 9:59 pm

It's the same review as posted on the xs forums: http://www.xtremesystems.org/forums/sho ... p?t=223391

Though especially the videos are helpfull. The Gentle Typhoon seems to be a good addition to the S-Flex 1200, 1600 and 1900 RPM, and undervolted at low RPM's it is actually more quiet than a S-Flex 1600 RPM for instance undervolted to the same RPM, as far I can tell from the video's.

jmke
*Lifetime Patron*
Posts: 495
Joined: Thu Apr 15, 2004 3:53 am
Location: In Front of PC
Contact:

Post by jmke » Wed May 20, 2009 3:41 am

Martin provided me with the test data and OK'ed it for full disclosure; the sound recordings are quite useful and CFM vs RPM , CFM vs Voltage etc are quite good;

CFM vs DB are do-able if you don't take the DB at face value but only as comparison tool between the results.

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Post by Olle P » Wed May 20, 2009 5:33 am

jmke wrote:CFM vs DB are do-able if you don't take the DB at face value but only as comparison tool between the results.
Or take the time to make it comparable:
[CFM] vs 10^([dB(A)]/10)

Cheers
Olle
Last edited by Olle P on Wed May 20, 2009 6:05 am, edited 1 time in total.

jmke
*Lifetime Patron*
Posts: 495
Joined: Thu Apr 15, 2004 3:53 am
Location: In Front of PC
Contact:

Post by jmke » Wed May 20, 2009 5:34 am

Olle P wrote:
jmke wrote:CFM vs DB are do-able if you don't take the DB at face value but only as comparison tool between the results.
Or take the time to make it comparable.
Either by comparing Lg([CFM]) vs [dB(A)], which is log-log, or [CFM] vs 10^([dB(A)]/10), which is linear.

Cheers
Olle
Hey Olle, I barely made it through math class in high school; how would I best approach this in Excel? :)

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Post by Olle P » Wed May 20, 2009 6:13 am

jmke wrote:Hey Olle, I barely made it through math class in high school; how would I best approach this in Excel? :)
Thought about doing it myself, but it seems like a difficult/cumbersome task given that the numbers are presented in charts and not in tables.

I think one interesting aspect of the test is that they use a radiator to cause a restriction to the air flow.
The result is that powerful fans provide the best air flows, while low output fans (that aren't hampered much by the restriction) still achieve good flow/noise performance.

Seems like the best way to use the charts is to pick a given, fairly high, flow and then see which fan produce the least noise at that flow.
The Delta seems (disrespecting sound quality) like a great contender in that aspect.
- With the Gentle Typhoon at 12.0V it matches a Delta at 5.5V.
- The Delta at 4.0V is almost as good as the Typhoon at 7.3V, and better than any of the usual "low noise" fans.
- The Delta and Typhoon seems to match each other pretty well in terms of RPM/Flow/Noise. (For a given speed they have the same flow and noise.)

It would be very interesting to see the Papst S-Force thrown into the fray as well, with a range of 500-11,000 RPM (by PWM control) it might perform well all over the range!

Cheers
Olle
Last edited by Olle P on Wed May 20, 2009 6:34 am, edited 1 time in total.

jmke
*Lifetime Patron*
Posts: 495
Joined: Thu Apr 15, 2004 3:53 am
Location: In Front of PC
Contact:

Post by jmke » Wed May 20, 2009 6:20 am

I can provide you with the raw .xls sheet if you want, PM your email addy if you're interested

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Post by Olle P » Wed May 20, 2009 7:57 am

I've now got the data and done the maths. (Ending up with the comparison parametre of cubic feet per minute of air per microwatt per square metre of noise.)

Noticed a few things:
- As expected: For all fans you get a better flow/noise relationship with reduced voltage (until the fan stop). This is due to noise being roughly proportional to (fan speed)^5 while the flow is almost linear to the fan speed.

- Best absolute flow/noise relationship is Scythe UK1, but its flow is so low that it's just about useless. (Relation starts at 47.6 and then goes up.)
- 2nd in line is Gentle Typhoon, just a tad below at the same flow, but can speed up and reach almost four times the flow at a relationship of 19.

- 3rd place is definitely contested. It all depends on the amount of flow you want out of the fan. It's tight around the 30cfm mark, but then some are better at lower flows while other fans do better with higher flow.
- The Delta does however seem like the jack of all trades, doing well at 13.6cfm (37.46) while taking the lead somewhere around 40cfm (14) and keeping it upwards. A great choice for maximum flexibility. At 1m and 4V with the given setup the noise level is an estimated 18dB(A). Then if the air flow isn't sufficient you can, according to the measurements, increase it by a factor 8 (even though the fan speed is only three folded... :shock:). (That last part seems like a flawed measurement. Got to come back on that after closer analysis...)

Cheers
Olle

jessekopelman
Posts: 1406
Joined: Tue Feb 13, 2007 7:28 pm
Location: USA

Post by jessekopelman » Wed May 20, 2009 4:30 pm

Olle P, why not be really fancy and take into account that perceptually +10dB is generally considered 2X as loud? Useful formula: Loudness_Difference = 2^(SPL_Difference/10). With that in mind you can do some normalization and create two charts. In both you will graph CFM vs. relative Loudness. One chart will have the lowest CFM fan as the baseline and the other will have the lowest SPL fan as the baseline.

Olle P
Posts: 711
Joined: Tue Nov 04, 2008 6:03 am
Location: Sweden

Post by Olle P » Thu May 21, 2009 4:09 am

jessekopelman wrote:Olle P, why not be really fancy and take into account that perceptually +10dB is generally considered 2X as loud?
To me it seems totally unnecessary, not least because of the simple math involved. Anybody can do that translation mentally directly from the presented charts.
A 10dB difference equals a tenfold difference in absolute sound power and thus also in my calculated comparison parameter.
jessekopelman wrote:One chart will have the lowest CFM fan as the baseline and the other will have the lowest SPL fan as the baseline.
Well, the Scythe UK1 does produce the least absolute noise as well as the least airflow. It's so low that I deem it (nearly) useless. The flow is insufficient for anything but the most modest applications, and the fan is so quiet you won't hear it anyway.

Plotting the 20-40 cfm range seems to be what makes sense. Useful flow at modest noise levels.

Cheers
Olle

ekerazha
Posts: 218
Joined: Thu Oct 16, 2008 10:42 am
Location: Italy

Post by ekerazha » Sat May 30, 2009 6:47 am

I have to buy 3 fans (2 x case, 1 x megahalems cpu heatsink) and I'm really undecided between 1200 rpm S-FLEX and 1150 or 1450 GentleTyphoon.

3 x 1200 S-FLEX?
3 x 1150 GentleTyphoon? 3 x 1450 GentleTyphoon?
1 x S-FLEX for the heatsink and 2 x GentleTyphoon for the case?

:?

Post Reply