Fan Roundup Worklog

Vapor · 12-16-2006, 03:24 PM

Okay, since the roundup is already in swing, I decided to make this thread.

I'll post some pictures eventually I guess....not in a picture taking mood right now, besides, if I throw all my work up on the first post, it wouldn't be much of a log, eh?

I'll be posting initial impressions of new fans in this thread, as well as explain some details of the process, and of course post results to this thread early

Anyway, CFM testing is a lot of fun suprisingly....with my airflow meter, it's just a simple calculation to get from FPM (feet per minute [velocity]) to CFM (cubic feet per minute [volume, which we want]), best of all my airflow meter can do it internally, no need for me to calculate anything

As many of you also know, I'm testing a restricted CFM value, where I take a medium-restriction radiator (Swiftech MCR120) and mount it flush against the exhaust of the fan, and also mount it to the CFM meter. This is one measurement that will aid us in determining which fans will be good for cooling on a HSF or radiator. How? If a fan loses a lot of its air-pushing capabilities when faced with some sort of restriction, especially to the point where it does worse (at the same noise level) than other fans, why use it? Now, due to other factors, like hub size, that won't tell the whole picture for how it actually cools on a radiator or HSF.

That's why real-world testing will also be performed. I've chosen a ThermalTake Big Typhoon as my HSF, due to it's ability to cool well with low CFM, yet also scale with high CFM. On top of that, it can mount any thickness of fans

Why not ONLY do real-world testing? Well, there are many instances where thickness is a non-issue for a user, in instances like this, hub size means very, very little because a shroud may be used, meaning that only CFM w/ restriction is the measurement they need.

I'll be testing on a yet-to-be-determined motherboard and CPU, but it will be Intel based, maybe C2D based....just have to decide which board I like better for 24/7, my P5B Deluxe, or the XBX2 Intel is finally sending to me from an ancient RMA. And whether or not my Celly D can produce that 'ideal' amount of heat where it shows improvement at better cooling, but also run at 90C with a no-CFM fan

Testing will be conducted at various voltage levels (12V, 9V, 7V, 5V, minimum) and I will report all RPMs (when possible, not all fans have RPM reads), CFM readings, dB readings, CPU temp readings, and considering reverse-CFM.

Reverse-CFM would be: how much noise/volts/RPMs do I have to make in order to get a certain CFM level? The 'goal' CFM level would vary per fan class, and I could also include a restricted CFM level.

For instance, for high speed fans, say you only want 100CFM free-air, well, I would test to see what kind of voltage/noise/RPMs would be needed, helping you decide which fan is best for your goals. Likewise, I could determine what fan is best for ~66CFM on the radiator.

Speaking of fan classes, here are the 7 categories, the names are just titles, if you don't like them, please ignore

Silent: <801RPM
Quieter: 801-1199RPM
Quiet: 1200-1599RPM
Medium: 1600-1999RPM
Loud: 2000-2399RPM
Very loud: 2400-2799RPM
Xtreme: >2799RPM

When I test fans, I'll also list, based on voltage response, which categories they 'spread' over, for quick reference.

For sound measurements, I've played with the idea of recording, but I'm just not into it (to be honest, I'm more into how fans PERFORM, not how their sound intricacies vary)....due to the fact all fan types will be tested, not just quiet fans. As such, I'll only be doing dB readings and subjective noise. If there's grinding, you'll know about it, if it's inaudible from more than a foot or two away, you'll know about it. I won't let the silent guys down much, but fwiw, this is for a whole-market roundup....something that's yet to be done anywhere, and with various techniques that have yet to be done anywhere.

I do plan to review as many fans as possible, but I'll be honest, everything's come out of my own pocket so far: 9 fans, multiple hundreds of dollars in CFM/dB equipment, passive fan powersupply, HSF for real world testing, etc., so if things slow down for awhile, you'll know why....a combination of me not able to get review/test samples and not being able to pay my own way (I am a student after all, lol).

Anyway, as always, any and all input is appreciated.

(oh, and I'll be copying and pasting results over from the other thread and locking that one down

)

Vapor · 12-16-2006, 03:31 PM

Here's the CFM results at 12V and 7V of all the fans I've tested so far

Saved for later

Vapor · 12-16-2006, 03:54 PM

Playing some more....and was wondering if anybody has experienced this with their Yate Loons.....

They seem to have a bit of a break-in period. I'm testing today, and my Yate is already pushing 52CFM, and has only been used for ~20 minutes. Gonna run it some more and see if it breaks in more

If it does, then it looks like all new fans will go through a mandatory breakin period before testing.

EDIT: Actually, it just exposed a slight flaw in an assumption I made....all previous results are invalidated unfortunately

Good news is that I've double checked everything and now it's a-okay.....guess I just got a little excited when I was first testing and missed a glaring error

(oh, and in case you're wondering, all volts are measured at the end of the cabling nearest the fan

....factoring in droop that way

)

Bbq · 12-16-2006, 06:41 PM

Minor typo: Silent should be <800, not <801.

BobMaximus · 12-16-2006, 06:53 PM

Sounds good :thumbsup: Make sure to state your opinion of each fan, and maybe which one you would have in your rig if you were to choose.

I'll be watching this thread for sure.

Vapor · 12-16-2006, 08:00 PM

Quote:

Originally Posted by Bbq

Minor typo: Silent should be <800, not <801.

No typo, I changed it since I could only find one fan that had a stock speed of <800RPM and figured since 1 was a lonely number, to expand the category

Anyway, that brings up a point I have.

Lots of fans don't have RPM reads....obviously, I could just guess what RPMs they go at.....but I don't like that.

For categories, how about I change it to CFM levels?

<30CFM: Cat1
30.01-50CFM: Cat2
50.01-70CFM: Cat3
70.01-90CFM: Cat4
90.01CFM+: Cat5

BTW, just finished CFM testing all 9 fans again at 1V intervals and with minimum working voltage....both open air and restricted. Will post results soonish.

Praxis1452 · 12-16-2006, 08:07 PM

couldn't you setup something like this?

laser ---fan---reader. However many times it hits divided by the number of spaces in a minute would be the rpm?

Vapor · 12-16-2006, 08:12 PM

No desire to get a laser RPM reader....RPM readings are probably the least important statistic anyway. Besides, spending more money on test equipment is not my desire.

EDIT: basic graphs...
Yate Loon SL:

Thermaltake Silent Wheel 130mm (VERY high loss%), well below CFM rating as well:

Thermaltake BTvx Stock Fan (no, there is no error....it really does scale like that). ~9.2V is what the included rheostat does for a minimum, so I included those numbers:

YS-Tech "108CFM" 120x120x25 fan....

Scythe Infinity stock fan:

Panaflo L1BX (all the Panaflos had the lowest loss%, as expected...but also all output below rated CFM values):

Panaflo M1A:

Panaflo H1A:

Panflo U1A:

Praxis1452 · 12-16-2006, 10:11 PM

how many more fans you got?

And how much would the laser reader cost? well w/e

Vapor · 12-17-2006, 12:14 AM

In my already-owned/just bought batch, 0.

How many fans do I have left that I plan to review? IDK, however many there are on the market minus 9.

Turns out only three of my fans (not 4 like I thought) don't have RPM reads, and two are Panaflos (meaning with interpolation and extrapolation, you can figure out their RPMs easily since two same-body counterparts do have RPM reads), and the other isn't made any more (meaning you can't buy it anyway

). Considering that, the laser and reader, if more $10, are just not worth the money or time to research/use them. I'm trusting that all 'new' fans will have RPM reads.

Anyway, just got all dB and RPM data on those 9 fans.

Looking at the RPM data, I find nothing useful, other than Yate Loon's CFM/RPM is pitifully low, not sure if that's at all useful though, since RPMs, just don't mean much, unless I'm missing something.

EDIT: looking at the dB data, it seems my dB meter does some sort of mode-switching between 46.7dB and 48dB as results get very weird there...results are normal both before and after.

Also, the way I'm graphing it, it DEFINITELY shows that buying a fan suited for your desired CFM is best....don't buy a high CFM fan if you want medium or low!!

Also, for restricted CFM categories....just use the same data bracketing as from open air? Or should they have their own categories? NVM, they'll have their own categories, looking at the data any other way makes little sense unfortunately.

<20CFM: Cat1
20.01-32.5CFM: Cat2
32.51-45CFM: Cat3
45.01-57.5CFM: Cat4
57.51CFM+: Cat5

Lenny_Nero · 12-17-2006, 03:23 PM

Nice, I have just had a quick look in here, as I am not going to use any of the old fans I have around and will be giving my new box a bunch of new ones.
Will you (or have you) tested any fan controllers ?

as for...

Quote:

RPM readings are probably the least important statistic anyway.

I have always just turned fans down 'till I cant hear them, the speed does not really matter if you know how much air it can shift v's the sound given off.

Vapor · 12-17-2006, 09:13 PM

Hmmm, small update: turns out there are a few fans that do hold true to their manufacturer dB rating:

Yate Loon SL
Panflo H1__
Panaflo U1

and that's about it....L1/M1/all-Thermaltake/Scythe* are off

*not actually sure about the Scythe since I can't find ratings, but I have a feeling it'll say ~24dB when really it's 28dB, just like the Yate.

Vapor · 12-20-2006, 07:59 AM

Okay.....input needed.

I have a TTBT and a Scythe Infinity here...and I have no preference for which goes into my 24/7 system (mainly because I'll be using my phase again once my vaca ends). I was originally planning to use the TTBT for testing, because it scales well with better fannage. BUT....I just did some testing, and its restriction is VERY high (~45% on a Panaflo [~50% on 25mm thick fans]), which means that lower speed fans will just fail. However, this also means that it will show a larger difference between fans.....

The main reason why I got it was because it could mount any thickness of fannage. Well, I just figured out a way with the stock Infinity fan clips to install any type of fan as well (two at that....heh).

So the decision is yours: Infinity or TTBT?

Infinity will show that low speed fans are actually better than it seems....while the TTBT will show larger differences in results.

Hmmmm, I guess I won't be using a Celeron D....the full-load difference between a U1A+Infinity and passive Infinity at 1.7V is ~7C.

EDIT: ah there we go....got it to crash from heat passively cooled.

Guess I can use this setup afterall

enz660 · 12-20-2006, 01:11 PM

Quote:

Originally Posted by Vapor

Okay.....input needed.

I have a TTBT and a Scythe Infinity here...and I have no preference for which goes into my 24/7 system (mainly because I'll be using my phase again once my vaca ends). I was originally planning to use the TTBT for testing, because it scales well with better fannage. BUT....I just did some testing, and its restriction is VERY high (~45% on a Panaflo [~50% on 25mm thick fans]), which means that lower speed fans will just fail. However, this also means that it will show a larger difference between fans.....

The main reason why I got it was because it could mount any thickness of fannage. Well, I just figured out a way with the stock Infinity fan clips to install any type of fan as well (two at that....heh).

So the decision is yours: Infinity or TTBT?

Infinity will show that low speed fans are actually better than it seems....while the TTBT will show larger differences in results.

Hmmmm, I guess I won't be using a Celeron D....the full-load difference between a U1A+Infinity and passive Infinity at 1.7V is ~7C.

EDIT: ah there we go....got it to crash from heat passively cooled.

Guess I can use this setup afterall

I like the Infinity more...(im going to buy one) but if it showes a large difference, go with the TTBT. Also when you get to the higher cfm fans with the Infinity, I doubt you will see more than 1c difference between the fans. I really like the reverse cfm idea becuase I was thinking of getting a high powered fan and turning it down but I want it to be silent at low rpms. Another thing...can you test this fan? This roundup is the best, keep up the good work :thumbsup:

serialk11r · 12-21-2006, 08:48 PM

OMFG...The thermaltake fans get down to 50 cfm @ 9 V??? Oh crap, I better just switch it off to a yl... VAPOR THANK YOU SO MUCH!!! I almost bought a thermaltake fan with an rpm monitor so that I could have good flow at the back of the case. Running it at 10V seems like a bit too much for me. When are the dba measurements coming? I need to order my fans tomorrow!!!

Vapor · 12-22-2006, 12:09 AM

I have all the dB measurements....but I'm on vaca and don't have all the graphs situated yet....PM me if you have any questions about the 9 fans above.

I can tell you, that a quick glance at the data says buy fans that are designed for your desired CFM. Don't buy high and run low, you'll be making more noise than you need to.

Vapor · 01-05-2007, 09:44 AM

Hmmmm, okay, question's up again.

TTBT or Infinity?

I'm leaning toward TTBT, but just a foreword with it: it will make slow fans look bad

Vapor · 01-05-2007, 09:48 AM

Ah screw it, executive decision: review gets TTBT, I want the Infinty for my quiet system

xlink · 01-05-2007, 01:00 PM

take the infinity and mount 2 fans on it...

should beat up the typhoon...

Lenny_Nero · 01-05-2007, 05:00 PM

Will you be testing the fan that comes in the Zalman ZM-F3 kit ???
I will try to find out the full name/builder of them
As I have just started up a new build with them as they were cheap and came with the rubber mounting kits and 7volt mode and leads, but so far they make almost no noise at 6~9 volt and I can just about hear them at full 12 volt with my ear near and this is mounted and working thru grills, not just running free.

I am impressed, very impressed with then so far

Vapor · 01-18-2007, 05:22 PM

Okay, bit of an update....just got a bunch of fans in today on loan from a forumer (not sure if he would like to be named).....one of them was a very special, very rare Delta EFB1324SHE.

What does that mean? Well, it's an EFB series (meaning standard three-blade design), but looks like an FFB, meaning it has 9 blades, 38mm thick, and a counter-angled diffuser at the exhaust side. 13 means it's 130mm (even though it's not....this also confused me, tbh). 24 means it's a 24V fan. SHE means it's a Super High Speed fan....somewhere among the EHE and GHE for speed.

Well, if the voltage info is true about the fan, I'm in awe. First, I don't have a 24V PSU, but with testing at only 12V, it put out 171CFM. It also draws 1.38A. That means at its rated voltage of 24V, you can expect it to put out over 300CFM and use 33W.

On top of that, it also had the best-as-tested pressure in a landslide.....

Anyway, thought I'd update everyone that things are coming along nicely and that the initial roundup will be very sizable (about 2 dozen fans of various speed/noise/thickness) and should hopefully be complete in about 10 days.

EDIT: well, researched that fan up....200CFM, only .92A....must be mislabeled.....oh well

EDIT2: also, I tested an EFB1212LE (rated for 2000RPM at ~75CFM) and my CFM meter COMPLETELY bugs out. First, the blades that measure airflow, SPIN THE WRONG WAY, second, they spin only fast enough for ~10CFM....Well, when I put a restriction on the airflow, suddenly it spins the right direction and spins for a respectable and expectable 43CFM

Praxis1452 · 01-18-2007, 07:07 PM

pics of that odd delta? loudness hehe

Vapor · 01-18-2007, 07:36 PM

Looks exactly like this: http://www.sidewindercomputers.com/delffb1212eh.html

Just the sticker says EFB1324SHE.....

In fact, I bet that's the fan that it really is (the one at sidewinder)....and yes, it is very loud

Spawne32 · 01-18-2007, 07:46 PM

Hows this thread coming, cus i still dont see a list yet int he first post

Vapor · 01-18-2007, 08:02 PM

I have a new system, ignore most of the first post

You'll see what it is.....very useful way to graph the data, IMO