Fan selection advice needed

[Gooserider]

As I discussed here, Am I estimating my airflow needs right? , I have arrived at an estimated minimum required airflow of 72.8 CFM in my new system (call it 80CFM to keep the numbers even ) I now need to pick out the best fans to use. I consider noise to be a major issue, so I want to get fans that are as quiet as possible while still moving enough air volume.

I am working on the assumtion that I will be using a total of 5 fans in the system. One is the 80mm thermally controlled fan in my PCP&C 510AX PSU. Unless I want to void the warrantee, I can't mess with that one, and don't plan to unless it really makes it necessary.

The other 4 fans I get to pick. I think they all have to be 120mm's as the 2 fan ports on the back of my AMS CK1100 Server Cube case are 120mm, and the radiator I'm using is sized to take 2 120mm's mounted in parallel.

I think the critical fans are the ones next to the rad. The rad is a FedCo 2-342 heatercore (9.5" x 6.125" X 2" core area) mounted in the lower front part of the mobo chamber, under the drive bays for that side. It is replacing a bunch of holy sheet metal and a panel of 3 generic 80mm fans. There is about 6" of free space behind it, so my plan is to put a duct on the rad to space the fans about 2" off the rad face. (I might line the duct w/ sound absorber while I'm at it, any guess as to how much good it will do? )

In order to discourage the dust bunnies, I plan to put a layer of washable HVAC filter material on top of the rad. I expect that between the rad and the filter, I will have fairly good damping on the fan noise out the front of the cabinet.

I would like to put a set of 'air shutters' on the output side of each fan so that if one is shut down or there is a big speed difference between them, the shutters will close on the slow fan and keep the fast fan from pulling backwards through it and bypassing the rad. (The air equivalent of a short circuit) For those not familiar with the term, air shutters are those louvers that you see on some building exhaust fans that automatically open and close when the fan turns on and off. Does anyone know a source for these, or should I plan on making them myself?

According to the stuff I've seen, the following statements are relevant:
1. To maximize suction, it is best to use deep (1.5" / 38mm) bodied fans.
2. Two fans in parallel will move about twice the volume at the same pressure as one fan of the same size.
3. Per some of the articles here and on Overclockers, a fan pulling through a radiator will move about half its rated free air volume.

This leads me to think that what I want for these two fans are deep bodied fans, approximately 80-90CFM capacity, with as good a Db rating as I can get for a fan like this. I also want to be able to monitor fan speed, so I see a tach wire as very important.

From looking at the mfg websites, and 'GiZzo's Fan Database' I come up with a bunch of candidates. Interestingly enough, Papst and Panaflo don't do very well on those lists. GiZzo makes a big deal about the Db/CFM ratio, and neither Papst nor Panaflo does very well in that regard.

Among the more interesting prospects:

Brand ... Model # .. ..A. RPM C.F.M. S.P. .DB. .N/R.
NMB ..4715KL 04W B10 .270 2300 83.7 .177" 37.0 2.24
Delta ..AFB1212 LE.. .190 2300 84.7 .180" 34.0 2.49
Delta ..AFB1212 ME.. .260 2300 95.0 .240" 38.0 2.50

Any suggestions / reccomendations (pro or con) among these? Or alternatives in the same CFM range?

The fans in the back I see as less of a critical issue, basically they are only needed to keep the case from being to positively pressured, and help suck the incoming air out. The fan on the PSU chamber MUST be a low profile (1" / 25mm) to avoid clearance problems with the internal drive bays. It is marginally the more important of the two since it will have to help pull air from the mobo chamber over into the PSU chamber, which won't have an intake of it's own.

I am assuming that I can use any reasonably high volume fans (approx 30-40 CFM each) to roughly match what I should be getting through the front fans. Any suggestions on these?

Sorry to be so long, Thanks in advance for any comments / suggestions.

Gooserider

[DMOS]

My questions are fairly similar.

I'm also going to be running a similar sized heatercore, with space up front for two 120/92mm fans, but I only have about 3" of clearance for the duct, if that. Where you planning on having the fans "suck" off the rad, or "blow" through it? In my current configuration I have the fans bolted to the case, and was planning on offsetting the rad from there, with the fans blowing. What is currently on there are two deafening @ 12V Delta 38mm deep 92mm fans, with the dual blades for focused flow. While that is great for the heatsinks they were on previously, is that a bad thing for the rad? Would it be better to have ones without the "tornado" set of blades? Just wondering mostly how that would affect things, or if it even would with the use of a shroud?

[Gooserider]

I now have two sets of fans, which ones I end up using will probably depend on test results.

The first pair is a set of Sanyo Denki, San Ace fans, 120mm X 38mm, 102.3CFM @ 39dBA from 12V, 0.52A.
($5.00 ea, on sale this month @ Electronic Goldmine)
The other is a set of OEM Panaflos that I got as part of a group buy from some folks on SPCR. These are rough equivalents of 120mm x 38mm L1A's (~70 CFM) but they are allegedly made w ball bearings instead of sleeve bearings. Their hubs are smaller, and the blades are different than the current stock L1A. Some folks say they are also quieter.

My rad is mounted through the case front, about 1/3 is in the case, and 2/3 sticking out. (The part that sticks out will be covered by the plastic front bezel) Both fans suck through the rad, with the bottom fan mounted horizontally on the floor of the case, blowing out the bottom of it (the case is on casters) The top fan blows into the case as it's primary air intake.

The bottom fan is at a right angle to the rad, and the bottom edge of the rad is touching the fan. The top fan is about 1.5" from the face of the rad.

Per Bill A, and some of the fan spec pages I've seen, it is better to have a fan suck than it is to have it blow, but the difference is not great. I'd set things up to suck if I could, but not worry to much about it as long as your duct is sealed up tight enough that you don't have any air leaks. Also look at if you can increase your available space by putting the rad on the outside of the case, or other such tactics. If all else fails, try just splitting the difference so that the fan is sitting in the center of the available space.

I'm not familiar with the Deltas you mention, but it does sound like there are quieter alternatives.

Hope this helps,
Gooserider

[Zhentar]

two things- first, fans are noisier when they suck. Second... Those panaflos at a full 12v aren't good enough for a SPCR person, but between undervolting and the high resistance of the rad, you'll probably end up dropping below that 80 CFM youwant. Maybe you should look into some automotive blowers?

[shiltz]

going by what you said you're looking for this may be a fan you would want to look at, http://www.directron.com/fanal12025b.html, 120MM X 25MM, Air Flow: 79.14 CFM @ 12V, Noise: <30 dBA, I use them in my watercooled dual athlon system and i'm gona be using them for the watercooling setup for my new system, they are nice and quiet with good airflow.

[Gooserider]

shiltz going by what you said you're looking for this may be a fan you would want to look at, http://www.directron.com/fanal12025b.html, 120MM X 25MM, Air Flow: 79.14 CFM @ 12V, Noise: <30 dBA, I use them in my watercooled dual athlon system and i'm gona be using them for the watercooling setup for my new system, they are nice and quiet with good airflow.

Possible, but there are two possible problems with that fan (you might want to think about the second yourself).
1. I've already got much of the duct work built, and it was designed around a 38mm fan.

2. According to everything I've read on fan selection, 25mm fans can't pull a good load against resistance. If you look at the P-Q curves on most fans you'll notice that at the same free-air CFM ratings, 25mm fans have a lower SP rating, and generally have a curve that drops much faster. Given that I'll be pulling against a heavy resistance, I want the extra Static Pressure that you get with the thicker fans. 25mm fans are good for exhaust fans and other places with low flow restriction, but they tend to loose it under load.

Zhentar two things- first, fans are noisier when they suck. Second... Those panaflos at a full 12v aren't good enough for a SPCR person, but between undervolting and the high resistance of the rad, you'll probably end up dropping below that 80 CFM youwant. Maybe you should look into some automotive blowers?

1. The 80 CFM number is based on a VERY conservative heat budget, I suspect I can get away with alot less in actual practice.

2. Keep in mind that I'll be using two fans in parallel, so each one only has to actually move 40 CFM to give me the 80CFM value. Judging by the PQ curves and the articles that BillA wrote on fan pressure drops going through a rad, I'm anticipating that I will only loose about 30% of the free air CFM, so again there should be plenty of margin. (BillA ranked on the Panaflos, but if you read the article real closely you'll note that it was only the 80mm's that he was having a real problem with. The 120mm's were OK)

4. Even if the Panaflos don't work, I'm sure the SanyoDenki's will, as they are much higher volume, and they also have good Static Pressure ratings.
5. Noise will have to make it through the rad in order to come out the front of the case - if I'd set the fans up to blow, then they would have been at the front of the case with nothing to block their radiation.
6. Automotive blowers have a right angle turn between the intake and output, which is OK for some applications, but in my setup it isn't - I need the straight through flow path you get with an axial fan.

Good comments though.

Gooserider