Mazing/Complicated Pathways: does it work?

Deam · Post by **Deam** » Sat May 27, 2006 7:02 am

In designing a new case, I was seriously considering using some form of complicated pathway/maze system for both the intake and exhaust in order to quite down my hard drives and the rad fans in my watercooled setup.

I am looking for advice and or opinions as to whether this technique, used in conjunction with an otherwise sealed case, has a significant impact on noise, and what it does to airflow.

My design idea was simply having a something like this:

Where the front is basicaly a U shapped airflow path, lined with foam.

Any advice would be appricated.[/img]

theyangster · Post by **theyangster** » Sat May 27, 2006 8:06 am

the design you show would work best with intake. Exhaust wouldn't be as efficient due to possible backpressure.

as long as there is sufficient room for both intake and exhaust (the pathway should be a bit bigger for exhaust), I see no no problem. Much simpler than my design

(involves "valves" and other stuff)

perhaps if you rounded out the maze, it would look nicer and be more streamlined for exhaust

Deam · Post by **Deam** » Sat May 27, 2006 10:15 am

Thanks for the response. I was definitely going to consider rounding out or modifying the exhaust side, given the reasons you note.

Post by **MikeC** » Sat May 27, 2006 11:16 am

IMO, mazes and complicated pathways don't really work. It's such a fine line between having enough airflow so that the added impedance doesn't kill cooling altogether and having too much airflow that turbulence noise is impossible to get rid of. Then there's cavity resonances in the mzes, the need to streamline to minimize turbulence.... by the time all the pros and cons are checked off, there's virtually no gain.

These comments come after building a massive PC enclosing cabinet with in/out mazes... and many other related experiments.

Today, I just go for the simplest low noise direct setup. Often, I think a caseless PC "sounds" better than any that is enclosed. If it is quiet enough, I prefer the quality of the unenclosed noise over that which comes from an enclosed cabinet or case.

Howard · Post by **Howard** » Sat May 27, 2006 11:48 am

You can line the walls with foam, Dacron, fiberglass, polyester fiber, etc. to attentuate the sound. The more you have, or the longer the pathway, the higher the absorption, especially with the higher frequency noise.

Gorsnak · Post by **Gorsnak** » Sat May 27, 2006 11:26 pm

In my opinion, there's no point in complicated pathways, but you can get some gains from some very simple sorts of things based on similar principles. For example, if you were to have an exhaust fan inset from the back of the case by a couple inches, or alternatively shrouded out the back, to focus the sound it emits to go mostly straight back, and then have a curtain or something that will act as a bit of a sound absorber a few inches back from that, you'd gain as much as from most any complicated mazed ducting, but with very little effort or downside.

Bluefront · Post by **Bluefront** » Sun May 28, 2006 1:38 am

Here's an example of a setup almost exactly like the the first drawing. The airflow enters at the bottom of the front bezel, travels upward between the bezel and the case, enters the case through the former opening to the floppy (enlarged 2x), makes a 180 degree turn at that point, then travels downward through a duct to an 80mm fan which is blowing downward into an internally mounted PSU.

The result? A PSU cooled by ambient air through a complicated path which completely mutes the sound. Works for me....

IsaacKuo · Post by **IsaacKuo** » Sun May 28, 2006 4:45 am

So far, I get my best results with simple pathways--my favorite being a U-turn so that the exhaust and intake are both to the rear.

However, based on some advice about sound absorbing "pillows", I've tried using pillows for sound absorbing with significant success. It's incredibly bulky, but I think I could use it to make a sound "maze" work pretty well.

In any case, a good rule of thumb is that airflow is restricted by the minimum cross section. Whether you fan is an intake or an exhaust, be sure that the cross section of the air pathway is never smaller than the fan itself. Sharp turns and length also cause extra resistance, of course.

Bluefront · Post by **Bluefront** » Sun May 28, 2006 5:06 am

There is a significant problem using a long duct as an exhaust, one that doesn't exist using that same size duct as an intake....it's back-pressure.

When you blow through a duct, you are actually moving a long column of air, trying to push it. And air has weight. So your exhaust fan has to work much harder/noiser. Ducting for an exhaust must be much larger to work as efficiently, and not increase the noise level.

Now if you put the exhaust fan at the end of a duct, sucking through that duct, the back pressure is eliminated, and the noise increase is lessened.

IMHO, of course.

Deam · Post by **Deam** » Sun May 28, 2006 6:42 am

theyangster wrote:the design you show would work best with intake. Exhaust wouldn't be as efficient due to possible backpressure.

as long as there is sufficient room for both intake and exhaust (the pathway should be a bit bigger for exhaust), I see no no problem. Much simpler than my design (involves "valves" and other stuff)

perhaps if you rounded out the maze, it would look nicer and be more streamlined for exhaust

My problem is that I now have 6 hard drives, and will have 8 in my new setup. They make a lot of noise combined, which is why I want to seal them up. I also will have a dual fan radiator for the watercooling. I was plannign to have my case in two independent section, the mobo with its watercooled parts, i.e. cpu, GPU, and chipset, so there really isn't any cooling necessary in that section but for maybe a small fan for minor air flow.

It was the rad/HD section that I was going to seal up and use these mazes for. While the air noise may be noisy, I guess I should've stipulated that the low sound of air isn't as bad to me as the high pitched sound of 10 hard drives. I should state that I'm not looking for total acoustic isolation, just a way to quite the drives with providing good enough air flow.

I figured that even though there may be some air turbulent/movement sounds, it cannot be worse than having the fans directly exposed, and the sound that travels through the fan holes from the interior.

IsaacKuo · Post by **IsaacKuo** » Sun May 28, 2006 7:05 am

Bluefront wrote:There is a significant problem using a long duct as an exhaust, one that doesn't exist using that same size duct as an intake....it's back-pressure.

When you blow through a duct, you are actually moving a long column of air, trying to push it. And air has weight. So your exhaust fan has to work much harder/noiser. Ducting for an exhaust must be much larger to work as efficiently, and not increase the noise level.

Now if you put the exhaust fan at the end of a duct, sucking through that duct, the back pressure is eliminated, and the noise increase is lessened.

But when you suck air through a duct, you're moving the same long column of air doing the same amount of work. Theoretically, the fan has to do the same amount of work so it's going to have to deal with the same pressure difference (assuming the same airflow rate).

All IMHO also...I haven't noticed any difference between "sucking" and "blowing" through a long cylinder but I've never made any measurements either.

Bean · Post by **Bean** » Sun May 28, 2006 7:51 am

Deam: Thats a lot a hard drives

. You definately would benefit from the maze and I'd have it's opening facing towards the rear or under the case.

In response to Isaac:
Makes sense in theory but cover the intake of the fan and it will still rotate. Cover the exhaust and it will stop rotation. From that you could draw the conclusion that a fan is less tolerant of pressure increases at the exhaust.

I wonder if in a vacuum the opposite will be true?

IsaacKuo · Post by **IsaacKuo** » Sun May 28, 2006 8:40 am

Strange, I never noticed a fan stopping when either its exhaust or intake were blocked. I just now tried with a few fans at 5v, and blocking either side of the fan didn't ever stop them. I don't have a rheostat fan controller handy at the moment, to test the very limits, though.

Regardless, this experiment can't really tell us much--if anything--about the normal situation. You're talking about blocking off airflow entirely, which is qualitatively different from merely introducing more or less airflow resistance. It's like trying to tell how fast a car can go by seeing how hard it can push against a brick wall.

Bluefront · Post by **Bluefront** » Sun May 28, 2006 8:56 am

Issac...it would be worth the experiment. But I'm not sure how to measure the result. Try sucking or blowing through a small tube (by mouth). It's easier to suck through the tube....noticeably. It's back pressure making the blowing more difficult. Same with a fan.....

Bean · Post by **Bean** » Sun May 28, 2006 9:12 am

Isaac,
Did you notice a difference in speed intake vs. exhaust blocked?

Maybe it didnt stop it entirely but I seem to remember there was a significant difference between exhaust and intake. Also, maybe I tried it at 12 volts. dont remember its been a long time ago.

I will take your results since you performed the test. So I stand corrected or diagosed myself with a shaky memory LOL.

IsaacKuo · Post by **IsaacKuo** » Sun May 28, 2006 11:00 am

I can't notice a speed difference, but that doesn't mean anything at all.

I'm just feeding the fans 5v with nothing at all to measure how fast it's going. It would take a really big speed difference to be noticeable by human eyes, and right now it's raining so it's too noisy for me to tell the difference by pitch.

When it's quiet, then it should be possible to tell the difference between faster/slower by pitch. The more rpms, the higher the pitch of the fan noise. Relatively minor differences in rpm should be detectable.

Bluefront, we humans don't use the same muscles to blow vs suck air, so I can't say the straw experiment is conclusive.

theyangster · Post by **theyangster** » Sun May 28, 2006 11:28 am

Bluefront, we humans don't use the same muscles to blow vs suck air, so I can't say the straw experiment is conclusive.

And what are you trying to suggest here?

Dream- what size and scope are we looking at here for the case?
how much are you looking to spend?

would you consider putting each HD into a aluminum enclosure? (it'd be expensive, but judging from the success from the other people who have done it,it might be worth it)

So the general consensus is that negative pressure is better, cooling/efficiency wise

what if one really would want to have an exhaust duct with the least amount of resistance, meaning no U turns just a straight duct. What if you put sound absorbing objects to break up the fan noise, with the least air turblence penalty..

I was thinking V shaped foam columns with the V poining at the fan
something like this
^^^^^^^
|.............|
|v.v.v.v.v|
|.v.v.v.v..|
|.............|
|.............|
\\\FaN////
^^^^^^^

Deam · Post by **Deam** » Sun May 28, 2006 2:06 pm

theyangster wrote:
Bluefront, we humans don't use the same muscles to blow vs suck air, so I can't say the straw experiment is conclusive.
And what are you trying to suggest here?

Dream- what size and scope are we looking at here for the case?
how much are you looking to spend?

would you consider putting each HD into a aluminum enclosure? (it'd be expensive, but judging from the success from the other people who have done it,it might be worth it)

So the general consensus is that negative pressure is better, cooling/efficiency wise

what if one really would want to have an exhaust duct with the least amount of resistance, meaning no U turns just a straight duct. What if you put sound absorbing objects to break up the fan noise, with the least air turblence penalty..

I was thinking V shaped foam columns with the V poining at the fan
something like this
^^^^^^^
|.............|
|v.v.v.v.v|
|.v.v.v.v..|
|.............|
|.............|
\\\FaN////
^^^^^^^

The case would be about 24" long, 24" tall, width is unkown yet, but wide enough to hold the HD's in that section at least, mayb3 12" wide.

That is an interesting concept, but I don't have the space for a duct. Negative pressure works better, but I have a rad that will need to bring in cool air into the case, so that is why I had negative pressure from my intake "U" turn.

As for the aluminum, that would be extraordinarily expensive for 8 hard drives. That is why I'm going for a sealed/maze design. As well, the duct issue could be problematic, as was discussed above, about backpressure and "moving air"

IsaacKuo · Post by **IsaacKuo** » Mon May 29, 2006 8:09 am

Bean wrote:Isaac,
Did you notice a difference in speed intake vs. exhaust blocked?

Maybe it didnt stop it entirely but I seem to remember there was a significant difference between exhaust and intake. Also, maybe I tried it at 12 volts. dont remember its been a long time ago.

Could someone with rpm measuring capability please try out this experiment? Basically, take a fan at 5v (or 12v) and measure the speed difference when blocking either the intake side or the exhaust side.

Because I simply can't believe my results. I'm only going by change in pitch, and I can't detect any speed difference at all! The pitch of the fan noise seems to be the same whether it's in open air or whether one side or the other is blocked! When I block the fan or remove the blockage, I don't hear any pitch difference at all of it slowing or speeding up. If there is a pitch change going on, it's completely overwhelmed by the introduction/removal of heavy turbulence noise.

Maybe the rpms remain the same no matter what, while changes in impedance merely change the amount of turbulent "recirculation". This would be like a motorboat's propeller, spinning at the same rate regardless of whether the boat is moving forward or it's anchored in place.

But I just can't believe it. Could someone with rpm measurement try this out?

In any case, blocking one side or the other of a fan definitely results in an incredible amount of extra noise (presumably due to the "turbulence").

Deam · Post by **Deam** » Tue May 30, 2006 5:44 am

Bluefront wrote:Here's an example of a setup almost exactly like the the first drawing. The airflow enters at the bottom of the front bezel, travels upward between the bezel and the case, enters the case through the former opening to the floppy (enlarged 2x), makes a 180 degree turn at that point, then travels downward through a duct to an 80mm fan which is blowing downward into an internally mounted PSU.

The result? A PSU cooled by ambient air through a complicated path which completely mutes the sound. Works for me....

What happens with the exhaust of that PSU?

Bean · Post by **Bean** » Wed May 31, 2006 5:42 am

Deam: Sorry to talk about fans again but I thought you might be interested in what I found. Specifically since you are watercooling. From my test, when I got real close to blocking the fans exhaust, they got louder, the pitch changed prob due to turbulence/cavitation like Isaac mentioned above. 1/16-1/8â€

IsaacKuo · Post by **IsaacKuo** » Wed May 31, 2006 6:18 am

The results you state are consistent with what I've found so far. That flourescent light method will be far more sensitive to small changes in rpm than hearing changes in pitch (assuming the rpms are close to a harmonic of the flourescent light frequency--otherwise it's just a blur).

I should hope that there's an effect on the level of airflow, of course! It would be very bizarre if a fan which has its intake or exhaust blocked off generates as much airflow as if in open air!

Anyway, I think it would be more useful and informative to experiment with the effect of fan placement in relationship to a long muffler tube. I'll start a new thread on that, I guess. The thing is, I don't have any means to measure the airflow. So really, I'll only be able to notice changes in noise level/quality. In any case, I've wanted to experiment with long tube fan muffling for some time.

Deam · Post by **Deam** » Sun Jun 04, 2006 1:37 pm

As a follow up, what type of foam is recommend for sound dampening/absoprtion purposes?

Bluefront · Post by **Bluefront** » Mon Jun 05, 2006 1:30 am

To answer Deam's question about the PSU exhaust in my previous picture.....it goes into the case. This is a highly efficient Fortron PSU. It runs cool, plus this is a low-heat setup. The temp of the PSU exhaust is usually around 30C.

Take a look at this picture....

There's an intake fan behind that rear muffler/duct. The intake noises are deflected downward.....much quieter. If you try to make an exhaust duct with such a sharp bend, the resulting back-pressure will severely restrict airflow, and make everything hotter and noisier.

As an intake, that muffler/duct does not severely affect airflow.....but it does make everything quieter.