Gentle Typhoon 4250RPM/116.5CFM

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

BonzaiDuck

Lifer
Jun 30, 2004
15,891
1,552
126
I gave the idea some considerations w/my last build but didn't have the room in my case.
I won't be building again for quite a long while other than upgrades or repair as my attention is elsewhere (engine swap). My PC, currently a Q8400, 8800GT, six gig ram and 4 notebook drives, one an SSD for the main will meet my needs for a while.
My rig is in a case I got when win 98 came out.

from this,


to this notice the AIO watercooler, I think I did that in '05



and this the latest itineration/ incarnation

That case has been resurrected many times between those pics and I've built many others as well but I think unless someone changes the ATX format it will likely be one of my last.

I know your using the HAFs now but what happened to that server case you had? The silver one, the name slips my mind,,,

That's pretty neat. I'm even more distressed now, that I didn't keep the Giant ProLiant. Frankly, I think it saves a lot of money to resurrect the old cases. The Proliant didn't have an ATX motherboard pan: only some of the standoffs and holes fit, so I had to get the 6-32 drill-tap kit and agonize over "precise measurement."

It looks as though the small independent computer stores have all but disappeared. Back in '03 or '04, I could go down to this little place near center of town, and the owner did a double-business in recycling old parts for the rare metals and other aspects. He had a pile of old cases in the back room, and I'd pay him a couple dollars for any one I wanted. That's where I got the server case.
 

WoodButcher

Platinum Member
Mar 10, 2001
2,158
0
76
Nice job getting that much fan to a tolerable noise level.

I was hoping to use 3 of them in that last build on the radiators. I found that 3 at low speed still put out too much wind noise for my tastes. I tried several variations, only the 2 front intake, only one exhaust (side rad and PS are exhaust, the PS being a low speed Yates) I tried Adda PWM, I don't recall the spec but I think top rpm was 2500 but in all situations I had wind noise.
When I tried just the one GT for exhaust when I ran a burn test and got it to ramp up it pulled air backwards through the power supply past the yates! Brought that poor yate loon to a dead stop...o_O
Because I built the case tight all the air was channeled and even the low speed yates made wind noise. I opened the vents at the back of the case and used the slotted expansion card covers and now with the yates downvolted to 5v I get little to no sound at all.
 

BonzaiDuck

Lifer
Jun 30, 2004
15,891
1,552
126
Nice job getting that much fan to a tolerable noise level.

I was hoping to use 3 of them in that last build on the radiators. I found that 3 at low speed still put out too much wind noise for my tastes. I tried several variations, only the 2 front intake, only one exhaust (side rad and PS are exhaust, the PS being a low speed Yates) I tried Adda PWM, I don't recall the spec but I think top rpm was 2500 but in all situations I had wind noise.
When I tried just the one GT for exhaust when I ran a burn test and got it to ramp up it pulled air backwards through the power supply past the yates! Brought that poor yate loon to a dead stop...o_O
Because I built the case tight all the air was channeled and even the low speed yates made wind noise. I opened the vents at the back of the case and used the slotted expansion card covers and now with the yates downvolted to 5v I get little to no sound at all.

Interesting!! There's another thread here about someone using a Corsair Carbide case, and the issue of "PSU-orientation" came up. If I didn't pressurize my case, I would probably have to flip the PSU over to isolate its airflow.

Truth is, with the long dialog among us enthusiasts about "air" versus "Water," EVERYBODY uses "air-cooling." What you describe suggests an equal consideration to radiators using case-intake air.

Hold on a minute . . . I'll continue this by editing . . . Lemme run a little burn-test . . .

Yeah . . . I suppose I can only offer proof to myself, in that I'd had to ask the rest of the family to mute their TVs so I could hear my computer in these "subjective tests."

With a little fine-tuning to the other fans in the case, it is virtually dead-quiet up to the mid-2,000's RPMs -- 2,400 or 2,500. You begin to hear just a little white-noise, but it could be less than the AC vent in your Home Theater Room -- before you turn on your 3D Surround. You begin to hear just the faintest hum -- what's left of the AP-30 tone after the acoustic refinements.

I have it now set up to reach a top-end of ~3,600 RPM when the CPU temperature registers around 65C. At first, I can really notice the change from 2,500 to 3,600, but only because it was so quiet at the lower speed. At 3,600, it is STILL just an AC vent, the tone a little stronger but really very faint.

I think this has been a success. I will now have to post a pic of my duct "noise-cone" suspended with the popsicle sticks. And I'll probably add another dab of Pit Crews' to the mounting just to make it "more permanent."

With this -- the only single reason I'd need to use water-cooling: to get another 10C+ drop in CPU temperatures. But then -- I'm also waiting on the reviews of the Noctua D15. I won't have to change any fans! I might have to build a new duct! But -- first, the reviews . . .

Here's the result of some few runs of IBT I just made: note the maximum CPU temperatures in the right column and the fan speeds. The 1,300 RPM unit is my AKASA CPU pusher fan, and the two slower units are 200mm intakes. On the temperatures, I'm convinced that there are built-in errors to the core sensors, so the high temperature and the lowest one only demonstrate that situation. Their average almost always equals the average of the remaining two sensors. Here, the total average of maximums is about 69.5C:

Fans%20n%20Temps%203-22-14.jpg
 
Last edited:

BonzaiDuck

Lifer
Jun 30, 2004
15,891
1,552
126
Here's the snaps of the last duct-box mods.

First, construction of the "noise-cone:"

Popsicle%20sticks%20and%20cork.jpg


The box is only 6cm wide, so with a 1cm offset from the fan-side, the noise-deadener needs some space on the cooler-fin side. So I cut down the cork to about 2.5cm. [DO NOT . . . attempt to cut cork with a dremel saw wheel. Just like "kick-back" with a table saw. Use a dovetail-saw, gently . . . slowly . . (Not that any forum members are going to try this at home, but I thought I'd warn you . . )]

Another "secret." I'd promoted my project to say I'd used Spire foam rubber exclusively inside the box. In fact, I'd used vinyl foam used to pack my 780 GTX retail box. Spire . . . is better. There were two remaining sides with the vinyl, so I replaced them with Spire during installation of the popsicle-stick contraption . .

duct%20box%20and%20noise-cone%20rear%20view.JPG


I simply cut strips of Spire -- different widths -- to the circumference of the cork, and slit them in the middle for the wooden popsicle-stick brace. The ends meet at the brace on the opposite side. As much as the adhesive on Spire is so effective to be a nuisance, I added some Super Glue just to assure nothing ever comes loose to foul the fan.

duct%20box%20and%20noise-cone%20front%20view.JPG


And . . . there it is -- the finished, complete duct-box with the popsicle-stick whatchamacallit: [What DO you call it?]


AN INCREMENT IN NOISE MITIGATION

As I said, the system is now dead-quiet up to about 2,400 RPM. The only reason to "notice" the noise above that level is the difference. Above that, it seems like an AC vent's white-noise with a faint hum. The hum gets a little louder as RPMs increase, but I think I can change the fan-profile to now allow for a maximum 3,750 RPM. It's not "silent," but it's pretty darn quiet -- given the Typhoon AP-30's established reputation.

Most of the remaining tone seems to be coming from the left side-panel upper corner. I can't add Spire to the box on that side, but I could stick another slab of art-board with a Spire layer in the remaining recess to rear of the side-panel's 200mm NZXT fan:

case%20side-panel%20NZXT%20fan.JPG


I already had blocked off additional vent-holes around the NZXT, since I wanted to pressurize the case, and the art-board lends some limited noise-mitigation. I figured it wouldn't hurt to put a couple strips on the art-board, as shown. I think that extrusion or "well" in the case side-panel could allow for another strip of art-board and Spire. Couple dabs of Pit Crew's, and I won't be fouling up my case with Spire that can't easily be removed.

I'm not encouraging people to do this sort of thing -- if you're worried about noise, you might defect to the "water-cooling" side, but then, you'll still have noise from fans, radiators -- pumps.

Things can be done to deal with noise, and you can have your airflow and enjoy the reduced decibels -- too. [Now, I'm going to cut myself a piece of cake.]
 

BonzaiDuck

Lifer
Jun 30, 2004
15,891
1,552
126
A FINAL NOTE . . . [which may not be all that "final" . . . ]

I took a metric ruler to see if there was room for one more Spire pad on the side of the duct facing the side-panel. There is -- a "perfect fit."

So I did that, and decided to add another Spire pad to the remaining three sides of the duct.

Now it seems that the volume of the "tone" I hear now from the AP-30 at ~3,900 RPM is more or less the same as the volume I'd heard last week ~3,000 to 3,300.

And as I said, the truth of the matter: Any further cooling with the NH_D14 owing to greater airflow reaches a maximum between 3,200 and 3,400 RPM.

What happens with this effort to reduce noise at the top-end of the fan's output: When the computer is operating in normal usage (other than stress-testing), it is now virtually noiseless. Someone else said they can't believe it is "noiseless" at 1,500 RPM with the AP-30.

Well, I'm sorry. But at the threshold of 2,400 RPM, I can't even tell if the computer is turned on and running -- but for the case LED lights.

Eat-yur-heart-out, all you water-cooling folks! I can't reach your temperatures, but my computer is as quiet as yours is -- maybe more quiet than that!!

Maybe I'll even find better temperatures under load after building a "better" duct. Should I? Or shouldn't I? Maybe when it's raining outside, I'll pick up another panel of foam board and my Xacto knife, and try again. For $8 a package, the Spire is cheap. Why . . . I might even be able to fix it so that it doesn't look "butt-ugly!"
 

BonzaiDuck

Lifer
Jun 30, 2004
15,891
1,552
126
Here's an update based on more recent testing I've done with regard to expanding the CFM and speed ranges of this AP-30 to appease a desire for relative silence. Again, since I don't have a decibel meter, I can only describe the sounds of the fan and system at certain fan RPMs with certain analogs and "metaphors."

Many may remember the film "War Games" with Matthew Broderick, Ally Sheedy, Dabney Coleman. The Defense Department's fictional "WOPR" super-computer inside Iron Mountain in some few scenes was shown in a close-up that included a certain audio effect: a sort of low decibel, low pitched rumbling or "lub-dub-dub." This best describes what I can barely hear of my Gentle Typhoon now at 3,000 to 3,200 RPM with the acoustic enhancements.

Another metaphorical description: a sleeping infant's snore in "deep-Rem, L-Tryptophan sleep-state," accompanied by the distant, barely audible whistle of a teapot in the kitchen at the far end of your 50-room estate mansion -- perhaps separated by a closed door.

By comparison in the quiet wee-hours of late-night or early morning, with the system showing 60C under Prime 95 LFFT testing and these same speeds, my refrigerator kicked in, perhaps 25 feet away in another room. I was pretty sure sitting next to my computer that the refrigerator sounded at least three times louder -- or louder by a greater factor than that.

THE PRESCRIPTION:

-- Rubber fan-rivets
-- Approximately one layer of foam board with three layers of Spire acoustic foam rubber around the ducted fan
-- Perhaps a 10"x10" plate of foam-board plus one matching layer of Spire on the right case panel, an a similar layer secured to the case top-panel. In my circumstance, the foam board in the top-panel was already in place to block off an unused 200mm fan vent
-- It is only likely that about four circular pieces of Spire matching the fan's hub-size stacked on the fan vent and case-exterior, together with the "noise cone" I devised -- help somewhat.

Even at 3,900 RPM, you might think the noise "tolerable" given the special situation of stress-testing an overclock, but for this overclock and this D14 cooler, your optimum CFM and speed beyond which cooling effectiveness remains the same is in the 3,000 to 3,400 range. I can gauge this by comparing temperatures at maximum stress-testing with the fan at those speeds as opposed to anything closer to the 4,250 RPM top-end.

ANOTHER UPDATE: For the price, and after this little project, I decided to order some more packages of Spire pads to use later. Seemed like a good idea to keep them around, since there doesn't seem to be a limited shelf-life. They arrived today.

They are now making the Spire material thinner than 5 mm.
 
Last edited: