Dissatisfied With Current AIO - Buy a different one or go Air?

[Vya Domus]

Tried that, but it didn't help. I even tried removing the second fan altogether and just running the center fan and confirmed for me that it wasn't phasing resonance between the two fans for the AK620, but between the poorly-balanced fans running through the resonant frequency of the heatsink itself.

The heatpipes attaching each fin stack to the clamped baseplate have their own oscillation harmonics and unless you run a manual, fixed fan speed that doesn't pass through the troublesome RPM range, you'll always hear it as the fan speed varies up and down through the speed range.

It would probably be fine if Deepcool's AK-series fans weren't poorly-balanced, or if the heatpipes were staggered rather than in a neat line in the same plane as the oscillation of the imbalanced fans. I wanted to like the AK620 because it looks and feels like a good bit of hardware - the fans certainly feel premium enough, but their lack of balance was really very noticeable, something I discovered when re-using the AK620's fans on some Thermalright SI-100 coolers.

So it was resonating with the heatsink itself not amongst each other, that's different, I guess there is not much to do with that other than try to find an RPM where it doesn't happen.

Actually what usually happens with fans isn't even really resonance, what happens is that that because the fans all run at the same RPM the frequencies they emit overlap and the sound adds up in amplitude and appears louder.

 

[freeagent]

I don't know if it's the same company as I used to be but prior to the last few years, Thermalright used to be very high-end air cooling that I last used in the mid 00's (Ultra 120 and XP-90).

Yup same company

 

[Shrek]

Actually what usually happens with fans isn't even really resonance, what happens is that that because the fans all run at the same RPM the frequencies they emit overlap and the sound adds up in amplitude and appears louder.

I belive the problem is beats, the fans are of slightly different speed so sometimes their sound/vibration is aligned, sometimes anti-aligned
Beats - Physics (physics-and-radio-electronics.com)

 

[Chrispy_]

So it was resonating with the heatsink itself not amongst each other, that's different, I guess there is not much to do with that other than try to find an RPM where it doesn't happen.

Actually what usually happens with fans isn't even really resonance, what happens is that that because the fans all run at the same RPM the frequencies they emit overlap and the sound adds up in amplitude and appears louder.

Yes, with just one Deepcool fan it was classic sprung-mass resonance; The spring is the small amount of flex in the heatpipes. You could feel the whole case shaking because the whole finstack was probably resonating. There's no way that much shaking could have come from the weight imbalance of the fan blade alone.

It's likely bad with the AK620 and not other towers with imbalanced fans because all of the heatpipes are aligned perpendicular to the plane of the oscillations from the fan (blue arrow); That means that each heatpipe has the same resonant frequency to oscillations in that plane as the other heatpipes, and the moment of the forces in that plane are uncountered by heatpipes in any other planes:

There are lots of different resonance possibilities for multi-fan systems and you're correct in that the thing you describe with two fans of nearly the same speed isn't resonance - it's low-frequency phasing between constructive and destructive interference.

The worst resonances IMO are when you have vibrations from imbalanced fan(s) that match side panels. Suddenly your whole side panel is acting like a buzzer... :\

I belive the problem is beats, the fans are of slightly different speed so sometimes their sound is aligned, sometimes anti-aligned
Beats - Physics (physics-and-radio-electronics.com)

Huh, whaddayaknow... phasing interference is called "beats".
Confusing to use a word that's associated with something else frequently used for sound waves, but less of a mouthful, at least...

 

[Shrek]

I'm still for vertical air so the VRMs get some flow; I am concerned that while a water-block (or horizontal air) is great for the CPU it leaves the VRMs to bake.

 

[FoulOnWhite]

I'm still for vertical air so the VRMs get some flow; I am concerned that while a water-block (or horizontal air) is great for the CPU it leaves the VRMs to bake.

I'm using positive air with my water cooling so i hope it helps with the VRM cooling

 

[Pictus]

I'm still for vertical air so the VRMs get some flow; I am concerned that while a water-block (or horizontal air) is great for the CPU it leaves the VRMs to bake.

Jiushark JF13K Diamond, good airflow for the RAM too.

 

[P4-630]

@Gmr_Chick did you buy a new (air) cooler yet?

 

[Chrispy_]

@Gmr_Chick did you buy a new (air) cooler yet?

Yes, post #99

Now that this thread is solved we're all going off-topic because that's what forum threads are for

I'm still for vertical air so the VRMs get some flow; I am concerned that while a water-block (or horizontal air) is great for the CPU it leaves the VRMs to bake.

I borrowed stole one of the SI-100's from work, because it's doing the VRM cooling in an otherwise passively-cooled case.

Thermalright, impressive quality for not much money - a very flat, nicely milled baseplate (yes, a baseplate, not some cheap direct-contact cop-out), plenty of surface area for the typical ~150W CPU and a fan that I personally dislike and replaced, but is actually decent. It's not a bad fan, but it's a bad pairing for a cooler that clearly isn't going into a bone-stock ATX build. For the price, I can't complain because the heatsink alone was worth the full asking price.

Going off-off-topic, Thermalright are a company I ignored for 15 years and have recently (post-COVID, post-Brexit) become a very appealing prospect in the UK. I am building so many PCs that I always know what the best bang-for-buck air cooler is and for a while that was Cryorig, then it was ID-Cooling. Both of those companies seem to have doubled their price whilst still being low-budget build-quality. Meanwhile, Thermalright have flooded Amazon.co.uk at rock-bottom pricing.

https://www.amazon.co.uk/Thermalright-Assassin-Cooler-Heatpipes-TL-C12C/dp/B0B4JCWHVG

£16 including tax is effectively a $17 cooler which is insane for something I can get delivered in under 12 hours. It's nothing special for a $40 air cooler but it's not $40, it's way less than half of that.

Shit, now I look like a Thermalright shill....
"Hey Thermalright, I'll shill for you for the right price and this one's on the house."

 

[AusWolf]

I'm still for vertical air so the VRMs get some flow; I am concerned that while a water-block (or horizontal air) is great for the CPU it leaves the VRMs to bake.

I'd love to see reviews on this matter.

 

[Space Lynx]

Tried that, but it didn't help. I even tried removing the second fan altogether and just running the center fan and confirmed for me that it wasn't phasing resonance between the two fans for the AK620, but between the poorly-balanced fans running through the resonant frequency of the heatsink itself.

The heatpipes attaching each fin stack to the clamped baseplate have their own oscillation harmonics and unless you run a manual, fixed fan speed that doesn't pass through the troublesome RPM range, you'll always hear it as the fan speed varies up and down through the speed range.

It would probably be fine if Deepcool's AK-series fans weren't poorly-balanced, or if the heatpipes were staggered rather than in a neat line in the same plane as the oscillation of the imbalanced fans. I wanted to like the AK620 because it looks and feels like a good bit of hardware - the fans certainly feel premium enough, but their lack of balance was really very noticeable, something I discovered when re-using the AK620's fans on some Thermalright SI-100 coolers.


For such cheap coolers, the Thermalright stuff of late has been impressive - both in terms of performance, and in quality of packaging.

I don't know if it's the same company as I used to be but prior to the last few years, Thermalright used to be very high-end air cooling that I last used in the mid 00's (Ultra 120 and XP-90).

Yeah I was impressed by how my fc140 was packaged, it was better packaged than my noctua NH d14 was many many years ago, noctua might be different now I don't know

 

[freeagent]

Shit, now I look like a Thermalright shill....

I am a huge Thermalright fanboy, so don't feel bad

 

[claes]

I'd love to see reviews on this matter.

Keeping the chip itself cooler does more for the VRMs than cooling them directly, at least in open bench tests. IMO it’s more of a case airflow problem than an advantage downdraft coolers might have, but I suppose it’d have to be tested in a case, where testing becomes more or less useless.

 

[Gmr_Chick]

Yes, post #99

Now that this thread is solved we're all going off-topic because that's what forum threads are for


I borrowed stole one of the SI-100's from work, because it's doing the VRM cooling in an otherwise passively-cooled case.

Thermalright, impressive quality for not much money - a very flat, nicely milled baseplate (yes, a baseplate, not some cheap direct-contact cop-out), plenty of surface area for the typical ~150W CPU and a fan that I personally dislike and replaced, but is actually decent. It's not a bad fan, but it's a bad pairing for a cooler that clearly isn't going into a bone-stock ATX build. For the price, I can't complain because the heatsink alone was worth the full asking price.
View attachment 308040View attachment 308042

Going off-off-topic, Thermalright are a company I ignored for 15 years and have recently (post-COVID, post-Brexit) become a very appealing prospect in the UK. I am building so many PCs that I always know what the best bang-for-buck air cooler is and for a while that was Cryorig, then it was ID-Cooling. Both of those companies seem to have doubled their price whilst still being low-budget build-quality. Meanwhile, Thermalright have flooded Amazon.co.uk at rock-bottom pricing.

https://www.amazon.co.uk/Thermalright-Assassin-Cooler-Heatpipes-TL-C12C/dp/B0B4JCWHVG

£16 including tax is effectively a $17 cooler which is insane for something I can get delivered in under 12 hours. It's nothing special for a $40 air cooler but it's not $40, it's way less than half of that.

Shit, now I look like a Thermalright shill....
"Hey Thermalright, I'll shill for you for the right price and this one's on the house."

Hahaha, it's perfectly fine. I'm actually finding all this off-topic discussion interesting and I'm learning things about fans that I never knew about before.

 

[Shrek]

Keeping the chip itself cooler does more for the VRMs than cooling them directly, at least in open bench tests. IMO it’s more of a case airflow problem than an advantage downdraft coolers might have, but I suppose it’d have to be tested in a case, where testing becomes more or less useless.

Interesting, but possibly deceptive:

• Low load: one is blowing warm air on them, so they don't seem cool
• High load: the warm air will keep them from baking

so it may be hard to judge the virtues of downdraft coolers without a high load scenario.

 

[P4-630]

If you have a quality board it should have already proper heatsinks on VRMs, to keep them as cool as possible.

 

[claes]

Interesting, but possibly deceptive:

• Low load: one is blowing warm air on them, so they don't seem cool

If they are overheating you have a problem. VRM are built to run very hot, as in 90*-120* before you see any performance issues or degradation

• High load: the warm air will keep them from baking

so it may be hard to judge the virtues of downdraft coolers without a high load scenario.

Of course? How else do we judge the effect of heat on performance?

 

[FoulOnWhite]

My VRM's seem to do pretty well. I have a custom loop and they don't seem to have any direct air flow. I do have my case setup for + though, which might help, not sure.

 

[Shrek]

Of course? How else do we judge the effect of heat on performance?

Just not taking anything for granted; safer that way.

 

[Space Lynx]

@Gmr_Chick i might have missed it, but just wondering, did you like the packaging of the peerless? Mine had a lot of foam cutouts in it, which I just thought was really nice. My cooler from them is bigger than yours though so maybe they only do the foam cause of the weight

 

[claes]

Just not taking anything for granted; safer that way.

What does this even mean lol

 

[Mussels]

The advantage of https://getfancontrol.com/ is that we can make the case fans react to both GPU and CPU temperature.

I do not see this as an advantage at all.

The fact that you're relying on software means theres a delay and they're spinning up after it's already heated up - your fans should already be at the highest state you can handle and not need to spin up more than that, other than to compensate for some sort of screwup or problem (Such as a fan eating a cable and getting stuck, or ambients hitting 40c)

That doesnt even account for what happens if the software ever errors or crashes, iCue proved that one a lot over the years by crashing out and locking peoples fans in 0RPM mode or whatever RPM they were currently running at - simply running Hwinfo still causes that error to this day, with a commander pro.

 

[AusWolf]

I do not see this as an advantage at all.

The fact that you're relying on software means theres a delay and they're spinning up after it's already heated up - your fans should already be at the highest state you can handle and not need to spin up more than that, other than to compensate for some sort of screwup or problem (Such as a fan eating a cable and getting stuck, or ambients hitting 40c)

That doesnt even account for what happens if the software ever errors or crashes, iCue proved that one a lot over the years by crashing out and locking peoples fans in 0RPM mode or whatever RPM they were currently running at - simply running Hwinfo still causes that error to this day, with a commander pro.

The advantage of any software over BIOS control is mitigated by the fact that it's software, with all the disadvantages software comes with. For that reason, BIOS level hardware control will always be superior.

 

[Gmr_Chick]

@Gmr_Chick i might have missed it, but just wondering, did you like the packaging of the peerless? Mine had a lot of foam cutouts in it, which I just thought was really nice. My cooler from them is bigger than yours though so maybe they only do the foam cause of the weight

Oh, I didn't get the Peerless. I got the newest Phantom Spirit 120. Came in a pretty good sized box, with thick foam cut outs for the heatsink and the fans.

@freeagent - I got it up and running today. The fans I bought for it, the TL-S12's are kinda....anemic when it comes to the ARGB halo -- they aren't thick and vibrant like the Light Wings from Be Quiet! so I'm probably going to take them off and just keep them for spares (they didn't cost much so no biggie). I'm going to put the stock fans on (and use the top ARGB header on my board for the ARGB P12's on the roof of my case) and see how that goes.

One thing I've always hated about tower coolers is the fan wires. My experience with them was no different here, specially when it came to mounting the fan that's meant to be sandwiched between the heatsinks. I fought with mounting it for at least 30 minutes. Other than that, I commend Thermalright on the installation process on AM4/5.

 

[freeagent]

The one thing I do to make life way easier, is to drop the cooler on its mount pegs and give each one a few twists until they stop turning. I saw the GN video where they struggled to mount it and it kind of boggled my mind why they were having so much trouble with a board that had no support in the center and flexed like mad every time they oust pressure down..