I've just bought a Corsair H50 Hydro Series CPU cooler, and I've to say that I'm not happy with it in terms of silence, altought it can be perfect for overclockers who don't care noise.

The fact: the H50's pump makes a very prominent sound in my quiet system.

I would personally compare the pump noise level to the i7 860 stock cooler fan at full speed, but the problem is that I thought the H50 would be quieter than most quietest CPU coolers because of its small size and low water volume, so I decided to spend this worthy amount of money on the H50, but even using the stock i7 860 fan, if not on heavy load, the i7 860 stock fan is definitely many times quieter than the H50 pump noise.

And it's not only the noise level, but the kind of noise too.

I want to make clear that it's not the H50's fan noise what I'm talking about (I really didn't install it), it's the pump's noise, because I used both the top and bottom fans of the Raven to cool the radiator, finally getting better cooling with radiator on top and the fan above it sucking hot air out of the case.

And I also want to make clear that's not the initial bubbling sound what I'm talking about, because that sound only lasts for a few seconds after system power up.

I'm talking about the persistent pump sound even after 5 hours of system up running.

This are the components of my new system:

- CASE: Silverstone Raven RV02-Window (Stock Fans at low - total 4 fans)
- GPU: Powercolor ATI Radeon HD 5750 (Passive)
- PSU: Seasonix X-750 (passive with low load / fan cooled on high load - can't hear it yet)
- HDD: Intel X25-M 160 GB SSD Gen 2
- MOBO: ASUS P7H57D-EVO (cheaper than P55 where I bought it)
- CPU: Intel i7 860 (not overclocked)
- RAM: 8 GB Kingston Hyper-X
- CPU cooler: Corsair H50 Hydro Series (radiator on top, but tried at bottom too)
- SO: Windows 7 Enterprise Trial


Before purchasing the H50 I read a lot of reviews and there was no mention to the highly rattling/buzzing sound of the H50 pump, so I decided to purchase it, but now I've found that noise to be THE ONE most prominent sound of all the system, and somehow irritating because of its special kind of noise (this is the first time I hear a watercooling system pump).

If I add some old 3.5" HDD's to the system (for example an old and very noisy Seagate 7200.7 160 GB), the H50 pump noise gets somehow masked and not so prominent, but it still can be heard above, although less irritating, because the mechanical HDD is vibrating too, and filling the somhow the intermitent silence of the pump.

But I've tried too using a 2.5" mechanical HDD (LaCie external USB) and the H50's noise is still highly above everything else, because I can't really hear the 2.5" drive in this system.

The cooling capabilities of the H50 are really impressive but that's not what I was looking for, because I really wanted an almost silent system, and because of some reviews, I thought the H50 could be very quiet, but now I realize it isn't right for my system. I think it's really noisy even with the case completely closed.

I must say that I think you can finally get used to that kind of noise because of its complete regularity. All the time makes the same noise, not higher, not lower, be it idle or high load, and finally your brain makes it dissapear if you're not looking for it.

Anyway, altought aesthetically the H50 looked very nice through the case window, I'm going to ask for a refund on this item and buy instead a Noctua NH-D14. Later I've read some people RMA this unit up to three times but the noise persisted, so that's why I'm going to ask for a refund.

I've experimented connecting the pump to several fan sockets in the Asus MOBO, some with fixed voltage and some auto-regulated by BIOS, but the pump noise is always the same

I just wanted to share my experience with other people who looks for a quiet or silent system too.

Hope it helps someone else.

Thanks you for your informative post, I already read the sound was noticeably but was wondering how much so, and your personal experiences nicely informed me of what I feared already

Now as for the actual problem, did you consider wrapping the top of the cooler in some sound-adsorbing material, after all the part on the CPU doesn't transfer the heat but the radiator does, and the part on the CPU has the pump in it, only problem is that around the CPU there's the area with voltage regulators and you need to keep clear of those, but even so there should be room and it should be possible to put a shroud over the part on the CPU to reduce sound, theoretically.

On a further search I found a somewhat conflicting report
on silentpcreview in /forums/viewtopic.php?t=55167
Where people say it is very silent, and it includes a youtube link to sample it.
But I think the issue is the kind of sound it is, it just annoys some people while others it does not and if it doesn't I guess it's much less noticeable.

It's because of that topic you're linking that I decided to post this one, so that other SPCR users have more information, although I couldn't actually watch the video due to my slow internet connection.

In global terms, in a standard PC, I think it won't be noisy at all, but inside a quiet PC as mine is (I also plan improvements in the future) it is very prominent, unless you attach some OLD and noisy 3.5" drives. Then you can barely discern the pump noise from the HDD vibration, although I think it's raising the total dB(a) of the system a little bit, but in this case, the noise/cooling balance may be it's favouring the H50.

About the shroud, I suppose it will dimm the noise a bit, but it won't look very nice through the case window (I think the Noctua won't look nice neither, but surely better that something I can make myself at home). Anyway I'm asking for a refund (2 days left).

By the way, I've tried to record the noise with a mobile phone (using video recorder) and it simply didn't record that noise at all, just some global noise.

Then I tried with a MP3 voice recorder and, although the recording doesn't represent what I really hear, at least shows the kind of noise you can expect.

I'll try to post better recordings, but I don't promise anything because I've a veeery sloooow Internet connection for now (2-3 KB/s, really, mobile phone and low coverage).

I'll thank anyone who can please let me know how can I link an MP3 to this post meanwhile.

Note: A few weeks ago I watched a TV documental that explained how different animals understand the same reallity (humans included), and the conclusion was that what we hear and see is not always the real thing, because each brain makes some modifications to that reallity, leveling up some sounds, lowering others (not just filtering frequencies), and so on... So, while the noise can be recorded, it's up to everyone how noticeably you'll finally hear it.

Anyway I find this kind of noise very irritating when I've no attached older 3.5" drives and even more when I open the case (as I do now) to do a better recording. I remember that I decided to purchase the H50 because I've always heard that, although the water cooling pumps made noise too, this kind of noise was nicer than the fans, but now I don't think so, at least talking about my only one water cooling experience with the H50.

so is it the pump or the cooling fan on the radiator that is making the noise?

You make me happy danimal, so you let me repeat it once more.

It's the H50 pump, that makes much more noise (and different) than the fan on the radiator (at 700~800 rpm, using Asus QFan in the Bios), although I didn't even put the Corsair's fan because I think the Silverstone's one was even more quiet.

I was tempted to disconnect the pump's power connector from the motherboard and the buzzing stopped at all , but the temps of the CPU raised from 25ºC to 60ºC , so I quickly reconnected it (I tried that because I read someone stating that the temps stabilized with the pump disconnected, but that's actually false, I tried myself).

If the pump's noise was like that of a fan, my system would simply just play the same sound but louder, but with the H50 in a quiet system you've actually two different songs playing at the same time .

I repeat anyway that using noisy 3.5" hard drives makes the pump's noise dimm a lot, but now I'm using a SSD drive and I'm just hearing the soft breeze of the quiet fans plus the buzzing of the pump, that would make me just hate it if I couldn't get a refund .

There's no doubt that for you it is best to find another cooling solution, but for people who didn't manage a REAL quiet system the H5O might still be an option.
Anyway I wish you success with finding a better cooling setup, the eternal search for all of us continues.

Yes, I said that... I'm going for the Noctua NH-D14 because of the excelent cooling capabilities as shown here in Silent PC Review, just removing the 120cm fan as for a 95W TDP CPU it isn't necessary.

But because this is Silent PC Review is that I tried to express the negative points of such a cooler, so that other potential buyers don't get tricky information about the "silent" fan of the Corsair, when the main source of noise of the H50 comes from the pump, not the fan, and its somehow irritating caracter.

By tricky information I mean measuring the dB(A) of the fan but not measuring the db(A) of the pump, because I really believe they were cheating on me.

i guess that you did make it clear in your first post, but i seriously did need to hear it again, because i was looking to get that unit... thanks for the feedback.

i'm wondering if the character of the pump noise changed as the load on the cpu ramped up.

or if there was any way to muffle the pump, i.e., wrap some insulation around it with a zip tie... unless the pump motor is air-cooled?

regardless, the noctua should be a better choice.

In my system (FT02, q6600), the Noctua blew the H50 out of the water (pun intended) stock vs stock. Core temperatures are 6-9C lower with the Noctua than they were with the H50. Here are my results, cross-posted from another forum as well as from another thread here:

Corsair H50 vs Noctua NH-D14

*FAIL = at least 1 core temp reached 71C so test was stopped
*Max performance of stock H50 indicated in RED
*Max performance of stock D14 indicated in BLUE

3.4 GHz (378*9)

1. H50, Stock Corsair Fan, Push-Push ==> Idle 44C; Peak 69C
2. H50, Stock Corsair Fan, Push-Pull ==> Idle 38C; Peak 66C
3. H50, 110CFM Scythe Fan, Push-Pull ==> Idle 37C; Peak 62C; LOUD
4. Noctua NH-D14, Stock (Push-Pull) ==> Idle 38C; Peak 60C
*At this clock, stock D14 cooled significantly better than did stock H50 (temp difference of 9C when H50 was push, and 6C with H50 as push-pull)


3.5 GHz (389*9)

1. H50, Stock Corsair Fan, Push-Push ==> NOT TESTED
2. H50, Stock Corsair Fan, Push-Pull ==> FAIL
3. H50, 110CFM Scythe Fan, Push-Pull ==> Idle 39C; Peak 68C; LOUD
4. Noctua NH-D14, Stock (Push-Pull) ==> Idle 38C; Peak 63C


3.57 GHz (397*9)

1. H50, Stock Corsair Fan, Push-Push ==> NOT TESTED
2. H50, Stock Corsair Fan, Push-Pull ==> NOT TESTED (nb: failed at 3.5 GHz)
3. H50, 110CFM Scythe Fan, Push-Pull ==> FAIL
4. Noctua NH-D14, Stock (Push-Pull) ==> Idle 40C; Peak 68C
5. Noctua NH-D14, 2x CM R4 Fans (Push-Pull) ==> Idle 40C; Peak 66C
6. Noctua NH-D14, 3x CM R4 Fans (Push-Pull) ==> Idle 42C; Peak 65C

3.58 GHz (398*9)
4. Noctua NH-D14, Stock (Push-Pull) ==> FAIL
6. Noctua NH-D14, 3x CM R4 Fans (Push-Pull) ==> Idle 43C; Peak 66C

3.60 GHz (400*9)
6. Noctua NH-D14, 3x CM R4 Fans (Push-Pull) ==> Idle 43C; Peak 67C


Summary

In my system, the D14 outperformed the H50 stock vs stock. The D14 also outperformed the "modified" H50 (upgraded fan). The Noctua performed well and was QUIET. The only speed at which the two systems could be compared in stock form was 3.4 GHz, where the D14 outperformed the H50 by 6C (H50 push-push setup) and 9C (H50 push-pull setup).

MaxVcore was 1.47V. My goals were to keep vcore below 1.5V and core temps below 71C.

Where I live, the price difference between the D14 and H50 is $10. The cost of the 4 CM fans (3 for cooler and 1 for case) was $22, total. For similar prices, given the performance in *my* system, I would go with the Noctua right off the bat if I had the chance to do it all over again. It's not as good-looking (terrible-looking, actually), but it works very well and provides that performance in a quiet manner.

I knew the great cooling properties of the Noctua, but never saw a comparison. Great news for me.

Here there's only 1 EUR difference between the Noctua and the H50. I've already returned the H50 and now I'm waiting for the Noctua.

Meanwhile I've saved the recordings of the noise so if someone is curious about it, please help me posting the mp3.

And yes, danimal, the H50 noise doesnt't ramp up on heavy load, it's high all the time. And about muffling it, try yourself... I'm sure any further questions, danimal, I've already answered that .

I finally got the refund on the Corsair H50 and used the stock cooler for a while.

The Noctua NH-D14 came today.

I seriously doubted about purchasing it for its high weight, but now I'm very happy because it gets even more securely fixed to the MOBO than what I thought. Really solid.

And about the cooling experience it's really amazing (it isn't cheap anyway, neither is the H50).

I don't have much data to compare, but it is really cool all the time.

At first, the fans were just toooo loud, but using the Ultra-Low-Noise-Adapters combined with the MOBO's QFAN feature they became really silent while throwing very much air through anyway.

I finally replaced my case's top fan with the Noctua 120cm fan that I removed from the heatsink.

On idle, system at 21-22ºC, CPU at 25-26ºC.

I remember that I've seen that numbers with the H50 but with the Noctua they seem to last longer.

When using the CPU (full antivirus scan with an SSD drive, really quick drive access and by the way very high CPU usage) it gets up to 10ºC higher, but instantly lowers again the same 10ºC for short periods of time that0s not using the CPU 100% meanwhile scanning. The H50 was slower to cool down again.

I've even given a try momentarilly to the Automatic Extreme Overclocking utility that came with the ASUS MOBO and it even reached higher values than with the H50. My i7 860, 2.8 GHz at stock, reached almost 4.05 GHz (45% higher) with all the case fans at its bare minimum and using only the 140cm fan with the Heatsink (with the ULNA + QFAN the MOBO says ~480 rpm). I can remember the H50 reaching 3.87 GHz at its full power.

I also tried removing the 140cm fan inside the heatsink, I mean totally passive, and it cooled really well, even with the Raven's bottom 180cm fans disconnected (5-6ºC higher on idle?).

The negative point comes in its ugly appearance, specially using the 140cm in a windowed case, but the low noise and cooling performance makes it perfect for me. The H50 really looked high end anyway, but the noise...

There's also another thing I'm still trying to solve, because when I connect the 140cm fan to the MOBO's CPU fan socket, it doesn't use the QFAN feature that raises or lowers the RPM of the fan as needed, emitting more noise than I want. The Intel stock cooler seemed to work right with the QFAN but no the Noctua.

For now I've connected it to the MOBO's Chassis Fan Socket that also uses the QFAN feature and seems to work fine with the Noctua fan, although it's very strange for me. I also tried that change with the H50 pump but it was at full RPM all the time even on the chassis socket, but seems to work fine with the Noctua, lowering its RPM to ~480 (as said by the MOBO).

Anyway this is not how it is meant to be, because now the CPU fan will only speed up if the System's temperature raises enough (which I doubt with the Raven case), but not when the CPU temperature raises

-----------------------------------------------------------------------------------

P.D.: When I asked for the H50 refund I also decided to replace my old 160GB Seagate 7200.7 SATA HD that I used sometimes for testing purposes with a new WD Caviar SE16/Blue 640 GB and I'm really surprised of the noise difference between them. The WD is really quiet. In my new system I decided to purchase 5.25" drive bay for the HD so that I could disconnect it when I'm using just the SSD drive, to get rid of the noise and vibration, but with the WD I'm really surprised that I don't need it. The old drive did also some kind of whining sound that seems to be inexistent with the WD, and the same with the vibrations that made every metal piece of the case resonate. I'm very happy with it too.

-----------------------------------------------------------------------------------

Thank you SPCR for the excellent reviews, I'm just sorry I couldn't buy these items through one of your affiliate retailers.

Did you actually try any kind of speed control on the pump at any point ? I have one, and yes the pump is noticable in a quiet pc.. but only if you run it at full speed. Turning it down it becomes practically silent (I can't hear it) with almost no impact on cooling, a few degrees at load at the very most. I'm using 3pin pwm control on mine without issue.

I also tried the Corsair H50 and it is way to noisy for my system and I can confirm that the cpu temperature is higher with H50 than with my Zalman CNPS10 Extreme. I got 44 C on CPU with H50 and with Zalman CNPS I get 37, Idle.
I did of course replace the standard fan on Zalman CNPS because the standard fan is way too noisy above 1080 rpm and even at 1080 my nexus or Noctua NF-P12 that I replaced the standard fan with is less noisy and it drops the CPU temperature with 2 degrees this is at the same noise level.
It is easy to replace the standard fan on Zalman CNPS extreme I just had to mod my Noctua and Nexus fans with a knife and then they fit easy it takes couple of minutes to cut a 0,6x0,3 cm in the frame. I cant decide which fan to use the Nexus or Noctua and I havent even tried the scythe slipstream that I also bought
I got a Zalman GT1000 with Noctua fans, the Noctua fans at 1080 rpm all three of them the 120 mm and the two 92 mm at front and it gives CPU 37 with MB temperature of 28 with CNPS 10 Extreme also that one with Noctua NF-P12 at 1320 rpm. Room temperature 24-25. CPU is i7 870 running at 178x22=3916 MHz and 1606 MHz idle with DDR3 2133 MHZ and Kingston V+ 325 with WD Black 640 GB and Zalman 850w PSU plus an ASUS ATI 5850. Going to mod the graphic card with Musashi when I find a high quality screwdriver that can unscrew those small screws holding the GPU heatsink, I already destroyed 2 cheap screwdrivers trying to unscrew them

darkb:


Yes, I tried the Asus mobo fan controllers, two of them, CPU and Chassis 1, and none of them made any difference in the sound level of the H50.

Although the Noctua fan didn't work with the CPU fan controller, it did finally with the Chassis 1.

For what I've read in some SPCR fan articles there are two ways of regulating a fan, one is by pulse regulation and the other one is by voltage regulation, so I think that the my Asus mobo fan controllers have both of them, one kind in the CPU fan controller and the other in the Chassis 1 one.

So I don't think an external fan regulator will lower the H50 noise.

Because I've already returned the H50 I cannot try it again, but I'm 95% sure that I've tried both fan controllers with the H50.


swift77:



Good luck with it. I decided to buy a passive GPU because I didn't want to void the warranty, a Powercolor ATI HD 5750, that is fine on Idle (~30-35ºC) but it gets very hot when playing (55-60-65ºC) so I have to at least raise the middle bottom 180cm fan of the Raven-02 to High for it to stay under 55ºC, otherwise the system locks between half an hour or so sometimes. I wished it would stay cooler, becase I thought that being passive since factory it wouldn't need anything else than good ventilation, but I've discovered it needs a fan anyway (when playing), but I'm happy that if I'm working or listening to music, I can have it passive with no noise at all. Let's wait to the summer to see how does it behave under hotter temperatures, but I fear the worst.

65c is nothing for a graphics card. You can say it's hot when it starts going over 80c but not before. GPUs are made to work at those temps unlike CPUs.
Most GPU don't even start to throtle before 90c+ !

Dropping the pump speed should lower the noise, just like this review says.

Yes he can hardly hear it but I can for sure hear it. Point is why use something that is more noisy and perform worse than a good aircooler.
I researched and read reviews of H50 prior of buying it and I just got so disappointed after spending all the time researching and installing just to discover that it IS noisy and perform WORSE than what I already had installed (Zalman CNPS10X Extreme with Noctua/Nexus fans).

I'm happy to hear that .

Anyway the problem I experienced was not throttling, but the whole system frozen and black screen, even the reset button stopped working. I thought it was because of the high temps on GPU because at lower temps the system didn't lockup (and playing even more time at once).

My system locked up 5 times (out of 10) now with the middle bottom fan at low and "none" with the middle fan at high. With the fan at high, only once a game stopped working yet but it was very different, because it seemed to be software related. The system really didn't lockup completely this time, just aborted with the blue screen of death while showing some debug info and now the reset button worked.

I'll keep an eye anyway on it without raising the middle bottom fan (kinda GPU fan in the Raven-02).

I doubt if the lockups were due to something GPU BIOS related, as some sort of automatic safety measure if reaching 65-70ºC or so . I couldn't update the BIOS yet.

This is some information from the Catalyst Control Center:

- BIOS date: 2009/11/04
- BIOS version: 012.014.000.004
- Controller pack version: 8.66 xxxxxxxxxx

Thank you for your feedback, Redzo.

First, I could NOT drop the pump speed with any of the mobo fan controllers, not by hardware, and not by software, although with the same mobo, yes, I could lower the speed of the fans (H50 fan, Noctua, Silverstone...), so I'm not sure when others say that they can lower the H50 pump speed, because I couldn't. Maybe someone missed it and said pump when they should say fan, I don't know.

Anyway I don't trust many of reviewers out there. Now that I've tried it myself, I truly think that SPCR are the best, and not only for that they have showed that a product from a brand that's not sponsoring them is better than another's competing product from a brand that actually sponsors them, but for the throughful tests they make, even breaking out the rules of previous reviews if they think the reader is missing some point, be it a good or bad point.

But what I think I've seen in SPCR is that they say (may be with others words) that, if they don't review a concrete product, there are high probabilities that the product is very noisy, so they don't even take the time to make a full review, but they are fair enough so as not to say that a product is "bad" if they don't make a full review. That's why I'm angry with myself because I bought the H50 although I didn't read such a review in SPCR, I just guessed myself that the H50 was quiet because it was watercooled (maybe I didn't guess it myself alone, may be I was "helped" by such a review that you link us to), but nothing more far from reallity, the H50 is not noisy, it's veeery noisy. So I wonder again why the Corsairs measure the H50's fan noise level but don't measure the pump's noise. I think that's cheating.

But I agree with Mats in that why would you use something more noisy while you can get something so much quietter for the same or even half the budget, while performing even better .

Like Mats, I was very disappointed with the noise too. Although I went for a different heatsink, but air cooled anyway. I chosed the Noctua NH-D14 because the SPCR ranking showed that the Prolimatech and the Noctua were the best (with difference from the rest), but in my retailer I couldn't buy the Prolimatech, but I could buy the Noctua NH-D14 and use the 120cm fan for the case, and with the already included Ultra Low Noise Adapters and anti-vibration mounting, the price was even lower than using the Prolimatech anyway.

The only problem I find now with the Noctua is that I can't attach the 140cm fan to the CPU mobo socket for auto-regulation of fan speed. I can only attach it to the Chassis 1 fan to lower the noise/speed.

Now that I think more about it, there are to factors on the H50 that make it sooo noisy, the low water volume and the small pump (so that it can be integrated in top of the CPU itself). I think that both factors make the pump spin very fast to give good enough cooling in high CPU usage although making much more noise than a traditional cooling system.

Aesthetically, the H50 looked very fine for a windowed case like mine, but the noise level and noise kind was really a drawback, so I would only recommend the H50 to people who like more to show his computer to others than actually using it, or to people who's computer noise base level is yet really high because of noisy overclocking (and again I think the noise is so different from that of a fan that even some people could think they don't like it).

Now that I replaced my old noisy dedicated hard drive for games with a modern Western Digital, I think I couldn't get used to the H50's noise ever.

While I have not tested the H50 Against the Noctua, I have tested it against two other coolers in a I7 DO system. My results are as follows:

Core I7 920 DO (CM Storm Scout case)

3.9GHZ

Zigmatek Dark Knight Cooler (Push/Pull CM R4 Fans)
Intel Burn Test max Load temp (20 passes) = 75C

Megahelems Megashadow Cooler (Push/Pull CM R4 Fans)
Intel Burn Test max Load temp (20 passes) = 70C

H50 Cooler (Push/Pull CM R4 Fans) Exhaust setup
Intel Burn Test max Load temp (20 passes) = 67C

4.0GHZ

Zigmatek Dark Knight Cooler (Push/Pull CM R4 Fans)
Intel Burn Test max Load temp (20 passes) = 78C

Megahelems Megashadow Cooler (Push/Pull CM R4 Fans)
Intel Burn Test max Load temp (20 passes) = 73C

H50 Cooler (Push/Pull CM R4 Fans) Exhaust setup
Intel Burn Test max Load temp (20 passes) = 71C

The H50 is a great cooler.

Get the Nanoxia PWM controller (Aerocooler among others sells it) You hook it up between the motherboard PWM fan header and the Noctua. It makes the Noctua behave as if it were a PWM fan.

Yea, but look at these test results, at Frostytech.com
http://www.frostytech.com/articleview.c ... 501&page=5

It is in 60th position on the list. It doesn't do all that badly - trailing the Proimatech by only about 5C. But those benchmarks aren't consistent with yours. It's Db rating isn't that bad either compared with others.

Still, if you go by Frostytech's benchmarks, the H50 wouldn't necessarily be your first choice.

ps: When you say "CM R4 Fans" what exact model and what speeds are you talking about? Are you familiar with the CM Blademaster? If so, how does it compare with the R4s you used?

Firstly, I'm not confusing the fan and the pump, I have them both on different headers controlled with different settings on my mini-itx dfi board. They are listed as CPU fan and PWM fan in the bios, and I can set starting points and slopes etc in the bios for each individually, as well as controling each via speedfan. The pump works all the way down to 45% in speedfan, which is just over 700rpm.

The pump is effectively silent at that speed, and max cpu temp only increases a few degrees, just topping 60C in speedfan under orthos load on my i5 750 @3.2Ghz. The only moving parts in this pc are the fans, and the pump: 5850 stock fan at 20%, 120mm arctic cooling PWM case fan at 700rpm, pump at 700rpm, and the psu of the SG05. I don't have a normal hdd in it, just a 120GB ssd.

Also because this is a small pc, it sits on my desk and the pump is 2 feet away from my ear in a straight line, with no acoustic dampening of any kind. It's very loud and annoying at full speed, but once it gets down past 60% or so I can't hear it, the almost buzzing type noise disappears.

In regards to the question of why to buy a H50, regardless of the fact that it can be quiet and highly efficient, it's also the best cooler that can fit in some cases, mainly ones where a 120mm fan tower design is too tall or is obstructed, but a 120mm case fan position still exists. Not every case can fit the H50's direct competition, and in my case the biggest air cooler I could fit would be the shuriken or the gemini II without a fan.

For the time being I have retired my (2) H50's. The buzzing was just too much when it was the loudest thing (even over my 5870 stock cooler). I picked up 2 used coolers, a venomous x and a tuniq tower 120 extreme for about what 1 of the h50's cost.

So right now they are both sitting in their boxes, I should go ahead and ebay them I guess. With the 2 air coolers I'm seeing about 5c lower at idle and load across the board pretty much.

My experience with H50 actually is very positive. My H50 pump doesn't make buzzing noise, and I cannot hear it over my case fan. The performance is also very good. The load temp is about 3C better than Thermalright Venomous X.

Nice choices. How does one go about buying nice coolers like that? How much off of their new retail price did you pay? What causes people to sell such good coolers?

Well this is my first message here, but I'll take to time to express my opinion about the H50.

The pump is definitely noisy, to the point that I replaced all my components inside an old P180 , but without great success.

The fan IS silent, but the pump emit an high frequency buzz . I could RMA the thing, but from what I've read it probably won't be better. In DBA, the noise isn't that lound, but since it's an high frequency it's annoying.

I never thought about slowing down the pump for H50. I haven't connected my H50 yet to see if I hear anything but would slowing down pump cause any unintended issue? Maybe such as damaging pump or any premature fail?

I can confirm everything that's been said about the noisy H50 pump. It is indeed very iritating. They accepted the RMA but I don't think things will get any better with a new one. Does anyone know how to lower the voltage on the pump without a fan speed controller (can't do it from the MB either). Thanks.

My H50 pump is the noisiest part of my system. If you've got it plugged into the CPU fan header, you monitor/adjust the pump speed via the BIOS or software. (Might need to disable the CPU fan failsafe to boot if the RPMs are low....silly BIOSes assume if the CPU fan speed is below a certain threshold they shouldn't boot.)

Or if that won't work, you can add an in-line resistor cable to decrease the voltage/RPM. My Noctua fans came with them so I've got spares, but you can buy them individually as well in various sizes. They look like this: http://www.acousticpc.com/zalman_3pin_f ... 6_ohm.html

You'll need to need to see if that will work in your case, and watch the temps for a bit. It will affect definitely affect performance, but how much I can't tell you as it's based on your case airflow.

I've heard that some H50s are quieter than others, but I suspect it has more to do with user tolerance to noise than defective units. Curious to know if the H70 is quieter or if I should go back to an ugly, massive air cooler...still undecided.