Noctua NH-U12P tower Cooler
Again, I can hardly believe to 33 decibels! Mike, did you measured the sound with the P12 fan mounted on the heatsink? I ask a forumer of the Italian site related to silent computing I usually attend and he told me the P12 fan is the best 12x12 fan he ever tested but once mounted on the heatsink (he uses an U12P heatsink) the sound changes dramatically. He tried to move down the fan about two fins to cool the northbridge but the fan started whistling really a lot. In his experience the P12 overrides the 30 decibels limit once on a heatsink.
But another forumer replied his U12P is totally quiet and he has 4 Noctua P12 inside the case ( http://www.pcsilenzioso.it/forum/showthread.php?t=5638 ).
But another forumer replied his U12P is totally quiet and he has 4 Noctua P12 inside the case ( http://www.pcsilenzioso.it/forum/showthread.php?t=5638 ).
"For continuity, we won't be using their thermal compound. We'll take their word for it."
I disagree with this policy. Even a difference of 2 degrees would have moved NH-U12P past HR-01. Thermal compound is part of the package and you pay for just as you pay for the cooler and fan. Therefore it should be included in the test.
I disagree with this policy. Even a difference of 2 degrees would have moved NH-U12P past HR-01. Thermal compound is part of the package and you pay for just as you pay for the cooler and fan. Therefore it should be included in the test.
It would have been fair if you would have used their thermal paste, but even a 2 degree difference would not have made the product much better or worse, it is still too minimal.
Last edited by Suosaaski on Fri Jun 27, 2008 7:05 am, edited 1 time in total.
But at least there "should" have been a comparison with their own compound and "reference".osmium76 wrote:Erssa you're right, but different termal compounds mean different scores and a good test should give an idea of the performances of the heatsinks by themselves (project, materials, manufacturing, and so on). For termal compunds there are a lot of excellent rounds up on the web.
Truly I can't understand the usefulness of a double test... If I have to buy a heatsink I look to the best results of the model by itself. Then I select the best termal greese. I will use the stock compound only if it's the best available. If you're looking for the BEST: http://benchmarkreviews.com/index.php?o ... &Itemid=62
Most will probably use the compound that comes with the cooler.osmium76 wrote:Truly I can't understand the usefulness of a double test... If I have to buy a heatsink I look to the best results of the model by itself. Then I select the best termal greese. I will use the stock compound only if it's the best available. If you're looking for the BEST: http://benchmarkreviews.com/index.php?o ... &Itemid=62
No it's not. It serves those who'd like to buy the best package (me) and it also serves those who buy their their compounds separately (you). Clearly a need for double test. Obviously it's also one of the reasons why they do a double test with fans.osmium76 wrote:... and so a double test is useless.
I'd be happy that Mike tested at all. I don't see any obligations here, unless of course you prefer "paid-for pages".
I'm wondering when the first .5mm-soldered Al-fin heatsink, with 5mm fin-spacing and proper mounting system will turn up. Possibly in combination with direct-touch, but I suppose that's a Xigmatek patent.
I'm wondering when the first .5mm-soldered Al-fin heatsink, with 5mm fin-spacing and proper mounting system will turn up. Possibly in combination with direct-touch, but I suppose that's a Xigmatek patent.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
The review of 33 TIMs linked by osmium76 is pretty good... it shows something like a 2C difference amongst major brands w/Intel E6550 Core 2 Duo 2.33GHz (oc'd to 3.23GHz) and a stock Intel HSF.
But let's make it clear why we don't test w/anything except the same TIM we've been using for years: The same reason we prefer to test w/the same fan we've been using for years. It's a reference, a constant. Yeah, some coolers come with TIM that may not be as good, some a bit better. But the TIM isn't going to change a bummer to a killer and vice versa. We already put a ton of effort into every review we do; at this point, we're not interested in increasing the effort for so little gain.
But let's make it clear why we don't test w/anything except the same TIM we've been using for years: The same reason we prefer to test w/the same fan we've been using for years. It's a reference, a constant. Yeah, some coolers come with TIM that may not be as good, some a bit better. But the TIM isn't going to change a bummer to a killer and vice versa. We already put a ton of effort into every review we do; at this point, we're not interested in increasing the effort for so little gain.
Yeah well, as I already said, the difference between TIMs is not big enough to make any real difference, 2 degrees is not gonna change anything, be the C rise for example either 15 or 17 at a given moment. Both are good. And if the C rise is 35 degrees, 2 degrees is not gonna help it so much...MikeC wrote:
But let's make it clear why we don't test w/anything except the same TIM we've been using for years: The same reason we prefer to test w/the same fan we've been using for years. It's a reference, a constant.
Hmm, sorry but I doubt those strange fan sounding results. I have tried Nexus and P12s at same speed (1000rpm) and Nexus is louder also moving much less air than Noctua.
Then those mp3 recordings...
Noctua NH-U12P with stock NF-P12 fan at 5V, 7V, 9V and 12V at 1m
Nexus "Real Silent 120mm fan" (at 5V, 7V, 9V and 12V at 1m)
Nexus and Noctua fans have been recorded using different mic sensitivities. That's very noticable even to the untrained ear. I don't know but something is wrong with this review.
Then those mp3 recordings...
Noctua NH-U12P with stock NF-P12 fan at 5V, 7V, 9V and 12V at 1m
Nexus "Real Silent 120mm fan" (at 5V, 7V, 9V and 12V at 1m)
Nexus and Noctua fans have been recorded using different mic sensitivities. That's very noticable even to the untrained ear. I don't know but something is wrong with this review.
In my opinion something's wrong with the heatsinks... As I said some posts ago two forumers of our Italian silent PC site had different experiences with the Noctua UP12. One says the P12 fan is extremely noisy once mounted on the heatsink; the other says the heatsink is the quieter he ever listened; now he keeps the P12 at 1120 rpm (LNA). I asked him to make a test with the P12 at full speed, as I have answer I post.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
There's no doubt in our minds that at 12V, the Noctua fan is much much noisier than the Nexus. We'll be happy to make a recording of both fans at the same time with the same mic in one recording.
You're right that the levels were not matched. We're having transition hiccups -- going from a Sennheiser ME66 to the new ACO Pacific 1". The latter gives us greater gain -- nearly 10 dB... and less noise at the same level. But the overall tonal balance is different, the ME66 is bright and forward with much less bass, while the ACOP is almost perfectly flat and extends much deeper into the bass. We're going to have to re-record our most popular/reference fans so that they can be fairly compared with future recordings. It's tricky because of the gain, noise and tonal differences. We are going to try and stay at the old ME66 mic recording levels, but this means experimenting with different level settings during recording until we get a good match. It's not as easy as it might seem.
I've posted a new gain-adjusted recording of the Noctua. The starting ambients in the 2 recordings are now within 1 dB of each other. You'll notice now that the Noctua recording's ambient is less hissy than that of the Nexus. That's the difference in the "self-noise" of the mics. Whether you can hear the brightness of the ME66 on the Nexus is hard to say -- it's just noise (not music or voices) and there's not much to peg any reference (to listen/compare).
Listen to the recordings carefully: The Nexus at 12V sounds quieter -- it has a much lower pitch because it's not spinning as fast. It's also smoother sounding. There's a touch of high freq. tone in the Noctua that keeps it always a bit more edgy sounding, and its pitch at 12V is higher, the blades are spinning faster.
You're right that the levels were not matched. We're having transition hiccups -- going from a Sennheiser ME66 to the new ACO Pacific 1". The latter gives us greater gain -- nearly 10 dB... and less noise at the same level. But the overall tonal balance is different, the ME66 is bright and forward with much less bass, while the ACOP is almost perfectly flat and extends much deeper into the bass. We're going to have to re-record our most popular/reference fans so that they can be fairly compared with future recordings. It's tricky because of the gain, noise and tonal differences. We are going to try and stay at the old ME66 mic recording levels, but this means experimenting with different level settings during recording until we get a good match. It's not as easy as it might seem.
I've posted a new gain-adjusted recording of the Noctua. The starting ambients in the 2 recordings are now within 1 dB of each other. You'll notice now that the Noctua recording's ambient is less hissy than that of the Nexus. That's the difference in the "self-noise" of the mics. Whether you can hear the brightness of the ME66 on the Nexus is hard to say -- it's just noise (not music or voices) and there's not much to peg any reference (to listen/compare).
Listen to the recordings carefully: The Nexus at 12V sounds quieter -- it has a much lower pitch because it's not spinning as fast. It's also smoother sounding. There's a touch of high freq. tone in the Noctua that keeps it always a bit more edgy sounding, and its pitch at 12V is higher, the blades are spinning faster.
I don't think anyone questioned this? 33 dBA vs 22 dBA is quite noticeable to anyone. Hardly something that can be explained even if hearing is subjective.MikeC wrote:There's no doubt in our minds that at 12V, the Noctua fan is much much noisier than the Nexus. We'll be happy to make a recording of both fans at the same time with the same mic in one recording.
That's like comparing Nexus @ 11v to Noctua @ 8.5v?ssb wrote:Hmm, sorry but I doubt those strange fan sounding results. I have tried Nexus and P12s at same speed (1000rpm) and Nexus is louder also moving much less air than Noctua.
Then those mp3 recordings...
P.S found a typo in page 5. According to the fan table, Noctua fan weighs 1700g.
Here what the "Pucc" said:osmium76 wrote:In my opinion something's wrong with the heatsinks... As I said some posts ago two forumers of our Italian silent PC site had different experiences with the Noctua UP12. One says the P12 fan is extremely noisy once mounted on the heatsink; the other says the heatsink is the quieter he ever listened; now he keeps the P12 at 1120 rpm (LNA). I asked him to make a test with the P12 at full speed, as I have answer I post.
"noctua P12:inudibile tra i 590 e i 910rpm, silenziosa fino a 1150 se non fosse per un deciso huuum (direi una nota sui 350/400Hz) che compare nell'intorno di 950rpm e 1200rpm.A questi giri è inutilizzabile...diventa veramente fastidiosa.
Per capirci a 950 và in risonanza e al minimo rumore d'aria che emette si aggiunge quella nota che accresce di molto i dB...sopra i 1000 il suono sparisce e torna ad essere un'ottima ventola (non + inudibile però)...sopra i 1100 ricomincia.Personalmente una delusione dalla ventola + costosa che ho".
"noctuaP12: not audible between 590 and 910 rpm, quiet until 1150 exept for a "huuum" (I think a bit about 350/400 Hz) that comes between 950 and 1200: in this range it isn't usable because it's too noisy. At 950 it goes in resonance... up the 1000 it becomes a good fan again but over the 1100 the problem returns. Personally I'm very disappointed by my most expensive fan".
The other forumer "Stalingrado" as I said uses both Noctua P12 and S12 fans and the tower cooler and he is totally satisfied...
I agree. Keep up the good work.MikeC wrote: But let's make it clear why we don't test w/anything except the same TIM we've been using for years: The same reason we prefer to test w/the same fan we've been using for years. It's a reference, a constant. Yeah, some coolers come with TIM that may not be as good, some a bit better. But the TIM isn't going to change a bummer to a killer and vice versa. We already put a ton of effort into every review we do; at this point, we're not interested in increasing the effort for so little gain.
I find it very annoying that every other website tests heatsinks with only stock fan at default speed. Even when they lower the speed, they often don't report the RPM, and of course even RPM info isn't enough (noise level, noise character, actual air flow, etc.).
Double testing stock fan vs reference fan is already above and beyond the call of duty.
As for thermal paste, I don't think it makes that much difference anyway. Even so, I usually just buy whatever is cheaply available and decent quality (happens to be Arctic Cooling MX2) and use it with everything I have.