[Navig]

Phase One of my Review is complete!



Here are some summary graphs:



Conclusion: bigger fans do give better temps, but expect diminishing returns with the super big fans.





Conclusion:
Best fan configuration: Push-Pull
Second place: Pull
Last place: Push



navig

[imposter]

How about some specs? hehe.

[Navig]

Sorry, the link is disguised in the title!

You want specs, you'll get more numbers than you will know what to do with!

[pak]

I would like to see how it does against the ultra-120

--pak

[Navig]

Sorry, don't have one, and for some reason now, not feeling like picking one up--the anandtech review showed good evidence that the eXtreme thrashes the older version.

[nd4spdbh2]

ya the temps on that e4300 are way to high on load for that cooler... i mean yea your at 1.5v but even with a yate thoes temps are way high... if i remember correctly tat and coretemp had problems with true allandale cores.

[Navig]

Originally Posted by nd4spdbh2

ya the temps on that e4300 are way to high on load for that cooler... i mean yea your at 1.5v but even with a yate thoes temps are way high... if i remember correctly tat and coretemp had problems with true allandale cores.

That's interesting, and not something I was aware of. Any links regarding this statement. Maybe I'll generate some numbers and graphs off the uguru temps, but they seemed to roughly track.

[Moto7451]

Yeah, TAT and Coretemp seem to read a bit high on Allandales, could you throw your conroe in there and redo a couple of the tests to see what difference there is?

[MadMan007]

Wnat to do me a favor and try it flat, test-becnh style, versus vertical as mounted in a tower? I don't think there would be a big difference but I'm always thinking of 'idealized' situations and how they affect results. Extra heatpipes that are vertical (test-bench) versus horizontal (mounted in tower) are more idealized for the way heatpipes work.

[Navig]

As long as its consistent, which it seems to be. I try not to base my conclusions on the overall external actual accuracy of the temperature, after all accuracy can always be called into question unless you embed an external probe into the ihs or the heatsink base--which won't happen because in this heatsink you'll rupture a heatpipe!

That's why everything I do is a comparison--starting at a baseline--and then systematically changing things in relation to the baseline.

Now, my baseline has tracked just fine, altho I was following it with uguru. I systematically started at 9*200, and worked my way up by 10s on the FSB. Needless to say, if you're really interested, I've got screencaps! Anyway, stock 1.325v was fine all the way to the 9*335ish, iirc. Load temps on orthos blend were in the hi 30s celsius. After that the Allendale wall hit--every 5mhz required at least 2 notches on the vcore, with temps appropriately climbing. When my load temps were in the high 50s on uguru, is when I deemed my system sufficiently hot for these top tier heatsinks. The pre-prep was done swapping several times between my Big Typhoon and the Thermalright 120. So I know that the hi coretemp values aren't say from a poor AS5 application. It may be a falsely calculated hi value, but it still should make for a solid baseline.

I'll be honest, I'd not be entirely comfortable putting my prize e6600 at 1.535v and loading up orthos!

I'm okay with my Allendale because, if you think about it, the Thermalright Ultra 120 eXtreme + 2 Panaflos costs more than the cpu!

When I'm done, I'll go back and do some easier tests with my e6600.


Re: vertical orientation, no probs, I'll add it.

[Retrolock]

reviews like these made me dump my scythe infinity and buy the 120 extreme!

thanks man! just bought mine and I'll install it later! w00t!

[Sleepy_Steve]

w00t -- confirmation that I was right to go nuts with air cooling over the winter and hook up two 120x38mm fans to my Utra 120. And that I was able to get into the threshold where deminishing returns were observed. For me at least that is as close as I get to being maxed out.

[Deanzo]

Hey Navig

Whats your view on the mounting system ?

I installed mine last night and I dont personally think you can put enough downward pressure the way the mounting system works. After putting it in my case and adding a fan (120X38) it now tilts to that side.

Though with the very little testing i've done my full load/oc temps seem to be 5 to 10c better than my SI 128.

This is on the rig in my sig (need to up date the heatsink)

Any how as they say a picture is a 1000 words etc etc.

[Sleepy_Steve]

Thats a bit much for twist... But as long as it lays paralel, which it seems to be doing, it should be fine.

But you may want to double check the spring loaded bolt things they attach with. I have no issues with mine holding up two 120x38mm fans and itself in the same position all zip tied together, but who knows... Perhaps you may have forgotten one of the bolts and are getting odd geometrys out of your HSF?

[Deanzo]

Originally Posted by Sleepy_Steve

Perhaps you may have forgotten one of the bolts and are getting odd geometrys out of your HSF?

roflao,
na mate, they are all done up as much as you can.
Its the mounting bracket its self, they couldnt use the kind that comes with the Ultra 120 or the SI 128.
The retail 775 bracket is not like the one in the anandtech review, wish I'd taken a picture of it now.
You put it through and then open it up.

But Im not alone

Click here, click here

[Mycobacteria]

Average of two experiments?

[Navig]

I don't really mind the little bit of twist, but I guess the majority of my cases are horizontal. The twist is actually necessary for me, I think there is a little bit of clearance issue with my ghetto'd HR-09 on the PWM.

[Navig]

Update: Part 2 Lapped vs Unlapped

Full thread here.

Summary tables:

Load temps, Coretemp






Load temps, uguru




Summary: Stock vs Lapped

Through 3 different configurations, the Lapped heatsink beat the Unlapped heatsink by 1.7c (Coretemp) or 0.7c (uguru). I’d call this a consistent but small difference.

[youngbuck]

With that small amount of difference, I'm not sure it'd be worth me lapping mine when it arrives in a few days.

What kind of differences are seen when lapping the cpu? I'm still debating whether or not I should do this.

[Sleepy_Steve]

Yeah, but every bit helps.

[Peepaw]

Where are you guys getting the Extreme at? I can't find it anywhere.

[Sleepy_Steve]

Originally Posted by Peepaw

Where are you guys getting the Extreme at? I can't find it anywhere.

Mine is extreme because it does an extremely good job.


I don't know where you would find the newer version you seek, check TR website and see if its actually being shipped yet? And then go through and check all the vendor sites they list?

I would like to help you, but sometimes doing your own legwork (as it seems you've been doing) generates the best results.

[Brando]

Originally Posted by Peepaw

Where are you guys getting the Extreme at? I can't find it anywhere.

Here ya go. 24 left at time of this post. Better jump on it.
http://www.heatsinkfactory.com/Therm...r-p-16580.html

[psi]

Navig trying to view your thread over at abit, but it keeps timing out? maybe the site is down? I was trying t oorder some extra clips and havnt found any? Also one guy said that either side you mount it to, it uses the same hole?

[Navig]

The abit forums are undergoing updating the last few weeks, so its a bit of hit or miss.

Yes, the fan clips mount to the single row of vertical holes that goes right down the center of the side of the heatsink. In stock configuration, it uses the top and bottom hole. You could probably shift the holes used to fit front and back clips--it might be goofy, annoying wrangling the clips (did I mention I hate the clips), but I think it can be done.



BTW, finished the heatsink comparison numbers, just got to crunch the numbers and make some graphs.

navig

[Navig]

Abit forums are running a bit slow, so I'll post my results here.


Part 3 ~ Heatsink Comparison
Thermalright Ultra-120 eXtreme, Thermaltake Big Typhoon, Tuniq Tower 120


The Setup:

All 3 heatsinks have been lapped to a mirror:
TRU120X
TTBT
Tuniq

The testing was all done in one session from about 1a-5a, which is about as constant ambient temperature as can be found in Boston. The TTBT was tested in push mode (air blowing downwards), and the TRU120X in Pull mode.

The TRU120X and the TTBT were all tested with the following fan settings:
Yate Loon 120x25, 7v
Yate Loon 120x25, 12v
Delta25 120x25, 12v
Panaflo 120x38, 12v

The Tuniq was tested with the following fans (only allows for 25mm fans):
Yate Loon 120x25, 7v
Yate Loon 120x25, 12v
Delta25 120x25, 12v

In addition, each heatsink was tested in what I call “Best Mod” mode. This was more personal testing. I was actually looking for the most reasonable setup that I would personally run. My personal preferences: medium noise—my maximal accepted fan noise would be the D25 or the Panaflo at 7v. In addition, to maximize performance, I’m fine with running physically modded configurations. Here were the actual Best Mod modes:

TTBT Best Mod mode:
Topped with a Delta 120x76 fan (GFB1212VHW). This enormous fan consists of 2 independently controlled counter-rotating 120x38mm fans in one housing, as far as I know, only available thru sidewindercomputers. Running both fans at 5v is an acceptably quiet solution (altho the loudest of the Best Mod modes) that generates a very strong pressure curve. Needless to say, the it is the downfiring layout of the the TTBT that allows for the mounting of such a fan, as opposed to the tower style coolers .

TTBT Best Mod pic.


Tuniq Tower 120 Best Mod mode:
I stuck with the D25 fan internally, but added an extra 80x15mm push fan. This extra fan is actually the AMD opteron stock fan. Both fans were set to 7v. This setup is actually the quietest of the Best Mod modes—the internal D25 fan gets significantly muffled.

Tuniq Best Mod pic.


Thermalright Ultra-120 eXtreme Best Mod mode:
No surprises here. This is the only heatsink that allows for any sized fans in push-pull. So I went with dual Panaflos in push-pull, set to 7v.

TRU120x Best Mod pic.

[Navig]

The Raw Data:

Thermaltake Big Typhoon

Configuration: TTBT, Yate Loon, Push, 7v
Idle: Core Temp: 37 / uguru CPU Temp: 31 / Pwm Temp: 37.2
Load: Core Temp: 78 / uguru CPU Temp: 62 / Pwm Temp: 49



Configuration: TTBT, Yate Loon, Push, 12v
Idle: Core Temp: 33 / uguru CPU Temp: 27 / Pwm Temp: 35.6
Load: Core Temp: 73 / uguru CPU Temp: 56 / Pwm Temp: 48.6



Configuration: TTBT, Delta 120x25, Push, 12v
Idle: Core Temp: 31 / uguru CPU Temp: 27 / Pwm Temp: 34.2
Load: Core Temp: 70 / uguru CPU Temp: 54 / Pwm Temp: 47.8



Configuration: TTBT, Panaflo 120x38, Push, 12v
Idle: Core Temp: 31 / uguru CPU Temp: 27 / Pwm Temp: 33.4
Load: Core Temp: 70 / uguru CPU Temp: 54 / Pwm Temp: 47.6



Configuration: TTBT, Delta 120x76, Push, 5v 5v ~ Best Mod mode
Idle: Core Temp: 31 / uguru CPU Temp: 25 / Pwm Temp: 33.6
Load: Core Temp: 69 / uguru CPU Temp: 53 / Pwm Temp: 47.8

[Navig]

Tuniq Tower 120

Configuration: Tuniq, Yate Loon, 7v
Idle: Core Temp: 34.5 / uguru CPU Temp: 29 / Pwm Temp: 36.2
Load: Core Temp: 73 / uguru CPU Temp: 58 / Pwm Temp: 48



Configuration: Tuniq, Yate Loon, 12v
Idle: Core Temp: 33 / uguru CPU Temp: 28 / Pwm Temp: 34
Load: Core Temp: 71 / uguru CPU Temp: 55 / Pwm Temp: 48



Configuration: Tuniq, Delta 120x25, 12v
Idle: Core Temp: 32 / uguru CPU Temp: 27 / Pwm Temp: 34
Load: Core Temp: 68 / uguru CPU Temp: 54 / Pwm Temp: 46.4



Configuration: Tuniq, Delta 120x25 AMD 80x25 Push-Pull, 7v ~ Best Mod mode
Idle: Core Temp: 33 / uguru CPU Temp: 27 / Pwm Temp: 33.6
Load: Core Temp: 69 / uguru CPU Temp: 54 / Pwm Temp: 46.8

[Navig]

Thermalright Ultra-1200 eXtreme


Configuration: Thermalright U120x, Yate Loon, Pull, 7v
Idle: Core Temp: 33 / uguru CPU Temp: 27 / Pwm Temp: 35.4
Load: Core Temp: 71 / uguru CPU Temp: 56 / Pwm Temp: 46.8



Configuration: Thermalright U120x, Yate Loon, Pull, 12v
Idle: Core Temp: 33 / uguru CPU Temp: 27 / Pwm Temp: 34.4
Load: Core Temp: 68 / uguru CPU Temp: 53 / Pwm Temp: 46.8



Configuration: Thermalright U120x, Delta 120x25, Pull, 12v
Idle: Core Temp: 31 / uguru CPU Temp: 27 / Pwm Temp: 34.4
Load: Core Temp: 67 / uguru CPU Temp: 53 / Pwm Temp: 47.4



Configuration: Thermalright U120x, Panaflo 120x38, Pull, 12v
Idle: Core Temp: 31 / uguru CPU Temp: 26 / Pwm Temp: 34
Load: Core Temp: 67 / uguru CPU Temp: 51 / Pwm Temp: 47



Configuration: TTBT, Panaflo 120x38 x2, Push-Pull, 7v 7v ~ Best Mod mode
Idle: Core Temp: 32 / uguru CPU Temp: 27 / Pwm Temp: 33.6
Load: Core Temp: 67 / uguru CPU Temp: 51 / Pwm Temp: 46.4

[Navig]

The Analysis:

Let us tabulate:


Core Temp Load





uguru Load Temp





And chart it:


Core Temp Load






uguru Load Temp





The second table, for whatever reason the excel tool is including the z axis all the way to zero. Does go to show you, overall, we are only talking about a few degrees of difference.



The conclusion:

Well, its pretty easy to summarize. The Thermalright beat out the competition on every test. The Big Typhoon is showing its age a bit—it was beat out by very significant differences (up to 7 degrees). The Tuniq performed close, but was consistently beat out by a few degrees.