TRUE & Megahalems: Performance with different fan configurations (Rd 2)

[ehume]

Didn't you used to have a non-OCCT graph on this thread? Or am I hinking of another review?

 

[muddocktor]

You are probably thinking of my original thread or perhaps the thread that baditude did comparing them on an i7 system.

 

[dejo]

wow, truly incredible write up. great job, and thanks for taking the time. I have a scythe mugen2 and the mega and have noticed similar results between them also. seems that the mega is the go to cooler for now. could change with the venomous x

 

[ehume]

You are probably thinking of my original thread or perhaps the thread that baditude did comparing them on an i7 system.

Thank you. It was the baditude thread.

 

[muddocktor]

My testing has again come to a halt and again it's because of motherboard problems. This time, I had at least 1 of the power fets around the socket fry tonight while making a run with 1 Panaflo M1BX fan installed. So tomorrow I will either try to finish up with my P5Q Pro board or I might also try taking the 4 pin ATX socket off the P35-DS3L board and soldering it inplace of the burned socket on the P35-DS3R board. I am having no luck at all with these Gigabyte boards with this Q6600.

I have listed the runs I did make with the Venomous X so far and they are looking pretty darn good too.

 

[muddocktor]

I've gotten the test rig back in action with the P5Q Pro board and have been listing my runs as I complete them. And it's looking real good for the Venomous X too. So far I have been seeing a pretty consistent lower temp with it versus the other heatsinks. After I finish the series or runs on the Ven X I will do some spot retesting of the others on this board to make sure the temps haven't changed significantly with the different motherboard. And I also plan to test the Venomous X mount with both TRUE heatsinks to see if the mount makes a difference on cooling efficiency.

 

[bing]

I also plan to test the Venomous X mount with both TRUE heatsinks to see if the mount makes a difference on cooling efficiency.

Thanks Mud !

 

[muddocktor]

All the results are up for the Venomous X now and presently I am making some runs with the TRUE Rev C, my original TRUE and my Cu TRUE while using the Venomous X mounting brackets. The base shape on all of them are the same, so it's no problem to sub any of the different mounting brackets on all of these heatsinks. Basically the Venomous X is just a slightly evolved TRUE with different mounting brackets.

As far as the mounting hardware goes, it seems to mount the heatsinks with good pressure, but applying maximum clamping force by screwing that nut in the middle of the crossbar down also gives it some room to pivot a bit. I would say the heatsink can rotate around 5-10 degrees and is pretty easy to move around at first. After a little while though it seems that the TIM "sets" a bit and it's resistance to heatsink rotation gets better. I think the new revision TRUE mounting brackets provide much more resistance to heatsink rotation than the Venomous X mount. After running all the heatsinks on the Venomous X mounting hardware I will give them all a spin with the new revision TRUE mount and see how they perform.

 

[ppe1700]

GREAT great post. Definitely gives me ideas.
Thanks

 

[Leonardo]

I will give them all a spin with the new revision TRUE mount

In more ways than one, it would appear.

I'm greatly looking forward to your posted findings, and just as much, your judgment.

 

[muddocktor]

Well Leo, after years of folks like us posting and bitching about their poor base finishes (and Prolimatech coming out with a worthy competitor), Thermalright has tightened up the QC on their base finishes finally. The TRUE Rev. C has a well done base that is finished much like a bowed waterblock with a good finish on it and the Venomous X has the same shape to the base, but done with as fine a polish on the base as I have ever seen on a heatsink.

 

[muddocktor]

Now here is some interesting info. I ran the other TR heatsinks using the Venomous X mount and here are the results I saw with them and the Venomous X mount:

Original TRUE with Venomous X mount, 2 Denki's in push-pull config:
Core 1 max temp - 58.5 C run 1, 59 C run 2
Core 2 max temp - 58.25 C run 1, 59.5 C run 2
Core 3 max temp - 54.5 C run 1, 54 C run 2
Core 4 max temp - 54.25 C run 1, 54 C run 2

TRUE Rev. C with Venomous X mount, 2 Denki's in push-pull config:
Core 1 max temp - 59 C run 1, 58.5 C run 2
Core 2 max temp - 59 C run 1, 58.75 C run 2
Core 3 max temp - 54.5 C run 1, 54 C run 2
Core 4 max temp - 54 C run 1, 54 C run 2

Copper TRUE with Venomous X mount, 2 Denki's in push-pull config:
Core 1 max temp - 58.75 C run 1, 58.5 C run 2
Core 2 max temp - 59 C run 1, 58.5 C run 2
Core 3 max temp - 54 C run 1, 54.25 C run 2
Core 4 max temp - 54 C run 1, 54 C run 2

Now compare those temps with the Venomous X run in the same configuration above (repeated here):
Venomous X with 2 San Ace fans in push-pull:
CPU1 max temp - 59 C
CPU2 max temp - 59 C
CPU3 max temp - 56.25 C
CPU4 max temp - 56 C

Looks to me that for all intents and purposes, it's a tie between the 4 heatsinks.

Note: This was using the Venomous X mount with the center screw all the way in for maximum clamping pressure for all runs. It's looking to me like the biggest difference between all the heatsinks is the mounting system for the Venomous X giving better clamping pressure, which is giving better heat transmission. I am presently running the TR heatsinks using the Rev. C TRUE mount and see what kinds of temps I see with all 4 heatsinks. I have a ways to go, but it's looking like the Rev. C mount gives worse temps, as the first heatsink I'm checking is the Venomous X and I am seeing higher temps across all 4 cores.

EDIT: Runs below are with the Rev. C TRUE mount that has been modded with using a Swiftech backplate instead of the one that came with the mount. That is because the Swiftech backplate gives a tighter mount to the mobo than the TR backplates, which yields more clamping pressure between the heatsink and processor. I tried with both backplates and the Swiftech one gives 1-3 C better temps.

Venomous X with 2 San Ace fans in push-pull:
Core 1 max temp - 62 C run 1, 62 C run 2
Core 2 max temp - 62 C run 1, 61.75 C run 2
Core 3 max temp - 56.5 C run 1, 55.75 C run 2
Core 4 max temp - 56 C run 1, 56 C run 2

TRUE Rev. C with 2 San Ace fans in push-pull:
Core 1 max temp - 60.25 C run 1, 61 C run 2
Core 2 max temp - 60.75 C run 1, 61.75 C run 2
Core 3 max temp - 54 C run 1, 54.25 C run 2
Core 4 max temp - 54 C run 1, 54 C run 2

Original TRUE with 2 San Ace fans in push-pull:
Core 1 max temp - 59 C run 1, 60 C run 2
Core 2 max temp - 59 C run 1, 60 C run 2
Core 3 max temp - 55.25 C run 1, 55 C run 2
Core 4 max temp - 54 C run 1, 55 C run 2

Copper TRUE with 2 San Ace fans in push-pull:
Core 1 max temp - 59.25 C run 1, 59.75 C run 2
Core 2 max temp - 59.75 C run 1, 59.5 C run 2
Core 3 max temp - 55 C run 1, 54 C run 2
Core 4 max temp - 55 C run 1, 54 C run 2

 

[Leonardo]

Ha, that's....I don't know what to say. Wow!

 

[ehume]

Here is a review where the guy tested the Prolima Mega Shadow with one and two GT-1450's.

 

[muddocktor]

OK, I just finished making 2 runs with all the other Thermalright heatsinks using the Rev. C TRUE mount and the data is posted in post #32. While they are close, it's looking to me like the Venomous X mount gives a little better temps than the Rev. C TRUE mount. I did try one other trick with the last 2 heatsinks I tested, which were the original model TRUE and the copper TRUE. I put a layer of duct tape across the top of the base, which helps tighten the mount just a little bit more. And both of those heatsinks seemed to run 1-2 C cooler than the other 2 heatsinks I tested with the Rec. C mount first. They did not have duct tape applied across the top of the base. So it's still looking like clamping pressure is reining supreme with all the Thermalright heatsinks instead of design shape or material it's made with. At least with this amount of heatload that is; it might make more of a difference with a highly overclocked i7 platform where it's not uncommon to be able to have 250+ watts of heat dump just from the processor.

 

[oc_aznboi]

Hi, muddocktor

I am in the market for a new heatsink for a core i7 build. Choosing between mega and VX since both manufacturer produce quality heat sinks.

Are the results on the first page, 2nd post updated? Like the same board used on all heatsink you tested including TRUE rev c and venomous x? Simply looking at the results, the TRUE rev c is better than the mega. And the venomous X has 3-5 degree advantage than mega depending which fan is used.

"Now for the TRUE Revision C results. These might not be directly comparable to the previous 2 runs due to the mobo change. I will do some test runs with the Megahalems and my original TRUE to double check their results on this mobo after I test the Venomous X:"

"Now for the Venomous X results, which will be worth waiting for I think. I also had to go to yet another mobo for this round of testing as the P35-DS3L cooked off a FET shortly after starting testing (3 runs completed then it fried). I re-ran all tests after the mobo swap to keep everything in the series constant. I also double checked the completed runs against the new runs with the Asus board and they were within a degree or so. "

 

[Brutal-Force]

Yay for my Venomous X and Panaflos!

While I would like to warn anyone buying Panaflo fans that you should also buy a fan controller. They are noiser (tolerable) than say.... an 1800 RPM Xigmatek fan, however they outperform by leaps and bounds. I run mine daily at 1500 RPMs through the fan controller, and I turn em up to 2500 RPMs for benching. My temps do not climb much over 80 degrees at 4.0 with 1.4 Volts.

 

[dominick32]

Mudd, I dont know what to say man. I remember you mentioning something to me about testing fans in a thread before, but this is just outstanding my friend. Simply outstanding.

Dom

 

[wickedout]

OK, I just finished making 2 runs with all the other Thermalright heatsinks using the Rev. C TRUE mount and the data is posted in post #32. While they are close, it's looking to me like the Venomous X mount gives a little better temps than the Rev. C TRUE mount. I did try one other trick with the last 2 heatsinks I tested, which were the original model TRUE and the copper TRUE. I put a layer of duct tape across the top of the base, which helps tighten the mount just a little bit more. And both of those heatsinks seemed to run 1-2 C cooler than the other 2 heatsinks I tested with the Rec. C mount first. They did not have duct tape applied across the top of the base. So it's still looking like clamping pressure is reining supreme with all the Thermalright heatsinks instead of design shape or material it's made with. At least with this amount of heatload that is; it might make more of a difference with a highly overclocked i7 platform where it's not uncommon to be able to have 250+ watts of heat dump just from the processor.

One thing we need to realize is that temps vary from PC to PC and ambient room temps also vary. Even applying the TIM correctly makes a difference to any cooler and CPU! I just got done watching a video about cooling! I left the video saying that the coolers base is very important so that you can gain lower temps and a idle setting or in a load setting. I never lap any of my coolers.

Thanks Mudd for the great information. I'm happy to say I love my Megahalems.

 

[muddocktor]

Hey all, I ran out of time to retest all with the same board on the C2D system. But I have a new i7 setup coming and hopefully I will get time to rerun the heatsinks on it. I have a 980X and Asus P6X58D Premium board and Kingston HyperX T1 DDR3-2000 ram that should be coming in tomorrow, so it should be waiting for me when I get home. I plan to experiment with it on air first, so after I get a reasonable overclock on it (4 GHz to start with?), I will try rerunning the heatsinks again and see how they stack up. I won't run through all the fans, but rather just concentrate on the Sanyo Denki fans in push and push-pull. You can always extrapolate the performance with the other fans from that I believe. And I will also play with my Denki counter-rotating beast on the TR heatsinks at least too, just to see how much further 300 cfm of absolute mayhem and noise cools better than 2 Denki 1011 fans in push-pull.