My H50 vs. H70 Test

so from those pics all they really did was make the rad thicker and add more fins per inch, basically it's a high performance 120mm rad that requires high noise, high pressure, high RPM fans in order to work well...

basically they need to make a 120x2 rad version with less fins per inch so that it can perform well with slower, quieter fans
 
definitely use your own fans on H50/H70 and I'm used to having 6 80mm fans going on my P4 system before he died.

Also found that the screw threads on the H70 seem to wear out quick, at the moment I have 1 maybe 2 screws that don't "catch" on to the threads. I probably over tightened at some point to wear them out or something. Just hope i don't have to un install it anytime soon, I may email Corsair about it and see what they say
 
As an owner of the H50 and someone who was recently looking at the H70, I have to say at this level its about noise. The single coolest benefit of my H50 is that it is dead silent. I didn't add any fans I lapped the copper base in 45min but thats all I have done.

I wouldn't want to add another fan, I don't need to. It keeps my temps in the sweet spot and is silent. I like how the H70 pump is much smaller, but thats the only advantage I see in it. Like others, I think they went the wrong way on the radiator, and noisy fans is just dumb.

Just my .02 cents.
 
To be honest it looks like they have got them round the wrong way.

The H70 should have come first with its bigger but noiser solution and then the H50 would come along, smaller, quieter and more refined.

Maybe the H70 was an initial H50 prototype they just decided to put into full production.:D

Here are my fans -

IMG_1585.jpg


IMG_1584.jpg
 
I suspected that the H70 wouldn't fare much better than the H50, and that it'd do worse with lower static pressure fans. I just didn't want to say anything because there seemed to be many Corsair H50 loyalists who would give me flak for my remarks, however true.

- Thicker radiators offer diminishing returns (double-stacked radiators, like the swiftech stack, offer very little benefit over a single radiator due to recycled air into the 2nd radiator).
- Not only is the radiator thicker, but the fin density is higher. Thus the need for noisier fans.

I think the H50 can be worthwhile... and I like Corsair, their rebate procedure, warranty, most of their products, and customer service. But the H70 just doesn't seem worthwhile when we have higher performing options like the Silver Arrow.

It's not even all that compact anymore... and that was the primary redeeming quality of the H50 (which I might add is probably slightly outclassed by the megahalems or Ven-X at the same noise level).

What they really needed was a 140mm version of the H50 for people with larger cases, or a 2x120mm version. It'd also be nice to have the option to purchase it w/o fans. I don't care what people say - that's a great option to reduce cost - so what if 2 different models are required.
 
Archmage said:
- Not only is the radiator thicker, but the fin density is higher. Thus the need for noisier fans.
Click to expand...

actually, the H70 is the low fin-density rad and the H50 has a higher fin-density. the main reason they don't show much any difference in performance is because they most likely have close to a similar surface area for cooling.

had the H70 had higher fpi lay out it would defiantly show greater performance, and the optimal fan layout would be like OP's setup with 2x 38mm fans utilizing shrouds.
 
Skull_Angel said:
actually, the H70 is the low fin-density rad and the H50 has a higher fin-density. the main reason they don't show much any difference in performance is because they most likely have close to a similar surface area for cooling.

had the H70 had higher fpi lay out it would defiantly show greater performance, and the optimal fan layout would be like OP's setup with 2x 38mm fans utilizing shrouds.
Click to expand...


You are right on with the results from my tests and photos.

I really thought the H70 would win but the H70 radiator appears to be the weakest link in the H70 build.
 
Wait... H70 has lower fin density?

Hmm... a quick search shows the H50 at 20 FPI, which is somewhat high. The H70 seems to be at about 15-16 FPI, but I'd need to confirm that. You were right.

Now to defend myself: I knew that the H70 had fewer rows of fins, but initial reports were that it was going to have a higher FPI as well: http://gadgetsteria.com/2010/07/30/...ry-level-water-cooling-kit-coming-august-1st/ (like this one. I just pulled a random article, but the same was said in some preview articles (where I developed that notion in the 1st place) : "...the new Corsair H70 is also said to have a dramatically higher FPI (fins per inch)..." )

I suppose Corsair felt the need for higher RPM fans on the H70 simply due to its greater depth.

I was mistaken... my fault for not fact-checking :( - I HATE being wrong, but I'm plainly so.
 
Before I did a google search to determine the FPI, I did see that photo. Upon first inspection, that photo does make the H70 look denser. I was fooled.

I'm not sure whether or not the apexes of the H50's fins are more rounded as you say, but perhaps you're right, and that contributed to the illusion (along with the angle at which the photo was taken, greater depth of the H70, and the light clearly showing through the H50).

I don't think I'll be doing any more research on this because it's of no consequence to me now.

SonDa5: Thanks for the comparison :) - it was well done.
 
From the pictures on the previous page even if the H70 has a higher fin density, it has less water passes through the core (or whatever you call them, I dont know the terminology). It has less sections where the water travels through. So even if the H70 has a higher fin density, it doesn't have more area for the heat to transfer from the water to the fins, so the higher fin density and increased thickness may simply be wasted as the heat can't get into the fins as effeciently.
 
Archmage said:
Wait... H70 has lower fin density?

Hmm... a quick search shows the H50 at 20 FPI, which is somewhat high. The H70 seems to be at about 15-16 FPI, but I'd need to confirm that. You were right.

Now to defend myself: I knew that the H70 had fewer rows of fins, but initial reports were that it was going to have a higher FPI as well: http://gadgetsteria.com/2010/07/30/...ry-level-water-cooling-kit-coming-august-1st/ (like this one. I just pulled a random article, but the same was said in some preview articles (where I developed that notion in the 1st place) : "...the new Corsair H70 is also said to have a dramatically higher FPI (fins per inch)..." )

I suppose Corsair felt the need for higher RPM fans on the H70 simply due to its greater depth.

I was mistaken... my fault for not fact-checking :( - I HATE being wrong, but I'm plainly so.
Click to expand...


Hey man, don't beat yourself over a simple mistake/misunderstanding. :) No one is ragging on you for it, you seem to be your own worst critic. You have certainly done your homework on this topic, and seem to know all the relevant sources/articles pertaining to it.

So cut yourself some slack, most people are wrong on these forums with at least some regularity, some might not even realize it; but others are willfully deceptive with what they present as fact simply due to overinflated egos. It would hurt their delicate feelings if they had to admit they were wrong about something. :eek:

So kudos to you for recognizing where you were wrong/confused, and actually taking the time to show us some sources for your claims. And not kudos because I like to see people being proven wrong; but rather I enjoy having quality posters who actually contribute to threads with meaningful information.

So don't sweat it man, thanks for all your helpful contributions! :cool:
 
Kinda intrigued by your testing Sonda, did you also test the stock setups ? I'm getting far better results with the H70 vs H50 dual fan setup...

I didn't test 4 fans as it ruins the concept of the design.
 
leeghoofd said:
Kinda intrigued by your testing Sonda, did you also test the stock setups ? I'm getting far better results with the H70 vs H50 dual fan setup...

I didn't test 4 fans as it ruins the concept of the design.
Click to expand...

My testing was done with a comparison of an optimized configuration of the H50. I started with stock H50 set up a long time ago. After improving the cooling by changing the configuration of the H50 I chose the optimized configuration for comparison on the H50 and the H70.


Bottom line is that these tests were done under the same conditions apart from one set up having the H50 and the other the H70.
 
Last edited:
I never had a H50, but I can tell you my performance comes no where close to yours, with an ambient temp of 25c my H70 at idle is 39 - 40c using Prime and Intelburn Test (1 hour each) my temps never go above 60c, I have no need to OC.

I have put 2 x Exalibur PWM fans on and do Exhaust out of the case. under normal daily use my fans are unde 1400 rpm at load thay will hit 2200rpm (loud but not noisy)

I think the point here is a lot of this testing is dependant on the environment of the item being tested.

David
 
Just read all this on trying to decide if the H50 i have in a box here I never used would go on my 2600K build. I already planned on a Push Pull setup and this helped push me over the edge. Side note to anybody maybe looking for a H50, the local Best Buys up Norht here have them on closeout for $50 YMMV.
 
djmorgan said:
I never had a H50, but I can tell you my performance comes no where close to yours, with an ambient temp of 25c my H70 at idle is 39 - 40c using Prime and Intelburn Test (1 hour each) my temps never go above 60c, I have no need to OC.

I have put 2 x Exalibur PWM fans on and do Exhaust out of the case. under normal daily use my fans are unde 1400 rpm at load thay will hit 2200rpm (loud but not noisy)

I think the point here is a lot of this testing is dependant on the environment of the item being tested.

David
Click to expand...


look at OP...

Voltage under 100% load of my i7-930@4GHZ cpu is right around 1.36v.
Click to expand...

leeghoofd said:
I didn't test 4 fans as it ruins the concept of the design.
Click to expand...

He didnt use 4 fans, he used two 38mm fans and two 25mm SHROUDS!

do people not read the whole original post? he even drew a nice diagram...
 
I thought I'd contribute a bit to the necro posting here in the hopes that it will help others make informed decisions.

Reading a couple of reviews elsewhere the numbers provided by the OP are consistent. For example, Hexus did a review comparing the H50 and H70, using push/pull on the H50 and the results between the H50 and H70 were very close.

If you have an H50 and are considering upgrading but are only using a single fan, just try a push/pull setup.

Unlike others I found the setup of the H50 to be a pita, the anchor bolts just wouldn't stay in place on the backplate side. Next build I'll try using a Xigmatek back plate.

For myself, the reason I like the H50: 1) It provides solid performance; 2) Footprint, much smaller then most heat sinks used these days; 3) Doesn't apply massive weight to the mobo. I've had heatsinks that bowed the mobo. 4) Clean. The traditional heat sinks have been massive dust collectors for me. Surprisingly, the rad is not collecting dust after 6 months.

Using a Corsair 800D case, I have the H50 pushing air from outside the case in. I have the two front fans pushing air into the case and the 3 top fans exhausting the air. Case temp is usually lower then ambient temp with this setup. (BTW I'll never buy a steel case ever again.)
 
I don't do any of our cooling dev or testing so my engineering knowledge is limited. I've never even driven a train before but I noticed something early on in this post. I think the shrouds may be hindering the performance of the H70. Spacing the fans back, in spite of the intended shroud effect, is likely reducing the static pressure at the rad fins. I suspect that the H70 would do better without the shrouds but, the only way to know would be to test it.
 
Mike Clements said:
I don't do any of our cooling dev or testing so my engineering knowledge is limited. I've never even driven a train before but I noticed something early on in this post. I think the shrouds may be hindering the performance of the H70. Spacing the fans back, in spite of the intended shroud effect, is likely reducing the static pressure at the rad fins. I suspect that the H70 would do better without the shrouds but, the only way to know would be to test it.
Click to expand...



Bottom line from my tests is that the conditions were as close as possible.

I could have tried a better designed shroud but I had a few old 120mm fans laying around which were cheap and easy to convert into "ghetto" shrouds. Thanks for noticing this thread.


:cool:
 
Last edited:
Hey, testing, benching, modding, etc are all fun. Good job IMO. I was just passing on that idea that I think that the H70 would benefit more from the static pressure you'd get from mounting the fans against the rad with no shroud.
 
Mike Clements said:
Hey, testing, benching, modding, etc are all fun. Good job IMO. I was just passing on that idea that I think that the H70 would benefit more from the static pressure you'd get from mounting the fans against the rad with no shroud.
Click to expand...

Before and after temps on H50 with ghetto shroud showed that the ghetto shrouds improved cooling performance.

Ghetto shroud improved performance.


Bottom line is that the conditions were the same and under the same conditions the H50 cooled better. Seems like the H50 is a better design on the radiator and pump.
 
Last edited:
I am with mike on this I think a simple test without the shroud would have been nice. The h70 with lower fpi should be better suited for lower rpm fan and h50 with higher rpm, unfortunately the rads aren't suited for the same type of fans. I prefer the h70 because I want quieter fans (gt ap-14 1450rpm), I had the h50 for a while as well, wouldn't dream of ghetto shrouding ;) I think the ap-15 or 38mm fans are better suited for the h50 and hence your findings.
 
I was thinking about getting an H70 as an upgrade from an H50, simply because the block mounted on the CPU is quite a bit smaller on the H70. But maybe I'll just stick with my H50 now.
 
Likely not defective read my post! Test condition was not optimal and fans not optimal for the radiators.
 
So do you think it's necessary to space the fans off the radiator to eliminate a center deadspot?

Or a thinner spacing?
Also do you think a fan with a higher then 60cfm is necessary on the H70?

What about a push pull setup w/ the pulling fan set at a higher rpm to suck more air through?
 
MrSneis said:
Likely not defective read my post! Test condition was not optimal and fans not optimal for the radiators.
Click to expand...



The fans and shrouds were optimal for the H50.

Bottom line is the conditions. My tests don't answer the "what if" situations that have been presented. Please point to a review that has done a side by side comparisons like my tests with the push/pull and back to back shrouds.

I was going for maximum cooling and this configuration worked best for my system.

I don't have the H50 or the H70 anymore so I can't do any more tests. I'd like to see some tests that answer the "what if" situations that have been presented.

In my mind the H50 works better. H50 costs less money. H50 has a 2 year warranty vs the H70 1 year warranty.

H50 is winner IMO.
 
SonDa5 -

I know you may not have one in front of you, but I would be very interested to see your results with your fan/shroud setup on a Cooler Master Hyper 212+

Personally after owning an H50, an H70, and several 212s.......I find it very hard to recommend anything other than the 212 for the money.

I went from a hyper 212 with push/pull scythes to an h70, and was pretty damn dissapointed
 
QuiteSufficient said:
I went from a hyper 212 with push/pull scythes to an h70, and was pretty damn dissapointed
Click to expand...


I have heard alot of good things about that cooler but have never tried it.

H50 not only cools well but it allows for other tricked out chipset/VGA/memory cooling configurations.

This is was a configuration I ran with my DFI P45 Jr that worked great.

GMSH50chipsetretouch.jpg
 
Just a quick note, I would drop that IC diamond shit like a bad habit and stick with MX3. Not doing yourself any favors with that garbage.
 
Also, might want to add that my 950 has full load temps (F@H) around 68-69c without shrouds at 4ghz. Using two Scythe GT 1850rpm's.
 
jarhead said:
Just a quick note, I would drop that IC diamond shit like a bad habit and stick with MX3. Not doing yourself any favors with that garbage.
Click to expand...

Never tried MX3 but I have used a few other TIMs and the IC Diamond TP has worked the best for lowering temps. Use it for cpu/GPU on my desktops and my notebook.
 
jarhead said:
Also, might want to add that my 950 has full load temps (F@H) around 68-69c without shrouds at 4ghz. Using two Scythe GT 1850rpm's.
Click to expand...


Nice temps. I posted voltage/speed and conditions of my tests. It's all about the conditions. I've seen some cpu's run cooler than others as well. Bottom line is the conditions of the test.
 
Great comparison, I'm actually a bit surprised how high your cpu temp is on the 930@4ghz. My friends [email protected] runs a little cooler using an h50 with push/pull 38mm width fans. I run an h50 for my amd 955@4ghz push/pull scythe fans and have had good results as well. The h70 really is a disappointment...better luck next time Corsair.
 
I think you guys are really missing the point between the differences in radiators and cooling performance of the h50 and h70.
 
I'm not dismissing anyone's individual results in this thread and I totally dig experimenting, ghetto modding, and good ole computer enthusiasm etc. But, when properly implemented, the H70 will cool better than the H50. It's simple physics.
 
Mike Clements said:
I'm not dismissing anyone's individual results in this thread and I totally dig experimenting, ghetto modding, and good ole computer enthusiasm etc. But, when properly implemented, the H70 will cool better than the H50. It's simple physics.
Click to expand...

You don't understand if you think this way. They are different rads, with different specs. Its obvious that with high pressure fans you will get more performance with H50 simply because the rad it uses is a high fpi one...whereas the H70 is not as high.

So, they are both optimized for different configurations, and thats why I don't get the point of the OP: at least try them with something reasonable, such as Nidec GT AP15 or the like, but going with shrouds and UK....:confused:
 
prava said:
You don't understand if you think this way. They are different rads, with different specs. Its obvious that with high pressure fans you will get more performance with H50 simply because the rad it uses is a high fpi one...whereas the H70 is not as high.
Click to expand...
I have seen the internal testing we have done and the info from the OEM. I understand that that the H70, when properly implemented, will cool better than the H50. ;) This may not be the case in every users situation as there are endless variables to consider.
 
You must log in or register to reply here.
Back
Top