Corsair H50 vs Prolimatech Megahalem

[layte]

I'd also add that my system has a 4870x2 with the Arctic Cooling Accelero installed, so no heat from the card is vented straight out.

 

[OLDPHART]

Also using Prime95 8HRs 76c i7>4GHz 1.260v Mx-2 Apache's - push>pull> air out, and this was at the height of the hot spell we had. When l ran Prime and as above my GFX Card heat is not vented out ATI 4890 950/4200MHz, but l would also add my PC is in the living room which gets the sun all day. It gets very hot l've bought a thermometer, for next year just to see how hot my living room gets out of curiousity.

 

[JonJ678]

it does not carry a large volume of water either so the system will get saturated fairly quickly and temps will be much higher.

This part looks like suspect reasoning, volume of water doesn't really affect equilibrium temperature. The rest I agree with completely though, and the H50 does appear to saturate somewhere north of 4ghz, after which the processor heats up rapidly with applied voltage.

The issue I take to the H50 plus two apache guys is that it comes out at £85. That's getting a bit too close to doing water 'properly', allbeit second hand. A 10W ddc can be had for 40 quid, a thick 120 radiator for 25. An ek supreme lt at 20, and a tenner on tubing/barbs. £95, and it wont need two apaches to get performance out of it.

I confess I'm looking forward to the first load of 'oh noes, my h50 is leaking/doesn't work anymore'. The owners have been altogether far too smug and anti custom watercooling for people who aren't very well placed to judge. The H50 is always going to perform worse than a ddc/120 feser combination however you look at it, and that's nothing to be very excited about.

Its a 'sealed' unit, fluid will evaporate from it. Not much liquid in it, so this effect is going to make itself felt. Bubbles in the system will be loud and cooling will suffer. As from the start I've been in favour of a large metal heatsink as it can't really break, I will feel vindicated when this occurs.

I'm stripping my loop down in a week or so, I'll put a shout out to see if anyone with a UD5/H50 combination is up for trying the same bios settings as me on a ddc and 120 radiator, to try to judge what the difference is in reality. Hardly scientific since they'll be different chips, but probably worth doing.

 

[gurusan]

No it does not affect equilibrium temp, my point was that it will reach the equilibrium temp much quicker vs a setup with a larger volume of water....and this "buffer zone" is one of the benefits of wc...

 

[Hotwired]

snipz

No need to be smug in anticipation either.

The H50 is not a passive cooler, it contains a pump and circulating fluid. One of which will eventually go **** up no question.

An almost solid metal air cooler is not going to do anything.

The H50 is a decent 3rd party cooler that can give vastly better use of space than the equivalent air system or full-on watercooling. That's as far as it goes.

I have no doubts whatsoever that a large aircooler in a case with good airflow will outdo it. Ditto a good watercooling system.

Also, what's with the pricing? The fan shipped with the H50 for it's rad isn't worthless but you've ignored it, put in two expensive fans (with a bad colour scheme) and then decided the rads for proper watercooling won't need any fans. Play fair will you?

 

[liamcrane]

megahalems all the way for me, have it set it with to apache fans keeps my phenom at stock 3.2ghz 30c idle and 40c load cant complain at all

 

[JonJ678]

Ah, I'm with you now gurusan. Shame the buffering effect doesn't survive running prime overnight, but it definitely helps for keeping fan speeds down.

Yate loons can be had for about £3 and work rather well on radiators with a low fin density. One of the issues with the corsair is the high fin density, hence people using it with high pressure fans. I see no issue with using it with the stock fan, but there's a lot of people on here who saw the need to strap two apaches to it. I'll let them answer why as I'm not really sure.

I guess you mean using a uniform, continuous block instead of something with fins? I can't see anyone suggesting using a block of metal anywhere so I'm not sure what you're getting at with the third line.

I'm not sure a large heatsink will always beat it. It has a considerable advantage in drawing air from outside the case rather than inside. I do agree it is far more likely to fail than a heatsink, and doesn't seem to perform any better. I don't really see the fuss, so put it down to feeling pleased that you're using liquid cooling.

Watercooling will cost more, of course. I'm comparing new to second hand as well. The price gap is small enough though that the custom approach makes a lot more sense to me, otherwise you're getting the risks of watercooling without the options to move bits around or the performance.

I'm pleased with the radiators in my case. Cant fit any hard drives or optical drives, and it weighs a ton, but I'm happy feeling smug about it nonetheless. Even though this does render me as a hypocrite. Thanks for calling me on it though.

 

[Nelly]

but I also think that if you are going to watercool then you should do it properly and not faff around with these little kiddy playkits.

lol!!!

 

[OLDPHART]

Posted By Gurusan, JonJ678>

but looking at it from an engineering standpoint it just doesn't have the cooling capacity of a megahelems

it does not carry a large volume of water either so the system will get saturated fairly quickly and temps will be much higher.

No it does not affect equilibrium temp, my point was that it will reach the equilibrium temp much quicker vs a setup with a larger volume of water....and this "buffer zone" is one of the benefits of wc...

This part looks like suspect reasoning, volume of water doesn't really affect equilibrium temperature. The rest I agree with completely though, and the H50 does appear to saturate somewhere north of 4ghz, after which the processor heats up rapidly with applied voltage.

Couldn't be bothered to put the wrap tag's on the above.. on the mumbo jumbo.

Most people will agree H50 will not beat a good water cooler set up, but it compete's with the like of Pro Meg etc. In the long run we will just have to see how long H50 last's, if it fail's so be it, at least l've tried one.

Posted BY gurusan
but I also think that if you are going to watercool then you should do it properly and not faff around with these little kiddy playkits

LOL, if thats not adig at the H50, l don't know what is...

 

[Hotwired]

Yate loons can be had for about £3 and work rather well on radiators with a low fin density. One of the issues with the corsair is the high fin density, hence people using it with high pressure fans. I see no issue with using it with the stock fan, but there's a lot of people on here who saw the need to strap two apaches to it. I'll let them answer why as I'm not really sure.

I guess you mean using a uniform, continuous block instead of something with fins? I can't see anyone suggesting using a block of metal anywhere so I'm not sure what you're getting at with the third line.

I'm not sure a large heatsink will always beat it. It has a considerable advantage in drawing air from outside the case rather than inside. I do agree it is far more likely to fail than a heatsink, and doesn't seem to perform any better. I don't really see the fuss, so put it down to feeling pleased that you're using liquid cooling.

Watercooling will cost more, of course. I'm comparing new to second hand as well. The price gap is small enough though that the custom approach makes a lot more sense to me, otherwise you're getting the risks of watercooling without the options to move bits around or the performance.

I'm pleased with the radiators in my case. Cant fit any hard drives or optical drives, and it weighs a ton, but I'm happy feeling smug about it nonetheless. Even though this does render me as a hypocrite. Thanks for calling me on it though.

I was referring to the fact that an air cooler has no moving parts and is essentially solid metal in it's structure therefore nothing to fail. Not the same as saying it's a solid metal block.

It's ability to draw air in is practically limited to what can be drawn through a 120mm fan opening, an air cooler isn't limited in the same way.

Anyways I'm going to be getting one for practicality reasons, It's undoubtedly better than this stock intel limpet clamped to my 750 but I'm going for a rather interesting approach in my mATX case and there is neither the room for a watercooled system or a desire to have a steel hedgehog of an airsink hogging the mobo.

I'll tell you how smug I am about it when it arrives ^_^

Mind you, one of the compact swiftech systems would be interesting to compare with.

 

[gurusan]

LOL, if thats not adig at the H50, l don't know what is...

Ok perhaps it was a bit

 

[nickrose]

Before we start I was not expecting miracles from the H50, but maybe a drop of a degree or 2 would be nice.

Specs.
i7 @ 4Ghz
Corsair 800D
Artic Silver 5 Thermal Paste

I recieved and installed the H50 on Saturday.

Previous I had been running on a Prolimatech Megahalems Air cooler with 2 x Noctura fans in push/pull.

Getting good temps - Mid 40's idle (about average I've been reading) and Mid-High 70's/Very Low 80's (Rarely) when running Prime95 (again about average at this speed)

I also reconfigured my other cooling whilst installing the H50:

Added - Silverstone SST-CFP51B Aluminum Bay Cooler + added extra Akasa Apache fan on back side of cage
Changed Corsair's 3x140mm case Fans for 140mm Sharkoon's
Changed top 3 exhaust fans for 120mm AKASA AK-174BKT-B Auto Thermal Fans

I then went on to install the H50, using 2 x Akasa Apachie 120mm fans in push/pull drawing cool air in from the rear case point.

H50- When I booted the machine up I let it Idle for 30 mins to bring up to normal working temp. I notice that using RealTemp the idle temps were about 4-5 degrees hotter than previous. I then ran prime95 for 1/2 hour and similarly those temps were 3-4 degrees hotter than on air.

Ran the machine most of the day whilst gaming and re-ran the tests again on Saturday evening and got the same results a earlier in the day.

Went back to the Megahalems Air cooler thinking on Sunday thinking that it was the new case fan config that I had not got right but to my supprise the fan re-config had lowered the temps even more by about 2-3 degrees, so I knew the input/exhaust fan air balance was O.K

This was not a scientific test, intended as a comparison or anything else, it is just my humble opinion on what I saw whilst changing to a H50.

So my 2 pennies worth - Stick with Megahalems - I have, and the H50 will go on e-bay.

 

[Yellowbeard]

It may be worth your time to let the TIM cure a bit if you prefer the H50.

 

[reygjabla]

So my 2 pennies worth - Stick with Megahalems - I have, and the H50 will go on e-bay.

Post when you do. I'll bid on it.