Date: September 1, 2016
Author(s): Ryan Perry
Cooling options for those building or upgrading an itty bitty mini-ITX system are few and far between, and even less so if liquid cooling is a must. Fear not small form factor lovers, Corsair is here to save the day with the H5 SF, the mightiest of all mini all-in-one liquid coolers, so read on to see if it can keep up with today’s pint sized powerhouses.
Not long ago it seemed like the size of PC chassis was about to spiral out of control. Full towers were all the rage, but would eventually give way to the monstrous ultra tower or super tower. These were chassis so big you could almost use them as a coffee table.
Sure, for those running a high-end rig with a custom water cooling setup, triple or quad GPUs, and/or a ton of drives, they certainly served a purpose, but before too long it’s as if someone flipped a switch and component manufacturers began releasing smaller and smaller hardware, sending us in the opposite direction.
Fast forward to the present day and we now have an abundance of mini-ITX chassis that can house some serious, yet pint-sized firepower.
This is all well and good, but what happens when users want to retain the cooling performance of a larger system while shrinking the overall footprint? Smaller, low profile air coolers would be one solution, but they rarely seem to get the job done, especially in a confined space. Enter the Corsair H5 SF, a unique mini-ITX only, all-in-one liquid cooler that features a design that’s very far removed from what we’ve seen in the past.
The H5 SF, or simply the H5 as we’ll refer to it going forward, is a factory-sealed liquid cooling unit that supports Intel LGA 115x and 1366, and AMD AM2, AM3, FM1 and FM2 sockets. It’s also compatible with all Corsair chassis, although certain allowances might need to be made as you’ll see further on in the review.
Non-Corsair chassis owners will need to do their homework first to ensure compatibility, as will those who use motherboards with risers or extra tall heatsinks, since there could be clearance issues at some point in the build or upgrade.
We start our tour of the H5 by looking at the block and pump unit. Under the shiny plastic top cover is the pump that moves the cooling fluid through a series of micro-channels in the copper block in order to pull heat away from the processor. To show this we’d have to ruin our review sample, but what we can show is the surface of the block that, to our eye, has remained virtually unchanged for as long as we’ve been testing these coolers.
The surface is very smooth with only light machine marks, and is perfectly flat on both axis when checked with a straight edge. Like all Hydro-series coolers, the H5 comes with some pre-applied thermal interface material to help speed along the installation.
Fluid is moved to and from the the block through a set of flexible, wide-diameter rubber hoses that appear to be the same ones used on beefier coolers such as the H100i. These hoses connect to the unique radiator that measures only 67mm x 40mm x 57mm, and sports a 120mm GPU-style blower fan (or squirrel cage if you prefer) rather than a more traditional high static pressure case fan.
Corsair has the fan listed as running between 1,000 and 1,800 RPM, while pushing a maximum of 24 CFM of air at 42 A-weighted decibels. Power for the PWM-controlled fan is provided by a 4-pin connector, while the pump is powered by a 3-pin connector.
The fan pulls in cool air from all directions across the motherboard and into the shroud where it’s pushed through the aluminum fin array. The now warm air is then exhausted out the port that faces the same way as the motherboard I/O. To ensure that as little warm air escapes into the chassis as possible, a foam gasket has been placed around the exhaust port.
Included with the H5 is the bracket to secure the radiator to the motherboard, a universal Intel backplate, and mounting brackets for Intel and AMD systems. There’s also a set of four thumbscrews to secure the block and pump, a set of metal loops to secure the block and pump to the stock mounts in AMD systems, a set of risers to secure the block and pump unit to the Intel backplate, and a longer set of risers to secure the radiator to the bracket in order to position it above the motherboard.
Finally there’s a detailed installation guide, and the usual papers showing what’s covered under the 5 year warranty and what to do if warranty support is needed. Strangely absent is a fan splitter seeing how many mini-ITX motherboards only have two fan headers and the H5 will use up both of those, potentially leaving any additional case fans unpowered.
Corsair sent along its mini-iTX Obsidian 250D chassis to ensure we had one that guaranteed support for the H5, but since my daily driver is a mITX system housed in a Graphite 380T, we decided to use it as the basis for our test system and see how the installation went in that chassis as well.
My rig also already housed a Corsair H100i, which uses the same Intel mounting system as the H5, so we decided to install it without removing the motherboard since some potential customers could be upgrading as opposed to building a fresh system.
So, we’ll tackle the upgrade first. After removing the H100i, but leaving the CPU backplate and risers in place, we secured the radiator bracket to the motherboard using the standard mounting holes that run around the perimeter. Then we threaded the three long risers into this bracket, which is where the radiator will sit, directly over the motherboard.
Next we secured the block and pump to the existing mounts with the thumbscrews, and connected the fan and pump power leads to the headers on the motherboard. The final step was to secure the radiator to the long risers resulting in the completed installation shown below.
Not removing the motherboard saved us a bit of time initially, and even with the openness afforded by the 380T, the shorter hoses on the H5 meant that we couldn’t really rest the radiator anywhere while we ran through the installation steps.
This made things somewhat cumbersome at times and a near constant balancing act. This, combined with the lack of working space compared to a standard chassis also meant that we didn’t get proper contact between the CPU and the block on the first try, resulting in 100 degree temperatures during our first round of testing.
In the end we removed the motherboard from the chassis and reseated the block and pump unit to ensure we had even contact. Our little shortcut ultimately wound up costing us extra time, so we recommend removing the motherboard as part of the installation process, even if you already have compatible mounting hardware from another Hydro-series cooler installed.
If nothing else it means you can clean the dust out your system to keep it looking tidy and to keep temperatures as low as possible. Judging by the look of the interior shots of the 380T, I should have done a better job with the housekeeping, so please excuse the dust.
We installed the same system into the 250D to see if there would be any change in temperatures since it has the dual 80mm fan mounts on the rear panel that Corsair recommends for optimal cooling. We won’t let the cat out of the bag yet with regards to any performance differences between the two chassis, but we will show you a shot of the H5 installed in the 250D.
It should be mentioned that in order to install the H5 in this chassis, the optical disk drive tray had to be removed in order to make room for the increased height of the radiator so as always, please do your homework before ordering your parts to ensure the compatibility of all components. Even though Corsair’s website claims that the H5 is compatible with all of its chassis, certain allowances might need to be made as we mentioned before.
With the power and data connections found along the top of the motherboard, we noticed that there was very little room for the hose to fit under the radiator in order to reach the block. While it did fit, we had an uneasy feeling as the unit was secured.
We don’t believe that the hose was pinched or crushed, but it almost looks borderline to us. We also found that the larger connector and thicker cable for the front panel USB 3.0 connection meant that the cable had to be severely bent in order for the radiator to sit properly on the mounts. Again, everything functioned fine, but it was a tighter fit than we would have liked.
All of our testing is performed under controlled conditions to ensure accurate and repeatable results. The test system is kept in a near steady 20°C ambient environment with readings taken before and after testing with a standard room thermometer. The Intel Extreme Tuning Utility (XTU) is used for monitoring and recording temperatures throughout the test process.
Stock CPU settings were obtained by reverting to the default settings via the motherboard BIOS, while a moderate stable overclocked frequency of 4.2GHz was reached by simply setting the multiplier to 40, leaving the bus at 100mhz, and increasing the core voltage to 1.22V.
A fresh, fully updated installation of Windows 10 is allowed to sit idle for 10 minutes after startup to ensure all services are loaded before recording idle CPU temperatures. CPU load temperatures are generated by performing a 20 minute run of the XTU CPU Stress Test to generate as much heat as possible.
|Chassis Test System|
|Processor||Intel Core i5 4690K 3.5Ghz, Overclocked to 4.2GHz/1.22V VCore|
|Motherboard||MSI Z79I GAMING AC|
|Memory||Corsair Vengeance LP 2 x 8GB DDR3 @ 1866mhz|
|Graphics||MSI 280X 3G GAMING|
|Storage||Corsair Force LX 256GB SSD|
|Power Supply||Corsair CX600M|
|Chassis||Corsair Obsidian 250D|
|CPU Cooling||Corsair H5 SF all-in-one liquid cooler|
|OS||Windows 10 64-Bit|
|Corsair H5 SF CPU Cooler Temps|
|Stock CPU Idle||36ºC|
|Stock CPU Load||60ºC|
|OC CPU Idle||40ºC|
|OC CPU Load||79ºC|
Once the Corsair H5 SF was properly seated, temperatures during overclocked load testing stayed well within the thermal limits specified by Intel, which was a huge relief. We also found that there was no difference between the numbers when the H5 was installed in the 250D or the 380T, which could be attributed in part to the foam gasket that runs around the exhaust opening.
The entire opening was covered by the dual 80mm fan mounts on the 250D, but an area of each corner of the exhaust port in the 380T was blocked by the narrower single 120mm fan mount. The gasket meant that what little warm air could have escaped back into the chassis wasn’t able to, meaning temperatures stayed constant between the two chassis.
While the temperature results are great given the compact size of the chassis and the cooler itself, we weren’t fond of the extra noise generated by the GPU-style fan compared to other Hydro-series coolers that use a more traditional case fan.
The short maximum RPM period that seems to go hand-in-hand with all-in-one liquid coolers when a system is powered on is no longer present, but during normal operation the increase in fan speed is clearly audible as the processor is taxed in short bursts. A faint whining noise can be heard when CPU usage reaches even 10%, which can get very annoying for someone trying to concentrate on, oh I don’t know. Writing a tech review?
Those who wear headphone or have music playing most of the time may not notice this, but my simple mind needs silence while I work, and I wasn’t getting it with the H5 due to the fan spinning up every 45 to 60 seconds or so.
It’s no secret among the TG staff that I like my rigs small. In fact my constant gushing over smaller hardware is one of the reasons why we’ve decided to move forward with building a small form factor rig designed specifically to test mini-ITX and micro-ATX chassis seeing how they are more popular now than ever before, so stay tuned for that. Thankfully we were lucky enough to have a rig with which to test the Corsair H5 SF, because it certainly has all the makings of a winner.
It’s easy to install, compact enough to fit into most chassis assuming the proper amount of home work is done first, and it’s capable of keeping mild to medium overclocked chips cool. Make no mistake, this isn’t an H80 or H100, but it is a solid, all around cooling solution for those who want more brawn than a low profile air cooler is likely to afford. Speaking of the H100, we had intended to see how the H100i compared to the H5, however it decided to give up the ghost just before we began testing. R.I.P. good buddy.
The major problem that we found with the H5 is the extra noise generated by the different type of fan. I’ve tested many Corsair Hydro coolers and they have all managed to stay whisper quiet while navigating around the OS, opening Photoshop, viewing pictures, and performing other basic tasks.
The H5 however, continuously spun up and down, making a faint whining noise each time that simply became too much to overlook. Don’t get me wrong, it isn’t a loud noise, but it is audible and frequent enough to make it standout among the usual noise made by other case fans, a second system about 4 feet away, and the NAS sitting beside me.
Also it would have been nice if a fan splitter cable was included seeing how most mini-ITX motherboards only have a couple of fan headers. Thankfully I was able to scrounge one out of my box of odds and ends, but the average user may not be so lucky and will need to tack on a few extra bucks to the total cost of their build or upgrade. It’s either that or say goodbye to any additional case fans.
Speaking of cost, the H5 comes in at a reasonable $75 USD at most major online retailers. That might seem like a lot compared to some air coolers, but with mini-ITX systems your options are limited to begin with. We feel that by spending a little extra on the H5, your chip will stay cooler since we’ve proven time and time again that the majority of air coolers, whether compact or not, simply can’t hang with all-in-one liquid coolers, and the ones that can aren’t likely to fit in a mini-ITX chassis anyway. There’s also the fact that all warm air is exhausted from the system, whereas air coolers leave some of it swirling inside the chassis, resulting in higher overall temperatures.
All in all, the Corsair H5 SF is a solid performer, but the noise is just enough to keep us from awarding it an Editor’s Choice Award. This might not make a difference to some, and it’s those people who will fall in love this cooler, but for those who absolutely need the quietest rig going, this simply won’t do the trick.
Copyright © 2005-2020 Techgage Networks Inc. - All Rights Reserved.