Swiftech H220-X Liquid CPU Cooler Review

Status
Not open for further replies.

xFolterknechtx

Reputable
Mar 20, 2015
2
0
4,510
A little bit strange these temperature results.

I think I remember the H220 beating the NH-D15 temperature wise in other reviews, not by a big margin but outside of the 1-2°C error margin.

"It’s very close to being flat, but we can still see ripples in its reflection."

Could it be the case that your CPU sample (lid) and H220 base arnt "compatible"? Did you try reseating, turning the cooler base 90/180° in either direction?
Or is the lid of your CPU totally bend and a prime candidate for lapping?


Also please give use dimensions/weights in the metrical system. I thought this was a tech site and not some farmers supply shop *sigh
 

rubix_1011

Contributing Writer
Moderator
With this system, you wouldn't want to simply just add a graphics card into the mix without also adding another radiator into the loop. Given that the TDP of most mid-to-high end graphics cards today hover in the 150-250watts TDP in addition to the 85-120watts of a CPU, you're looking at a 240mm radiator that cannot dissipate that much potential heat at full load on its own.

The pump in the H220-X is based on the Laing DDC used by Swiftech as the MCP350/355/35x line of pumps rather than an unknown pump used by all the other closed loop coolers on the market. It also doesn't use an aluminum radiator like the cheaper, closed loop coolers - the H220-X is brass tubes/tanks and copper. This means that if you expand the loop, you don't have to worry about galvanic corrosion by introducing mixed metals into the loop together.
 

Eggz

Distinguished
Hasn't this thing been out for a while now? And isn't the x61 a 280mm? That's like comparing the H100 to the h110. It seems like a misplaced comparison for deciding "better" or "worse," which presumes you can fit either a 240mm or 280mm in your case. Comparing those seems more appropriate for informing a case purchasing decision - i.e. "Should I get a case that first a 280mm or just a 240mm?"
 
this is the first review of the swiftech h220x i've seen where it didn't slaughter everything else against it.

something strange happening in your test system. probably air in the lines or a malfunctioning pump.
 

endeavour37a

Honorable
I have read pretty much every review of the H220x and something seems a bit off with performance results with this one. They were also testing against the same hardware as you did here, oh well.
The H240x seems to perform just slightly better than the H220x even with a larger radiator, but would be the thing to have if adding a GPU block. I would think anyone who planned on adding a graphics card would have the common scene to add more radiator area also. It's about the pump, one strong enough to handle a couple blocks and radiators is what this unit is about. Most AIO have just enough pump to take care of the CPU block, unless one wants to spend significantly more on a open loop kit.
With that said, I feel anyone planning on adding their graphics card to the CPU would be hard pressed to find a better value in performance/dollar.
 

Crashman

Polypheme
Former Staff
That kind of thinking gets us motherboards that don't line up with case standoffs. ATX was based on inches. Also it's easier in the US to measure clearance in inches because that's how our tape measures are scaled.

You're right, we really need to put this up against a 2x120mm unit, other than the Zalman which is an alternative design. I'll get right on that.

 

Eggz

Distinguished


Maybe I should have prefaced by first saying that the review was informative and did include the relevant comparisons, as you point out. I guess my intended point was that the 280mm seemed like the high schooler playing football against pop warner kids - not necessarily better, just bigger.

That aside, though, the Zalman's unique pipe design seems to be playing smarter, not harder, given that it keeps up with the x61.
 

PEJUman

Distinguished
Apr 27, 2008
58
6
18,665
Also please give use dimensions/weights in the metrical system. I thought this was a tech site and not some farmers supply shop *sigh

I thought tech enthusiast knows how to convert imperial to metric units... or maybe you're a farmer? :p
 

laststop311

Distinguished
Nov 8, 2010
281
0
18,790
If you have a case with good circulation you can't beat huge air cooled radiators. I use a noctua nh-d14 and a triple slot arctic accelero gpu cooler inside a fortress ft02 case. the 3x 180mm intakes move so much air that even with all the fans on dead silent low speed my temps remain excellent. Always more noise with water cooling and more money.
 
Confusing??
Why do I always find CPU cooler (air and liquid) comparisons so confusing? I have access to various coolers so can make my own comparisons but I usually end up staring at these charts like they were in a different language.

For example, the Noctua NH-U12S spins at 300RPM in idle so is completely silent and barely audible under load for an i5/i7 at 4.2GHz. When I compare online I see max fans or other comparisons which I find difficult to relate to the real world testing I do.

There's also the question of whether the pump is noticeable or not since the dB measurements aren't directly comparable (annoying buzzing can have the same "loudness" as large fans that are barely noticeable).
 
1. This cooler's been out almost a year..... so there's a lotta stuff to compare with. Since there's a hint of a follow up, \I'd like to make some suggestions

2. I don't get this part.

Only the most die hard liquid cooling enthusiast could enjoy constantly checking coolant levels, topping them up, and flushing out algae growth when their coolant isn’t “toxic enough”. Regular users, including part-time enthusiasts, are better off with a system that doesn’t have evaporation holes or air exposure.

I don't know anyone who does this. The Swiftech, just like every CLC uses an "engineered coolant" designed specifically for PC water cooling. My coolant level, also using an engineered coolant (custom loop) is exactly where it was 20 months ago, I have never needed to top it off, never needed to flush out algae. Evaporation holes ? air exposure ? Is this perhaps a reference to the 0.5 to 1.0 inch of air peeps sometimes keep in there reservoirs ?

3. Some reference to the materials would have been instructive, especially the lack of mixed metals in the loop. Would also have liked to see power draw to get an idea of just how much oomph each pump had.

4. One expects variations but this is completely out of line with everything I have seen before. The Swiftech H220-X is normally about twice as loud at full tilt than the D15 at full tilt and about 5C cooler than the Noc.

5. Another thing about that test that I would rethink is ... not loading the GFX card. This seems perfectly logical for testing 240mm coolers against 240mm coolers but gives air coolers a huge advantage. A properly installed OLC / CLC will have the radiator cooled by incoming ambient air and is little affected by interior case temperatures.

Definitely not so for an air cooler. One test showed 4C advantage to the H220-X over the D15 at both stock and 4.2 Ghz..... the Zalman Reserator lost by 6C in TPUs test....In another it had 2C at stock and 6C at 4.4 Ghz over the Noc. Again, one expects difference between tests but this stands out. As the others all show pretty much the same thing, I think something was amiss.

6. As was said before comparing the 240 mm H220-X against the 280mm Kraken is like comparing a Hyper 212 w/ 1 fan against bigger heat sinks with 2 fans. I could understand if there was no H240-X but there is. The H110 - GTX would also be nice it being one of the newest ones out there. In the H240-X review on the same site as the 2nd review referenced above, the 280 mm H240-X for example:

- 280 mm Kraken X61 lost by 5C and 10 dbA
- 360 mm Thermaltake Water 3.0 lost by 4C and 8 dbA.

7. Assuming this was a test rig used before for testing .... and I'm therefore assuming fans on top were exhausting air. After switching the fans to push air into the case so that cooler outside air is used as per manufacturer's instructions, how many case fans were blowing in / blowing out during the test ?
 

rdc85

Honorable
I would like to ask for some additional test using some AMD power hungry CPU..

What came to mind is there not much heat generated so the cooler perform just so. so.. It may different if there more heat generated by cpu..
the cooler may not useful for general using since noctua is still better..
but it my have some use in some market (AMD power hungry cpu maybe)
 

Crashman

Polypheme
Former Staff

It's been here a while, but when it showed up I was told it was time to change test platforms. So, it got put on the backburner with other coolers over the first few months of 2015.

I've personally owned a few coolers with vent holes. My coolers are poisoned with Ethylene Glycol and don't get algea, but I still have to top them up about once a year.


Thanks, I finally remembered to do that with another upcoming cooler :)

I didn't believe the numbers myself so I rechecked the installation, the coolant, and the numbers. Second test was consistent with first. There could be an internal interface problem between the core and heat spreader though.

But what would it do with the case? Most cases support radiators only on top, this case was chosen in part because it represents a "generic" layout. I thought the problem might be low liquid, so I checked that. I thought it might be a paste issue, so I checked that. I thought it might be a pressure issue, so I opened the side to check that too.

I have to admit the size difference slipped past me until the test was submitted. After checking the things above and writing the thing up as is, I looked at the specs and said "OMG it's a different size". Aside from the oversized Kraken, the only reason for the performance deficit compared to other coolers might be the size of the vent holes and the fan's proximity to them. I'd have flipped the fans over and put them under the cooler, if possible, but the H220X design doesn't permit this.

Kind of going at this question from the third degree right? I mean, I did get the thing back from my tester with numbers, check for installation issues, retest, open the case, re-test, and re-write the description. Anyway, there are 4 fans blowing in and 1 blowing out, two intakes by Swiftech design and two intakes by the case design. Opening the case was already cheating but I did it anyway just to investigate the pressure issue, as mentioned in the article. And so I'll say a second time, I think the most likely "fixable problem" was that these fans are factory configured as intakes, putting their fins in close proximity to the case's integrated fan grill.

 

Wexed

Reputable
Jun 8, 2015
1
0
4,510
One thing I would like to see in future tests would be to have all these radiators tested with a standardized fan like noctua NF-F12s. For one it really helps in reducing the noise of these systems, second we'll have a standardized air pressure for the radiator and can compare how effective they really are.
As much as this idea destroys the idea of unboxing test or initial comparison between units; it is such an easy conversion and a rather cheap one that it can, depending on how units fare, change one's purchasing decision.
 

rubix_1011

Contributing Writer
Moderator
I've personally owned a few coolers with vent holes. My coolers are poisoned with Ethylene Glycol and don't get algea, but I still have to top them up about once a year.

There shouldn't ever be a liquid cooling/full watercooling system that is left open and vented to the ambient air...there is no point to this. I've also never heard of any form of liquid cooling that is, by manufacturing design, developed to have open venting.

1) Leaks, coolant loss to evaporation or transport.
2) 'refilling' or topping off coolant means you run the risk of diluting the existing coolant's anti-microbial properties to the point where it can support growth, if you aren't careful.

A liquid cooler isn't a fish tank. You don't want things living in it and you don't want sustained exposure to ambient air. Reservoirs have O-ring G1/4 capped fittings for a reason.

If you are having to 'top off' a watercooling loop once a year, you are losing coolant/water some how...either evaporation (due to exposure to ambient air) or a leak. Or, you just had a lot of air in a loop that wasn't correctly purged in the first place.
 


1. It's very hard to compare results from different sites because test rigs change. An air cooler vs water cooler comparison will suffer....if GFX loading changes....

Test Rig 1 - Nvidia 730 (23 watts)
Test Rig 2 - Twin 290x (310 watts)

I think it's going to be obvious that the air coolers will be hampered by the extra 600 watts of "case heaters" boosting up inside case temps where as the water cooler gets to use nice cool outside air. So that's why I figured that seeing previous results might have caused a rethink in test conditions.

2. Given the aesthetic draw, while purists still cling to DW for that extra half a degree, I think most folks today used engineered coolants. Reasons to top off closed loops.

a) leaks
b) replacing volume lost be releasing entrapped air
c) water absorption.... plastic tubing will absorb a certain amount of water, Tygon 2475 is one known for it's low % of absorption.

3. Some nice pics here on what happens when a loop has mixed metals
https://martinsliquidlab.wordpress.com/2012/01/24/corrosion-explored/

Pump draw will give an indication of how much ooomph and how big a loop you might use it on.

5. As above, I think the lack of load on the GFX card puts an added strain on the air coolers. One think I found absolutely necessary with the H240-X .... granted this one was modified ..... I had to set the case on it's rear and get all entrapped air out of the system which involved some tilting ..... a bubble caught somewhere can reduce flow significantly.

6. I'm still wondering where these vent holes are..... there was a bleed screw but it has since been removed from the design. As for the fans..... based upon Martins testing, push seems to work best up to about 1500 rpm and pull works better from 1800 rpm and up..... in the middle it's a tossup. One thing I have found with my test rig, you can get a whopping decrease in performance from inlet air restrictions. Examples:

On the Enthoo Primo, I saw a big increase in Delta T across the bottom radiator (thermal sensors on inlet and outlet of each rad) when I stuck a book under the case (one at front and one at back) which allowed more air to "get under" the case..... That difference tripled when I removed the air inlet filters. Air coolers don't suffer this dilemma.

7. Looks like the inlet air restriction was something you also suspected in the test. I left the grilles, removed the filters. We have forced air HVAC so the whole house gets filtered, dust isn't really a problem. I have 14 fans blowing in (10 on rads) and 1 blowing out.

If ya curious about case and air restriction effects, I'd be happy to run some tests for ya .... test rig has 6 thermal sensors (4 water / 2 air), digital readouts for each, rpm monitors for all the fans and power meter to read system draw.



 

endeavour37a

Honorable
Ya know Martin is not in business anymore, what a loss indeed. The thing is, if he did not say it then it is questionable, if he did then you could take it to the bank. I think he is persuading another hobbie, perhaps buggies or something. He sure had a great set up and loads of great equipment, most of all the respect of the WC community. Perhaps it would be nice for someone to take the gavel and run with it Jack, you seem very passionate about this indeed.

I read many of your post and am impressed with your focus on the point at hand without being distracted on meaningless tangents. Just sitting here trying to follow this thread and what it is doing, thinking on who will read it and benefit. Funny thing about a lot of readers here, they come and ask if the 780 is a good card and will it do this or that or beat this card or the other one, yet the review is posted to read right here where they are coming for answers. We fill them with our ideas, performance charts and links galore to vindicate our point of view. Yet the basis of everything is knowing how to ask the right question, often missed by the OP and discussions evolve about the relevance of how a card looks in a windowed case.

I personally don't care for this review or the one about the Deepcool product, but I am not going to hammer on Thomas about either one of them, they are what they are. Water cooling is different than CPU's or graphics cards, different animal that needs revues differently. CPU and GPU you plug in and see what comes out, water cooling consists of parts all working together that can be looked at for what it really is, a pump is not a pump. So what do you guys expect for this guy? Something on the level of what Martin did on the original H220?

You think most readers would even understand that level of technological dessication? Maybe most readers understand temps, sound levels, price, looks (ya) and value. Let them roll with it, if it floats their boat why are we beating a dead horse here. I would like to see water pressure, static fan pressure, and everything about the material in the system, but it is what it is. I mean nothing by this comment to anyone, no argument or anything, just like the thread and nothing else to do now.

I think Tom's Hardware should rethink reviewing water cooling AIO's, not changeling Thomas on his test results as per the criteria set to him by this site to conduct his review. I feel something is wrong here as you do but what benefit to anyone to raise this from the dead, let's fix how reviews are done on water. You guys know what you want to see, work together to make them better for everyone, just a thought.
 
Martin still pokes his nose in over on OCN .... I spent a lot of time trying to take his results and make easy-to-use tools out of them. Martin chimed in on this one

http://www.overclock.net/t/1457426/radiator-size-estimator


Others have tried to fill the void, not adequately so IMO.


Tho I have always relied on these guys too.

xtremesystems does great CPU WB roundups
http://www.xtremesystems.org/forums/showthread.php?290225-Radiator-Round-Up-2015

xtremerigs does nice GPU waterblock roundups
http://www.xtremerigs.net/2013/10/03/nvidia-gtx780titan-water-block-roundup/
 

endeavour37a

Honorable
Lots of great valuable info in all of those links there. WOW, a whole other level than this review here. So what are we doing here to adjust things and where are you going with your knowledge? It's a hard thing indeed to understand the human concept of reality and what one requires to transverse the mind in creating a common level of communication to address the question of understanding the basic requirement of a solution to any inquiring person. Just trying to say your research and results are just great reading for someone like me, yet temps and sound are about as far as most readers here can absorb or even want to know. They present a simple question and understand a simple answer, what is best.

WOW, you got some great stuff on water, it's even over my head unless I do some serious thinking about the math! People are inherently lazy, want a link cause they are to lazy to search, where does that leave us? Giving them simple answers to their simple questions. You have done the math, looked at specs, looked at material, so you know (we know because we look at this differently than what is here). When we conclude ourselves the H220/240X are the best AIO on the market today for any price how do we resolve our conflict with this review?

We don't, why should we, who would understand why? The one trend I dislike here as I have seen on a few threads is gang tackling and questioning the integrity of the person who present this. If we question his honesty then let us nail him to the wall, if not then do not question his results as he has presented them. His method of testings perhaps is in question, I trust every word he is saying on how he tested, they do not add up with what I have read but that is not even relevant here. You believe his findings or you do not, done, no need for challenge or argument. If you do not then so so clearly, he is wrong.

Jack, I would love to see you review the H220X, properly against every dam thing you cant throw at it. Why not? I think it's the best thing since sliced bread, Martin says so. So how can we do this, Jack's review, why not? Perhaps not to TH standards or inherent misunderstanding of what AIO are all about but something that explains in plain English what these parts are are about. What the pump and block are all about, the radiator and fans are just there, it's the pump and block.

Perhaps do a H240x to have something different, why not?
 
Well you have to remember that each reviewer's conclusion is going to be based upon the test they did and the things they observed. I have read the author's memory reviews and they are very well done, informative and professionally written.... with other memory articles I have read here, I feel like I am watching Game if Thrones and the cable is going in and out every 10 minutes and I am left wondering how we got from here to there. I have asked for more info.... wud love to know whose modules are inside and was given a logical answer ...would still love to see min frame rates and SLI / CF conditions when examining system memory impacts on gaming tho.

It's quite a conundrum doing reviews in how one approaches them. Do you ....

1. Read everything out there and look for what other peeps have observed thereby taking all potential pluses and minuses into account. or

2. Go in cold so as not to have your conclusions potentially colored.

Do you use the TIM that comes with the unit or same one for all.

Do you retest with all same fans ? Frankly, I'd stio reading if ya had to go out and spend $60 on fans to make a $150 cooler perform better.

Ideally, I like to start with option 2 above and then go back and test with what I pick up from option 1 in mind. After doing that rad calculator thing, I was sort of puzzled as to why my 280 rad had a 30% bigger Delta T than my 420 rad (shuda been other way .....as I'm testing rpm accuracy and other potential causes, a fellow Primo case owner noticed that after he built a rolling case stand his temps improved.... he concluded that the air inlets at case bottom weren't big enuff. I confirmed this, by using a technologically advanced approach .... I put 2 text books under front and back of case....and now my 280 was 40% better than the 420 ????

Then I took the grilles and filters out top and bottom and I picked up 4C ! And the 420 wa sbigger than the 280 as one would expect..... the dimpled case top grille was a big deal..... bigger holes fro version 2 methinks.

Today there was a review on the Acer Freesync monitor which didn't pick up that Freesync disables the overdrive feature .... kinda important on a gaming monitor. So after I completed my draft, reading about this in other reviews would have sent me back to the work bench as it's kind of an important issue since it's been reported that it will take a updated driver, firmware upgrade and possibly more to correct.

It was indicated that the sample sat around for a while.... and a "vent was mentioned so i wonder if it's the current model. The bleed screw was removed I think because stress cracks were observed in the acrylic reservoir and it might have been considered a potential point to focus any stress (google notch stress test or Charpy test). If you have ever "cut tile" by making a light score and whacking it, you have seen it in practice.

I forget what the test is called but at a conference we were given a test. The task was to rank 25 things from most important to least important as "things to have" if you were standing on the moon. I remember duct tape was on the list which made me laugh (more on that later)... we were 1st asked to rank the items individually, then in groups of 2 and then in groups of 8. The most common error was air tank being ranked 1st. I actually had that correct as before you had a chance to take a breath, w/o a pressure suit your body would explode. They way they paired you off tho....w/o us knowing, they paired the top quarter scores with the most correct answers with the bottom quarter lowest scores and put the middle quarters together. In every single case, the pair's score on the retest with 2 peeps was better than all but a few of the individual tests. The guy who got the highest mark, did better having the lowest score guy sharing his thoughts. Point being, even the smartest guy in the room can learn something from a guy who might not have had a good grasp on the subject.

So looking now with the author's explanations .... I gotta think this vent port might have affected the results. Methiunks if it was opened...or perhaps not tightly closed while in storage, some fluid may have evaporated and there was a bubble stuck somewhere. I pointed out that the no GPU load thing seemed to tilt the results but OTOH, taking the side panel off probably eliminated most of that effect. Just looking at air cooler comparisons, we side wide results but again, I'd lobe to see someone take an Enthoo Pro with (8) 140mm fans, shove in PH-TC14-PE and test it.... then while maintaining the same conditions take 1 fan at a time outta the equation and restest. Then flip the top fans (exhaust) and change them into intakes and see if that changes anything.

As for your request to test the H240-X myself ..... here's my son's build. He's 1 year outta college, finally landed a job that he could use his degree for and lived at home long enough to afford the components before he moved out ...so I hadn't had a chance to do anything :).

LL


Can't just yet anyway....there's something afoot in that the MoBo control software is not reading the rpms..... I narrowed it down to either PPCs fudged something sleeving the wires or the splitter is a dud....Swiftech folks sent a new splitter.... and we'll check it out meybe next week. Other issue is I wanna get outta the way first is the typical "top card in SLI is 10C hotter than bottom"....despite the fact that he has 3 top, 2 bottom, 1 front and 1 rear fan. I expect it will be solved by adding a 120mm fan on the back of the HD cage (usually does) but he's been enjoying his bachelor pad too much i think to get around tackling it. he hasn't even OC'd the CPU yet and the build is about 5 weeks old. One conundrum here is what to do with the 3rd fan spot on top which the 240-X doesn't use....

I'm anxious to see follow up tests .... given the explanations provided, I think some of the variables could be removed and we'd see more of an even keel.... one test I always find useful is side by side with rpm set at the same noise target.
 
Status
Not open for further replies.