Aquacomputer Airplex Modularity AMS Copper 360 Radiator

Posted: May 12, 2012 in Radiators
Tags: , , , , , , ,

Thermal Testing

I’m using my new V2 radiator test bench for testing which nets significantly lower watts dissipated numbers than my old open air testing.  This new bench is both insulated and shielded from getting any external cooling help and likely more accurately simulates a closed case condition.  The tested parts are as follows:

Test Specifications:

  • Temperature Probes: Dallas Digital One Wire DS18B20 probes.  These are good to about .2C absolute accuracy in normal water cooling temp ranges and have a nice fine .0625C resolution.  Also since they communicate digitally, you can string the power, ground and Vdd wire in series between all the sensors limiting the amount of wires significantly.
  • Pump: Swiftech MCP-35X2 at 40% PWM.  Simulates medium pumping power and results in roughly 1.5GPM +- depending on radiator restriction.
  • Block: Danger Den MC-TDX block
  • Heater: Standard Aquarium 300W Heater with safety switch soldered in the closed position so heat remains on regardless of temperature.  This requires removing the heater element from the glass tube, soldering the two tabs together and putting it all back together.
  • Heater/Reservoir Bath:  I fabricated this from 1″ schedule 40 PVC with a T and elbow to 1/2″NPT threaded fittings and then used NPT nylon barbs to connect tubing.  The cap I had to use a 1-1/4″ threaded cap and turned it in my lathe to fit the exterior of the 1″ T fitting.
  • Insulation was a combination of 1″ and 1/2″ pipe insulation cut to fit.
  • Tubing is 3/8″ ID x 1/2″ OD tubing. Koolance 3/8″ barbs are also used as the test standard for thermal testing.
  • Case Material – 1/2″ x 8″ Pine it’s a little over 24″ wide to barely fit 4x140mm rads.  The bench overall height is about 18″.
  • Inlet port was fabricated from a 4″ flange material making an ID opening of 4.540″ Diameter.  I shaved the threads out of it and tapered the inlet on the lathe for a smoother inlet.
  • Acrylic panels are all .100″ thickness and I dado cut slots into the shelves for them to fit into.
  • Current Fan Controller is a Scythe Kaze Master, fed by a Koolance SPD-24 to slightly overvolt fans to 2200RPM.
  • Heater Control is via a 3A  generic Variac, although 5A would be better for higher heat loads.  I am using the variac to dial in the Watts into the variac to 100W, 200W, and 300W for each RPM.  5Watts is then added for the pump heat minus variac heat, so each test is targeting approximately 105W, 205W, and 305W.
  • For Watt Metering – P3 Kill-A-Watt and just manually observing and correcting wattage is used.  Wattage normally does not vary by more than 2-3 Watts.
  • Fans- Titan Kukri H PWM fans – I picked these fans because they had a good RPM range and I thought would better represent 25mm fan performance over using 38mm fans.  They do have a more dense 9 blade fan design similar to the Gentle Typhoons. I’m not using them because I think they are superior in noise over other fans, I am simply using them because I can run them from about 650RPM clear up to 2200RPM to get a good broad range of RPM levels tested with a single fan type.  They seem like a pretty good fan, but I wouldn’t suggest they are superior over other 25mm fans.
The radiator is installed into the radiator cabinet with only the bottom three fans push condition.  The fans are all permanently fixed to the acrylic template below with a foam gasket so the fans are installed, run, and logged exactly the same on each radiator. The system is filled with distilled water and allowed to bleed and fans allowed to warm up.  Then the variac is turned on to apply heat load and dialed in to 100W and the fans are adjusted to 640RPM.  The Crystalfontz is then used to monitor fan RPMs and adjusted to a good stable level and all 17 of the thermal sensors are also turned on to start recording air in, air out, water in, water out over those 17 sensors and the fan RPM.  All of these are then recorded each second and allowed to run and log temperatures for approximately 1 hour while carefully controlling ambient temperatures.  After about 30minutes the data log is brought into excel and reviewed for stability and ambient results.  After the system reaches stability while ambients are nice and level, the data is saved and trimmed and an average of the stable data is averaged.  This averaging includes the average of the fan rpms at the time of the test as well as all the temperatures.  That data is then copied to the below chart to calculate the results.  At that point the radiator fan speeds are increased to 1400RPM and heat load to 200W and the test is repeated.  Next the fans are increased to 2200RPM and heat increased to 300W and data logged and extracted again.  The most critical and difficult part of this is maintaining a constant ambient temperature.  It doesn’t have to be any one number, but it needs to hold constant with no more than about .5C rise or fall in any 30minute period.  This is why logging is used to evaluate and continue logging data if needed to collect a good data point.  Simple things like someone opening the front door for a minute will cause a sudden drop in ambient and the test has to be redone. In a nutshell, I am spending a great deal of effort to review logged data and extending the data logging period as needed to ensure a good stable ambient temperature which results in more time per data point but a much more accurate data point.
Summary of 5 tests:
Since the initial 3 tests came in lower than expected, I actually went back and did two additional points on this test at 1000RPM and 1800RPM for validation purposes for a total of 5 points.  The testing is all the same, just a bit more detailed with two more logging sessions.
Using grey shading, I’ve outlined my range of results previously obtained, and you can clearly see the relative performance. It did it’s best (relative to other radiators) at the very slow speed 600RPM area.  Of coarse you do naturally get more performance with increased fan speeds, but it seems to be more tuned for very slow speed fans.
While performance was very good at 600RPM, I expected more at higher speeds considering the 140mm radiator width, extra thickness, and copper tubes.  Unfortunately, that’s not what I got and I can only concluded that while round tubes can be made to work well, your conventional flat tube/folded fin/soldering type construction method is superior in performance per material.
I even went above and beyond my usual 3 point test method and collected two additional points in between the others and my data points have all fallen within 2-3 watts of the trendline.  You can see how the logged data points compare to the smoothed trendline…no major errors jumping out, seems ok.
It is nice and strong in the 600RPM category, but really we are only talking 6 watts separating the top 11 radiators there, so it’s largely becoming a tie at the low end.  Also the fin spacing is a rather extreme low density, especially when you consider it’s extra wide 146mm width at only 8FPI, so it makes sense to me that the radiator would be optimized for very low speed fans based on fin density.  I just thought it would also hold strong in the higher speed ranges and it didn’t which is likely due to the round tubes just not transferring heat as well or as quickly as more conventional flat tube radiators that have soldering.

Pages: 1 2 3 4 5

  1. Dear Martin,
    Throughout your site, your performance analysis of PC components is stellar. But results for this radiator are truly puzzling. By now, you must have seen testing of this reservoir by PureOC.
    ( )
    And it is very strange that their study contradicts yours. As a heat exchanger, this reservoir shows mundane performance at higher fan speeds in your testing. And these results are entirely opposite those shown by PureOC.
    As you know, if the reservoirs you are testing are equivalent, thermodynamics requires disparity is limited only to the accuracy of measurements and the testing process.
    But your revults versus those of PureOC show extreme disparity.
    Where’s the smoking gun ?? ?

    • Martinm210 says:

      I would have to see their logged data to understand, but one methodology spec is alarming. They not testing for 30 minutes and averaging. I have found with larger radiators and water volumes that you can need upwards of 30-40 minutes just to fully stabilize.

      I have also found that slight rise or drop even if steady can throw results off and typically found it takes at least one trial hour of logging followed by a solid hour of logging to QC and pull out the portion where ambient is level.

      a typical radiator test require a good weekend to test properly.

      30minutes is likely only going to give you results of the warm up.

      Other things I learned is ambient probe location and logging of rpm.

      My v1 bench had the ambient sensors too close to the radiator which picked up radiant heat of the radiator. My latest testing I made sure the bench kept them a good 6″ away with air in coming from a collection port with several probes to average the air temp.

      RPM is the other thing I log because despite your best effort RPM does vary as fans warm up and samples vary as well. My latest bench installed the same exact fan permanently with rpm reading and logging from the same fan.

      Then there are heat load issues. Voltage in you house also does not remain constant and can easily make a 100w heat load vary by 5w or more. I used a variac as a line conditioner and even with all these extra precautions found the results to only be a out 2-3w in repeatability.

      Most rad reviews simply do not spend enough time trying to do testing QC work. I also test over multiple fan rpms and plot out a trendline which is another “error” check.

      Rad testing is by far the most difficult product to do accurately, so there are bound to be conflicting results. Not saying mine is perfect, it’s just another for your information. But you should look for details on the methodology.


  2. kit808chris says:

    air flow through the rad makes a big differance, due to its multi pass, if your drawing air from the hot side to the cool side it will keep the cooler side warmer, the cooler side needs to be in the front line of air flow.

    • Martinm210 says:

      There are actually three rows of internal tubes. I am not certain, but I also believe the flow path interconnects these from one side to the other on the ams making one long left to right path?

      The rad certainly did not come with any instructions to suggest air flow must be one direction. I have also seen that suggested with other rads that was debunked with good testing.

      Anyhow, This was just a loaner test sample so I couldn’t look if I wanted to. In general I haven’t found any radiator prefer flow direction a preference for either water or air when tested, they seem to work exactly the same no matter what orientation.

      Case direction could matter, but that’s due to case heat build up and how cold the rad intake air is.

  3. PaleRider says:

    I am so disappointed in the results for this radiator.
    Especially since I own a pair of these, but have yet to use them.
    This is in fact, only the second review I’ve found.
    There was something I just felt wasn’t right about the review at Pure OC.
    I had no idea what it was, the review just didn’t seem as if it were performed correctly.
    Yours, as usual, I trust. I was hoping you’d get to review this radiator, I just wish you could have done it a lot sooner. I definitely wouldn’t have bought these.
    Even the fans I bought, a set of NoiseBlockers best, are the wrong speed.
    I guess I’ll be shopping for a new set-up. Probably Alphacool NexXxoS XT45 Full Copper 360mm rads and a set of slower NB fans.
    Not sure what to do with these AMS 360’s. What’s copper going for these days?

  4. PaleRider says:

    I guess the least I should do is give these ago.
    Any suggestions as to use push-pull or just push?
    These will be mounted in the top of my CaseLabs T10.
    I was planning on using a dozen NoiseBlocker fans: Already purchased – 12 NB-eLoop B12-4 88.5 CFM, 2400 RPM, 34.29 db. Like I said, they’re probably too much fan for this rad.
    It’ll be cooling an i7-3930K and a pair of EVGA GTX Titan’s. Maybe a motherboard block later.
    Any help would be much appreciated.
    Oh, ambient temp. is 72F and 22.22C.

    • Martinm210 says:

      Differences between rads is pretty small. Push plus pull is always a benefit no matter what rad, but how much is hard to guess and depends on heat load.

  5. kit808 says:

    make sure you have the air flow the correct direction.
    i have mine set up on push

    link to pic

    • Martinm210 says:

      Haven’t seen any worthwhile testing show much if any measurable difference on flow direction so I would worry about it much. My sample did not have any directions or suggestions about flow direction either so I wouldn’t worry about it much.

  6. PaleRider says:

    Did I misunderstand this review? My 2 radiators are just fine, just not the best there is?
    I bought them to go with my ultimate water-cooled case, a CaseLabs T10. It’s the computer I’m building with the best parts I can find.
    I have a dozen Noiseblocker NB-eLoop B12-4 120mmx25mm Ultra Silent Bionic Blade Fan – 2400 RPM, for these rads. Either push and use the other 6 fans elsewhere, or use all 12 as push/pull.
    Reading the review I suppose I’ll just be using 6 in push.
    I have a Swiftech MCP35X2 pump I’ll be using for redundancy.
    I’ll be using either an ASUS Rampage IV Extreme or an MSI Big Bang II, I already own both. I have a pair of i7-3930’s for these boards and may possibly buy an i7-4960X whenever they become available this year. That depends on my finances at the time.
    I’ll most likely add water-cooling for the boards, and I’ll be using a pair of water-cooled EVGA GTX Titans I already own. Or, maybe, the pair of EVGA GTX 690’s I also own. My ASUS Ares II already has its built on hybrid water-cooler.
    All the single VGA cards have Alphacool blocks and the cpu block is an Aquacomputer Cuplex Kryos XT Series – .925 Silver Edition for LGA 2011.
    I was hoping that the pair of these stainless steel and copper AMS rads would be enough for these items, but I have room enough for 2 more if necessary.
    I’m taking my time with these so I can afford the best parts and buy them as I can afford them.
    Right now, I’m down to the fittings as I just bought a new BluRay recorder and already have four 500GB Samsung SSD’s for this build, plus a 3TB Toshiba HDD for internal backup and storage. Also a set of Corsair Dominator Platinum 4x4GB 2400MHz sticks.
    Will these 2 rads be enough or should I buy one or two more?
    Ambient temp. here in my computer room is 72F 24×7. The room has its own very large air-conditioner I bought just for my computer room.
    Any suggestions? As I said, this will be my ultimate water-cooled computer case. The motherboards, video-cards and other components may change over the years but this is the last, finest case I’ll ever build.
    If these rads aren’t going to be up to snuff for at least a decade, please tell me now so I can change to another type as soon as possible. They haven’t been mounted yet so I can sell them on Ebay if I have to.
    It really doesn’t matter, but I bought the case in reverse configuration as I have already prepared an area to the left of my computer table for it.
    If I can get CaseLabs to cut me a panel for a Watercool Mo-Ra Pro 120×9 I may just use it. It sure would be easier.
    Thanks for your help.

    • Martinm210 says:

      They are fine. As with most watercooling parts the differences are very small and hard to measure and we all tend to put too much emphasis on those small differences. Heat load and fans used will have a much larger impact on your result than which brand or model radiator. It’s all good.:)

  7. […] review do not say so good on AMS…. Aquacomputer Airplex Modularity AMS Copper 360 Radiator | (adsbygoogle = window.adsbygoogle || […]