In our continuing quest for the best top-end air cooler it has been interesting to see the heatpipe towers pull to the front of the pack. Coolers like the Thermalright Ultra-120 eXtreme and the Ultra-120, the Tuniq Tower 120, and the Scythe Ninja Plus 2 with push-pull fans have topped our all-time performance charts. They are all very similar in concept, with a vertical heatpipe array supporting horizontal fins with cooling from a side-mounted 120mm fan. More heatpipes have generally meant better cooling performance, at least in coolers from the same manufacturer.

Readers have pointed out that this cooler configuration probably is not the best for cooling motherboard components as well the CPU, and yet we find this type of cooler doing well enough in cooling the systems to consistently rise to the top of the performance charts. Logically a cooler with heatpipes and a fan blowing down should perform better, but our testing of the Cooler Master GeminII showed this was not the case. The GeminII is a decent cooler at stock speeds, but it just does not compete very well or cool well enough at higher Core 2 Duo overclocks.

The answer, according to some readers, is to test more of the down-blowing coolers, as the ones we've tested just can't be representative of a class that has to be better. So here we go again with the down-blowing coolers, and this time we are testing two of the most highly advertised on the market - namely the Scythe Andy Samurai Master and the Thermaltake MaxOrb.


As you can see in the side-by-side comparisons, both these coolers are fairly massive for CPU toppers. The MaxOrb is smaller and is similar in appearance to the older Zalman coolers like the 7000 and 7700. However, the MaxOrb is still large enough to mount an integral 110mm fan. As you will see in the specifications, the MaxOrb is also much lighter than the Andy Samurai, weighing in at a very svelte and moving friendly 465g.

Another significant difference in the two coolers is that any 120mm x 25mm fan should mount on the Scythe, where the Thermaltake 110mm fan is embedded and not changeable. To make up for this Thermaltake has thoughtfully included a rheostat for fan speed adjustment right on the fan, and the adjustment range is specified as 1300-2000rpm with a 2000rpm output of very high 86.5cfm. The on-cooler speed control is handy - at least until you close your case.

Thermaltake MaxOrb
POST A COMMENT

50 Comments

View All Comments

  • Ver Greeneyes - Tuesday, June 05, 2007 - link

    That's exactly what I said a few posts above.. I don't understand this setup. I think the best setup for a top-mounted fan would be if you've got another fan that blows air into the heatsink, which said fan then pulls away out of the case. Reply
  • SurJector - Monday, June 04, 2007 - link

    I suspect the components on the MB do not need that much cooling. Some air, even warm, is better than none, but there is probably no need for much more. Reply
  • MageXX9 - Monday, June 04, 2007 - link

    Does anyone else stop even considering a heatsink when I see those horrible push in clips that are the same type as the retail heat sink fan? I recently built my first Core 2 Duo system and was horrified at what a horrible design. The instructions had in big bold letters that it should only be installed when the motherboard is already in the case, but the amount of force needed to get each one to click, and the way my motherboard flexed made me vow to never use those types again.

    So, if I don't see a screw down design that isn't plastic I immediately write it off.

    What does everyone else think?
    Reply
  • kmmatney - Monday, June 04, 2007 - link

    It's not just you. I was horrified when I built my first Socket 775 system. What a pain those plastic clips are! I'm always afraid I'm going to break something, or break something on the motherboard with the force needed to snap them into place. I've been putting off pin-modding my E4400 because I don't want to go through the hassle of removing my HSF. Reply
  • Imnotrichey - Monday, June 04, 2007 - link

    I just dealt with those clips for the first time, 2 weeks ago on my new system. What a hassle! First, I couldn't get them all in together at once. Then finally when I do get them in, one stays out!! so I try to restart, and then I can't pull one of the pegs out, I felt like I was going to rip the mobo out before I was going to pull out the stock HSF. Luckily, I got it once I turned the case at a certain angle so I could get a good grip. Turns out one of the pegs wouldn't go down all the way. A little piece of plastic was coming up in between the peg, pushing them apart.

    I had to get an Arctic Cooling Pro 7, still had some issues, but eventually got it right. But definetly never want to have to fool with those screws again :)
    Reply
  • sofarfrome - Monday, June 04, 2007 - link

    ...what was ambient temp during this (and other) tests? Everytime I look at the chart that compares 22 or so HSFs I see where 3 products I use always are at the top of the list (Tuniq, Scythe Ninja RevB, and now the TR Ultra 120 extreme). However, obtaining the temps Anandtech claims at 1.5875vcore is a little difficult to believe. That must be one hella cool running x6800. Reply
  • Wesley Fink - Monday, June 04, 2007 - link

    The ambient room temperature is maintained at 20 to 22C, which is 68 to 72F. We measure ambient room temp before we begin any temperature tests. In the summer we have to turn off almpst all the equipment in the test room to keep the temperature from rising during the tests.

    The fans used with this 3 top coolers definitely improve the cooling with these heatpipe towers. You might want to refer back to the original reviews.
    Reply
  • DaveLessnau - Monday, June 04, 2007 - link

    quote:

    We do not use auxiliary fans in the test cooling case


    If I'm reading that correctly, that means you tested without a case fan. This is definitely a problem. Without a case fan, the only way to get hot air out of the case would be because of overpressure. With nothing moving air into the case, there'd be no overpressure and thus no heat exhaust. Properly oriented side-blowing heatsink fans would provide some exhaust, but the down-blowing ones wouldn't be able to do that. Essentially, without that case fan, this test is designed to cause down-blowing heatsinks to fail.
    Reply
  • Wesley Fink - Monday, June 04, 2007 - link

    The large area behind the CPU is perforated in the test bed case so air can definitely escape due to heat rising and gravity flow. We just don't use a case fan to push the air out. It also seemed a possibility to us that we were not exhausting air as well with the down-facing fan coolers, so we also ran a few tests with the case on the side and the side (now the top) off. Cooling performance and overclocking did not improve at all.

    We are looking at all your suggestions to incorporate the best ideas in the new cooling test bed.
    Reply
  • lopri - Tuesday, June 05, 2007 - link

    Incredible argument. (umm.. Gravity?) Are you suggesting that we can do away with the probably single most important fan in ATX design philosophy? Reply

Log in

Don't have an account? Sign up now