Scythe Mugen 5 Rev.B CPU Cooler Review

  • Thread starter Thread starter Guest
  • Start date Start date
Status
Not open for further replies.
I presume this all falls down to the fan. Replace the fan with its, more than likely, cheap sleeve bearings with a better fan, and the 1 year issues will start to weaken. Of course that will probably increase noise and improve cooling performance... depending on the replacement... but once this is done, you've just spent more $$$.

Sell just the Heatsink for around $30-40, and let us choose our own fan to buy? Then we fall responsible for its noise/performance ratio. Toms, would be nice to see these heatsinks done and compared also with a specified fan of the given size (ie. Noctua NF-A12x15 PWM for a 120mm fan, or other agreed upon model)... just to even out what the heatsinks alone can do (cooling performance, and its own contribution to noise, subject to fan blade design.)
 
I've stayed away from Scythe coolers since the Ninja 2. Had installed it in a friend's rig some years back and within a year, the fans failed 3 times. You also had to install the fans after the heatsink was secured to the mobo, otherwise you wouldn't have enough clearance to fasten it to the socket otherwise, and the retention clips / fans were a pain to install after the fact.

Perhaps Scythe coolers have improved since my experience with them, but based on past experience, I wouldn't recommend them.
 
I have to yell out a "yeah, me too!" to what SHRAPNEL_INDIE said. I would indeed find it very useful to see a test of a group of heat sinks all being tested with the same fan. That would really empower us, the "poor" enthusiasts. It would be nice to know for once that we could overclock to our hearts' content with the potential of not needing to spend double on an AIO for every build.
 
Is there any way you could check the minimum sustainable fan speeds? Just knowing how slow they can go would help in designing a quiet build, especially for the coolers with higher max speeds.

After all, the best cooler for quiet builds is one that has decent performance at slow fan speeds as well as enough headroom to push the CPU when needed. This cooler is quiet, but I wouldn't want to run the CPU at 85 C (61+24C).
 
Each cooler review I've seen over the last year or so just makes that cheap Gammax look better and better. Slowing its fan down, or using a better fan, would likely quiet it down nicely while still providing decent cooling performance.
 


If you replace the fans on this cooler with faster models with similar airflow and noise at a given RPM, it would outpace the Gammaxx 400 in all realistic cases. This is based on the measured values in the article and a 40*log relationship between the delta-T and noise (usually quite accurate for estimating cooler performance).

Based on that model, this cooler even gives the NH-D15 a run for it's money. The D15S would probably turn out ahead, though.
 
I think anybody looking for a quiet fan will probably buy a Noctua U14-S (140mm fan) which moves at least 40% more air at the same RPM, so it can run at much lower RPM. We recently got one shipped from Amazon for $70 including tax ($64 retail). I think it would be a better cooler and potentially, just as quiet. It is low profile and can fit into many cases.
 
I have Scythe Zipang it also takes upto a 140mm fan, i have a noise blocker fan on it now and i think it is one of the most robust cpu coolers of it's time. got it about 9yrs ago and it still fits all sockets including the new Amd AM4 socket. so if my cooler is this good maybe this new one will be even better.
 
Can't see any of the data on any of the carts in today's reviews so comment has to be limited but reviews have to be done in context and today's context is that cooling needs for CPUs are dropping as we find ourselves hitting the voltage wall more and more often before hitting the temperature wall .

And as noise has become a significant "feature" vendors are seeking to put products on the market that address user's wants / needs. The desired feature of low noise and higher cooling performance are competing. You can not increase one w/o decreasing the other. With that understood, a product that excels in one will invariably lose ground in the other. Therefore when a manufacturer seeks to hit a market niche by excelling in a specific category, it must be understood that the other is not high on the design goals. One one looks at a Porsche and says "Geez, it's mpg isn't all that great" or an economy car and complains that it doesn't get to 60 mph in < 4 seconds. Both are doing exactly what was expected of them by their designers.
 


Summary of data:
Delta T @ Max: 55C for CPU, 23C for VRM.
Delta T @ 50%: 61C for CPU, 28C for VRM
SPL @ Max: 20.8 dBA
SPL @ 50%: 12.2 dBA

The 50% SPL is darn close to the noise floor of the meter prior to the distance correction (which is applied before TH presents the level).
 
You're both somewhat in luck! SPCR (linked) do this when they test heatsinks.
1. Test the stock fans to find out their speed range and actual noise, including any tonalities and other nasties.
2. Run the heatsinks with both any stock fan as well as their own reference fan (or fans, depending on what sizes fit).

Then the results are presented in a list based on cooling performance at specified noise levels.

The only issue now is that the man who did all the testing has been moving, and not done any reviews at all in the last year or so... 🙁
 


Benchmarkreviews does this in every review

https://benchmarkreviews.com/28173/scythe-mugen-max-cpu-cooler-benchmark-review/6/

They also have the best detailed TIM and TIM application articles out there and age has in no way reduced there usefulness

https://benchmarkreviews.com/206/best-thermal-paste-application-methods/

If you want to see just how drastic changing fans can be, look here to see what happens when you take the NF-A15's off a Noctua CPU cooler and replace them with slower Phanteks Fans

http://www.overclockersclub.com/reviews/phenteks_f140/3.htm

Noctua NF-A15 = 52.9C @ 1500 rpm
Noctua NF-A15 = 56.3C @ 1200 rpm
Phanteks PH-F140SP = 50.0C @ 1200 rpm

Noctua showed some new "low tolerance" fans at Computex, with the tighter clearance tolerances, science says they should be able to retake the title ... we'll ahve to wait for testing to see if it holds true. Here's silentpcreview.com's Top 20

http://www.silentpcreview.com/article1345-page7.html

Phanteks PH-F140HP/TS
Noctua NF-A14 FLX
Noctua NF-A15 PWM
Noiseblocker B12-2
Noctua NF-P14 FLX
Scythe Gentle Typhoon 120-12
Noiseblocker M12-S1
Corsair AF120 Quiet
Corsair AF120 Performance
Thermalright TR-TY150
Scythe Gentle Typhoon 120-14
Xigmatek XAF-F1453
Nexus Real Silent
Xigmatek XLF-F1453
Noiseblocker B12-PS
Corsair SP120 Quiet
be quiet! Silent Wings 2
Noiseblocker M12-S2
Antec TrueQuiet 120
Noiseblocker M12-P


 
Just thought I'd point out that the list omits a fair number of newer, high-performance fans.

The main ones that come to mind are the ML120 Pro and Silent Wings 3 in all variations. Otherwise, it's about right.
 
It's 4 years old so yes, would love to have newer data. The new Noctua fans, 5 years in the making, would be a welcome addition. But as that example demonstrates, bringing a better design to market and getting into the top 5 is a lengthy process

The ML120 pro has proven itself to be a very quiet fan but needs more performance to crack the top 5. Also at 2400 rpm, it's a design well suited for high fpi radiators that have long since gone out of vogue. I think they'd make a good replacement say for Corsair's H100i, but for case and copper rads of today, I'd be looking for 1200 rpm. Seems the in doubling the rpm however, there should be more cfm output

The Silent Wings are a very good fan, tho again a bit hi rpm for what I would want to use on today's copper rads (8 - 12 fpi) or as in case fans. In a case with insufficient fan mounts where one had to move up to 1500 - 1800 rpm, they would be at the top of the list. But, since we design for a max rpm of 850 under all but stress test loadings, with a top speed of just above 50%, it leaves too little of a control range.
 


I didn't mean to imply that the ML120s belong in the top 5. They're good performers, but have enough issues to cause concern in a lot of builds. Just ask Corsair about using speed control with them. The requirements are as strict as the most problematic of water pumps.

The Silent Wings 3 is more deserving of a place on the list, though. Regarding your particular application, it comes in a high speed and low speed version. All of the PWM models offer reliable control down to the 250 RPM range, and are very consistent from model to model. They're among the few fans out there that offer this wide of a range to work with. The only other ones that come to mind are the iPPC 3000s with a range of 450 to 2800 RPM on average.

It's worth mentioning that the Silent Wings 3 have a very peculiar PWM vs fan speed curve. If you haven't seen it before, it's worth a look:
http://thermalbench.com/2016/09/05/be-quiet-silent-wings-3-120-mm-fan/3/

The low speed models might actually be a perfect fit for targeting 850 RPM as a max in normal situations while still keeping some headroom on tap for stress tests (note: this statement only makes sense after you see the PWM vs RPM curve above).
 


No I didn't think that you implied that either, just mentioning that if it doesn't break the top 5, i want going to be looking at from 6 on down.

I like the Wings, but aside from good performance at higher rpms, which is outta my target area, we have since stopped using PWM. As you mentioned with Corsair, they are problematic when ganged as Darla explains here:

http://www.overclock.net/t/1506812/the-best-pwm-rad-fans-going-to-be-very-exact-spec-needs/30

With the advent of PWM fan Control PCBs, you can gain all the speed control advantages of PWM w/o any of the disadvantages Cost / low speed clicking or hum). The other thing that figures in, we used to have as many as 5 different case models for users to choose from in each price niche but last few years the list has dwindled because all most are choosing Phanteks ...and when the case comes with 3 - 5 fans *and* a PWM => DC hub that controls them. make sit hard to justify tossing them and getting something else. I am a bit concerned about the price creep on them...tho still at $15 hard to complain... used to jump on the newegg specials when they were $10 each but can only recall 1 or 2 in 2017. Another complaint ... buv them in a box and they come with the rubber mounts ... pre-installed case fans ... no rubber mount, tho if ya call and ask they mail them to ya.

The thing about the Silent Wings twas that it's kinda unusual that the DCV and PWM ones differ so widely which makes the 120 and 140 deliver very close to the same 1ir flow

120 is 1,450 rpm
140 is 1,000

120 HS is 2,200 rpm
140 HS is 1,600

If doing an ITX build, haven't dome one in ages, I'd prolly go for the Silent Wings .. right now I'm anxious to see how Noctuas new low tolerance fans perform ... on paper, they sound great, logic says that the reduced space between shroud and housing should improve things greatly.... whether we actually see that in practice remains to be seen and I hop other vendors follow suit.

The other thing I like about the wings is that the PWM and DCV versions were selling for about the same price. perhaps that's a recognition of the current state of the industry where, since the arrival of fan control PCBs, PWM lost its speed control advantage.







 


Maybe for the 112 watt models. There's no way it will work with the Skylake-X models. They need water cooling to avoid throttling at stock settings.
 
Status
Not open for further replies.