Scythe Ninja Mini experience

[Toan]

I recently put together a new PC and thought I'd share my experience and test results with the new Scythe Ninja Mini. This is actually a normal mid-tower build, but I didn't want a heavy, hulking HS on my MB.

I narrowed down my decision to either the Scythe Ninja Mini or the Thermalright Ultima 90I. I went with the Ninja Mini because the fins were farther apart, so it should do better in a low airflow environment. I was hoping to cool my C2D E6750 passively.

Anyway, system specs:
Core 2 Duo 6750 (overclocked to 400x8)
Abit IP35 Pro
EVGA 8800GTS 320M Superclocked
WD Raptor 150G
Corsair HX520 PSU
Antec Solo with 1x Nexus 120mm outtake, 2x Nexus 92mm intake fans

(I mildly overclocked the CPU with a 400 MHz FSB so that it could run synchronously with my 800 MHz DDR2 memory. The E6750 didn't need any extra voltage to run at 3.2GHz.)

NOTE: Unfortunately, the Ninja Mini's lowest fin came into contact with the NB's heatsink, so I bent the Ninja Mini's lowest fin down so it was perpendicular to the motherboard. This made it extend just short of the NB heatsink, so it didn't matter that the fin came down closer the the motherboard.

The IP35 Pro has 5x 3-pin fan headers (SYS + AUX1-4) in addition to the CPU fan header. It can control all 6 fans based on CPU, "PWM", or system temperature. You can set individual parameters for each fan -- they can each run off different temperature sensors / thresholds.

(My AUX3 fan header was acting up -- it would intermittently power down the fan connected to it, then power it back up during normal operation. It did this regardless of which Nexus fan was connected. I switched the fan that was connected to AUX3 to use AUX4 instead, and everything was fine.)

I played around with varying the stock cooler's fan based on the CPU temperature, but it was extremely annoying. The CPU temp rose and fell quickly depending on what I was doing on the computer, so the fan audibly ramped up and down continuously.

(I now have all my fans controlled based on the "PWM" temperature reading. This is the most stable indicator of system activity and avoids the annoyingly quick changes in fan speed. I found the other sensor ["system"] varied too little and too slowly to be useful.)

Anyway, onto the test results. All temperatures are reported in Celsius. These are all "load temps" running Orthos with small FFTs at priority 1 for 30 minutes. I disabled C1E and EIST (in the BIOS) to make sure the CPU was running at 3.2GHz. "CoreTemp" shows the temperature which that program displayed for both cores (they tended to be the same). I also got a set of temps using Abit's uGuru program. (SpeedFan reported bogus temps, and Intel's TAT did not work on my setup -- it reported an error on startup.)

Stock cooler @ 100% fan speed, all Nexus fans @ 12v
Ambient 28.7
CoreTemp 53
uGuru CPU 61
uGuru PWM ???

I forgot to record the uGuru PWM temp with the stock cooler. Oh well.

Scythe Ninja Mini (fanless), all Nexus fans @ 12v
Ambient 27.4
CoreTemp 62
uGuru CPU 69
uGuru PWM 64

Fanless was a bit too hot for my liking. I tried replacing the 120mm Nexus fan used as the outtake fan with the Antec Solo's 120mm TriCool fan.

Scythe Ninja Mini (fanless), replaced outtake with 120mm TriCool on High, intake Nexus fans at 12v
Ambient 27.8
CoreTemp 50
uGuru CPU 59
uGuru PWM 56

The temperatures were much more acceptable, but the TriCool is LOUD on high. So I put the 120mm Nexus back in, and attached the Ninja Mini's 80mm fan on the HS.

Scythe Ninja Mini with stock fan @ 10v, all Nexus fans @ 12v
Ambient 26.8
CoreTemp 46
uGuru CPU 55
uGuru PWM 58

The stock fan on the Ninja Mini is very quiet. At 10v, I don't hear it above the Nexus fans running at 12v. However, running that 80mm fan @ 12v makes it barely audible above the Nexus fans.

For reference, the idle temps are:

Scythe Ninja Mini with stock fan @ 8v, all Nexus fans @ 10v
Ambient 27.0
CoreTemp 25
uGuru CPU 39
uGuru PWM 50

(The CoreTemp reading seems to be a little low, but they are included since the program is more widely available.)

(I mounted the stock fan so that the bottom of the fan clip fit between the Ninja Mini's last and 2nd-to-last fins. This keeps the fan high enough to not interfere with my memory slots. It also ensures the fan won't move up/down very far -- though the fan clip holds the fan pretty firmly anyway.)

I was really hoping to run the Ninja Mini passively -- after all, both the case's outtake fan and the PSU intake fan would be drawing air from the HS. Unfortunately, it didn't work out passively. However, the stock fan works very well, and sound-wise, it's as good as passive. With it, I get noticeably better-than-stock cooling, and I don't worry about bending my motherboard.

[cmthomson]

I was really hoping to run the Ninja Mini passively -- after all, both the case's outtake fan and the PSU intake fan would be drawing air from the HS.

I think with a bit of ducting you can run without a fan on the HSF.

[Lensman]

I think a lot depends on the airflow in your case. On the advice of Pierce from this thread, I tried running my Mini Ninja passive in my Antec NSK2400 case. It works great! I think the airflow design of the case works well with the Mini Ninja. I have both side exhaust fans running and they're 20mm or so from the Mini Ninja I think they're drawing significant airflow over the heatsink. I also have the little plastic fins configured to force airflow from the back vent through the Ninja.

I'm running the Antec tri-cool fans at low.

[daaron415]

Okay, I'm about to go crazy here. I just installed a Ninja mini on my BE2350, that I installed on the Biostar690G and I get so many conflicting temperature reports, I really have no clue what to do. Using CoreTemp, I read an average of 8C for core1 and 12C for core2. Using the bundled hardware monitor program, I read 53C for my CPU, and when I use Speedfan, I read 13C. I have no clue which program is inaccurate, if they're all inaccurate, or if there is something else entirely going on. If anyone else has run into this problem, give a shout out, or if you know what's going on, a little enlightening would be greatly appreciated.


Thanks!

[NeilBlanchard]

Hello,

The low temps are way too low -- the 53C sounds about right.

[daaron415]

Oh, so the mini with the BE2350 should run about 50ish degrees? Thanks for confirming!

[NyteOwl]

I wonder how well the mini would do on a AMD 6400+ passive but in front of a pair of 60/80mm intakes?

[daaron415]

Well it should probably be pretty damn good, because where I live, ambient temperature is near 40C, and I get around 50 running passive, inside a P150, and I am only using the tri-cool fan that comes bundled running at low.

It's kinda higher than I would like, but the comp is stable, so there's nothing to complain about.

*EDIT* Okay, I meant 40CELSIUS, not Fahrenheit %_%

If you can reach even 50C, I will pray for you.

[NyteOwl]

Even in winter my ambient is a lot higher than 40F! By a good 25-30 degrees

If you meant 40C I'm enouraged

[zerok66]

Second post regarding the E6750 and the scyth ninja mini...

see here:

viewtopic.php?t=42893&highlight=

I am using an awfully flowing case (OrigenAE HTPC) using the thermalright SI-128 cooler + 120mm fan.

I have not done any load testing, but initial idle and general running test have shown:

CPU Temps - 25-30 degrees
Core 1 - 12-18
Core 2 - 13-19

REALLY cool... this is a great cooler - I had thought it was awful as my AMD 4200+ X2 s939 CPU in the same case, setup was running at 50-60 degrees c at IDLE in the same case!!

Think the ninja mini is not a great cooler

[Badger]

Second post regarding the E6750 and the scyth ninja mini...

see here:

viewtopic.php?t=42893&highlight=

I am using an awfully flowing case (OrigenAE HTPC) using the thermalright SI-128 cooler + 120mm fan.

I have not done any load testing, but initial idle and general running test have shown:

CPU Temps - 25-30 degrees
Core 1 - 12-18
Core 2 - 13-19

REALLY cool... this is a great cooler - I had thought it was awful as my AMD 4200+ X2 s939 CPU in the same case, setup was running at 50-60 degrees c at IDLE in the same case!!

Think the ninja mini is not a great cooler

Unless you're running that outside with snow falling, you're going to have to add 15C to each core.

[sunspot]

Hey, I'm also running the Mini Ninja (a Minja?) on the Abit P35 Pro board. Just set it up last week. Haven't run the Abit uGuru utility yet, but with Core Temp I'm also seeing approximately 45 degree temps on the CPU (a Core 2 Duo E6750 at 2.66GHz), pretty much full-time (I'm using the BIOS to control the fans). Even under load the core temp never seems to climb above 48, even in a warm room.

I've got the stock Scythe 80mm fan on the Minja, I've got a big Scythe 120mm fan on the back of the case (an Antec Solo) which is very quiet, and two 92mm Antec TriCool fans up front, set to medium & controlled by the motherboard. They aren't terribly quiet, but until the case really heats up the motherboard keeps them turned down pretty low, so they're effectively silent most of the time. I'm running the Antec Phantom power supply - so far I haven't heard its fan switch on.

This machine isn't any quieter than my old system - an Athlon 64 3000+ with a giant Thermalright XP120 heatsink and a Seasonic Super Tornado 300 power supply - but it's much better ventilated, has a much more powerful video card (a GeForce 8600GT vs. an ancient Radeon 9600), and I'm running a 3 drive RAID array. My drives are much better ventilated in this case - the Antec Solo leaves a lot of space between your drives and the fans really push a lot of air over them - and I think there's still a lot of room to tune the fans. I might also try to duct the Minja at some point an run it fanless. There's room for improvement, but so far I'm quite happy with the results.

FWIW, the Core 2 Duo runs 10-15 degrees cooler than my old Athlon 64, which is impressive given that the Core 2 Duo 6750 chip is about twice as powerful and I'm using a smaller heatsink and much smaller fan. According to Intel, the max core temp for this processor is 100 degrees centigrade, so there's plenty of headroom if a fan should breakdown or something.

I think we're gonna see a lot of smaller coolers sporting designs similar to the Minja over the next year - lots of heat pipes connected to a smaller set of radiating fins. Looks like there are some skinny tower designs coming out now too that are just big enough to accommodate a 120mm fan on one (or both) sides, and are packed with heat pipes. My guess is those will become the new standard for cool and quiet operation - they look to be very efficient, and would probably work well in ducted implementations (just hook a duct to one side and all of the case airflow could be sucked thru the cooler - very efficient).

[Jason W]

What programs are you guys using to monitor your temps? For those of you with the E6750 or any Exx50 CPU, I believe the Tjunction is 100 degrees C, not the 85 degrees C that SpeedFan or Core Temp 0.95 report. Therefore, for those two programs, you would need to add 15 degrees C to the temps for accurate readings.

If you download and use Core Temp 0.95.4, it reports the E6750 Tjunction as 100 degrees C, therefore it is accurate.

If my information is not correct, please correct me!

Regards.

[ronrem]

I wonder how well the mini would do on a AMD 6400+ passive but in front of a pair of 60/80mm intakes?

No better than getting rid of the small intake fans and attaching a slow 120mm to the HS.

The whole plus of a passive heatsink is lost if you have a whole lot of small high rev fans. Your main goal in life is to run your computer with a TOTAL of 2 fans or less. That includes PSU.

I would go Full Ninja unless you REALLY need a small cooler (like the old lady wants a HTPC case that's pretty) A Mini on a 3600X2 or a Sempron may do okay but I figure the 6400 + is probably about as hot as AMD makes. Even a full size Ninja would have trouble,though a heavy undervolt job just might cover it.

[sunspot]

If you download and use Core Temp 0.95.4, it reports the E6750 Tjunction as 100 degrees C, therefore it is accurate.

My copy of CoreTemp reports the Tjunction as 100 degrees C.