Katana 3 is one of Scythe’s most economical models. Among the giants and skyscrapers of the brand’s extensive lineup, it stands out by being relatively small, light and inexpensive. But Katana 3 supports every current socket, including 1366 for Intel i7.
July 3, 2009 by Mike Chin
Product | Katana 3 CPU Cooler Model SCKTN-3000 |
Manufacturer | Scythe |
Sample Supplier | Anitec |
Street Price | $26~30 |
Sometimes, Scythe’s marketing copy is so amusing it’s impossible to resist repeating it verbatim:
“The Japanese long sword ‘Katana’ is produced in a unique, manufacturing process, which has been brought to perfection over centuries – until today, only a few blacksmith master have the knowledge to manufacture a Katana. The new Scythe Katana 3 CPU cooler also arises from a consistent further development and differs from other CPU coolers and is a worthy successor to his precursors, Katana and Katana 2.“
Even with the minor grammar issues, it’s a heady brew. Not only does the copy neatly sidestep the question of what could possibly be exciting about a third version of a heatsink, it puts the Katana 3 in the glorious context of Japan’s martial tradition and history. My hat is off to the marketing boys at Scythe.
Katana 3 is one of Scythe’s least ambitious and most economical models. Among the giants and skyscrapers of the brand’s extensive lineup, it stands out by being relatively small, light and inexpensive.
|
Scythe Katana 3: Key Features (from the product web page) | |
Feature & Brief | Our Comment |
Easy installation without any tools The smart and patent-pending mounting mechanism “VTMS” (Versatile Tool-Free Multiplatform System) allows mounting of Katana 3 CPU heatsink without using tools. | Good, and secure enough with the low weight. |
S.P.S. (Pipe Slant Structure) Already developed with the first Katana, the slightly curved shape of the Katana 3 allows both the direct heat transfer from the cooling fins of the cooler as well as an additional cooling of the mainboard components. | It’s a reasonable approach to the task of VRM cooling. |
F.P.S. (Fast-Phase Structure) The various cooling fins on the baseplate allow an additional heat dissipation and optimize the overall performance of the cooler without changing the compact dimensions of the entire system. | It’s not clear whether this really helps, but the term sounds impressive. See text at bottom of p.2 for more discussion. |
Scythe Katana 3: Specifications (from the product web page) | ||
Model # | SCKTN-3000 | |
Heatsink | Dimensions | 94 x 108 x 143 mm 3.70 x 4.25 x 5.63 inch |
Weight | 495 g (without accessory) 15.98 oz (without accessory) | |
Compatibility | – Intel LGA1366 (i7), LGA775, 478 – AMD AM2/2+/3, 939, 940, 754 | |
Fan | Dimensions | 92 x 92 x 25 mm |
Weight | 115 g | |
Speed | 300 ~ 2,500 rpm (via PWM) | |
Noise Level | 7.2 – 31.07 dBA | |
Bearing | Sleeve Bearing | |
Air Flow | 12 – 95 m³/h = 6.7 – 55.55 CFM |
It’s noteworthy that this modest heatsink is compatible for use with Intel’s i7 processors. It is, in fact, compatible with almost every CPU socket type going back a decade. The reference to noise level is probably sound pressure level (SPL) at 1m.
PHYSICAL DETAILS
|
|
The reason I question the supposed merits of Scythe’s Fast-Phase Structure is that heatpipes are phase change devices that work best with high temperature differentials between the evaporator end and the condensor end. The hotter one end gets, the faster the internal liquid boils and turns to vapor, moving to the condenser end. Conversely, the cooler the other end gets, the faster the vapor turns back to liquid and moves down to the evaporator end. The extruded aluminum piece (FPS) is on the hot or evaporator end. Rather than being transferred up into the liquid of the heatpipes, some of the heat will go up to the FPS and be dissipated by airflow. This would reduce the total amount of heat going into the heatpipes, and it would have the effect of slowing down rather than speeding up the phase change process in the heatpipes. Have I missed something? Or has Scythe’s marketingspeak gone one step too far?
Having said all that, if the mass of the FPS was converted into more fins on the main tower stack, the difference in CPU cooling would probably be marginal. The improvement in phase change speed would be offset by the absence of cooling via the extruded aluminum piece. It’s just that Fast-Phase does not accurately describe the role of the secondary aluminum piece.
MORE DETAILS
|
INSTALLATION
The most critical aspect of installation is for the heatsink
to be securely mounted. The more firmly it is installed, the better the contact
between the heatsink’s base and the CPU itself. It’s also less likely to fall
off. Ease of installation is also important — a simple mounting scheme
means less time spent installing, and a reduced likelihood of screwing up.
The pushpins employed here for Intel 775/1366 socket boards work well enough. Access to the pins is unimpeded by the heatsink itself, and the total weight of 530g with fan and mounting hardware does not seem onerous for the plastic pins. The heatsink seemed quite firmly secured in the end.
For AMD processors, there is only one way that the Katana 3 can be mounted, but as most AMD boards have a north-south orientation for the heatsink retention frame, the fan ends up blowing in the right direction, towards the back of the case where there is usually an exhaust fan, rather than up towards the PSU intake (in a typical case).
|
TESTING
Before thermal testing, we took some basic physical measurements.
Scythe Katana 3: Physical Measurements | |
Weight | 530 g (with fan and mounting brackets) |
Fin thickness | 0.31 mm |
Fin spacing | 1.72 mm |
Number of fins | 48 |
Vertical Clearance | ~55 mm (from motherboard PCB to the heatsink’s bottom fin) |
Horizontal Overhang | -30 mm (that’s 3cm in from edge of heatsink to the top edge of our test motherboard) |
Comparison: Approximate Fin Thickness & Spacing | ||
Heatsink | Fin Thickness | Fin Spacing |
Scythe Ninja 2 | 0.39 mm | 3.68 mm |
Thermalright HR-01 Plus | 0.45 mm | 3.15 mm |
Noctua NH-U12P | 0.44 mm | 2.63 mm |
Noctua NH-C12P | 0.47 mm | 2.54 mm |
Prolimatech Megahalems | 0.50 mm | 2.00 mm |
Xigmatek HDT-S1283 | 0.33 mm | 1.96 mm |
Cooler Master Hyper N520 | 0.38 mm | 1.72mm |
Xigmatek HDT-S963 | 0.33 mm | 1.72 mm |
Scythe Katana 3 | 0.31 mm | 1.72mm |
Thermalright Ultra-120 | 0.45 mm | 1.42 mm |
Testing was done according to our
unique heatsink testing methodology, and the included fans were profiled
using our standard fan testing
methodology. A quick summary of the components, tools, and procedures
follows below.
Key Components in Heatsink Test Platform
- Intel
Pentium D 950* Presler core, C1 stepping. TDP of 95W; under our test load, it measures
78W including losses in the VRMs. - Asus P5Q-EM motherboard. A microATX board with integrated graphics
and short solid-state capacitors around the CPU socket, and a diminutive northbridge heatsink for maximum compatibility. - Intel X25-M
80GB 2.5″ solid-state drive. - 1GB of Corsair XMS2 DDR2 memory. 2 x 512MB PC2-8500.
- FSP Zen 300W
fanless power supply. - Arctic Silver
Lumière: Special fast-curing thermal interface material, designed
specifically for test labs. - Nexus 92 fan (part of our standard testing methodology; used when
possible with heatsinks that fit 92x25mm fans)
Nexus 92 fan measurements* | ||
Voltage | Noise | Speed |
12V | 17 dBA@1m | 1240 RPM |
9V | 13 dBA@1m | 980 RPM |
7V | 11 dBA@1m | 770 RPM |
* measured mounted on Katana 3 |
Measurement and Analysis Tools
- Seasonic
Power Angel for measuring AC power at the wall to ensure that the
heat output remains consistent. - Custom-built, four-channel variable DC power supply, used to regulate
the fan speed during the test. - PC-based spectrum analyzer:
SpectraPlus with ACO Pacific mic and M-Audio digital
audio interfaces. - Anechoic chamber
with ambient level of 11 dBA or lower - Various other tools for testing fans, as documented in our
standard fan testing methodology. - SpeedFan, used to monitor the on-chip thermal sensor. This sensor is not
calibrated, so results are not universally applicable. - CPUBurn
P6, used to stress the CPU heavily, generating more heat than most
real applications. Two instances are used to ensure that both cores are stressed. - Throttlewatch
2.01, used to monitor the throttling feature of the CPU to determine
when overheating occurs.
Load testing was accomplished using CPUBurn to stress the processor, and the
graph function in SpeedFan was used to make sure that the load temperature was
stable for at least ten minutes. The stock fan was tested at various voltages
to represent a good cross-section of its airflow and noise performance.
*An Aside: Pentium D at 130W TDP? The low 78W (minus VRM losses) drawn by our Pentium D 950 chip has been a curious oddity since we began using it on our heatsink test platform a few years ago. As far as we knew, it was supposed to have a TDP of 130W; we assumed ours was an exceptionally cool sample. Recently, curiosity followed its inevitable course: Web research after identifying the exact details of our Pentium D — model 4, stepping 4, revision C1, 12+16kB/2048kB L1/L2 cache, ‘Presler’ FSB200x4, XD, VT, EIST. It turns out there were two main versions of the Pentium D. Ours was the last iteration, not surprisingly, the most energy efficient. The C1 stepping identified it on Intel’s Processor Spec Finder database as a Presler core with 95W TDP, not the 130W we’d always thought. At less than 78W at full tilt, it still seems like an exceptionally cool running processor. |
TEST RESULTS
The Fan
Some basic measurements were made of the Scythe SY9225SL12M-P fan included with Katana 3.
Fan Specifications | |||
Brand | Scythe | Power | 0.17A (2W) |
Model | SY9225SL12M-P | Airflow | 6.7 – 55.55 CFM |
Bearing | sleeve | RPM | 300~2500 |
Hub Size | 34mm (1.32″) | Noise | 7.2 – 31.07 dBA |
Frame Size | 92 x 92 x 25 mm | Header | 4-pin |
Weight | 115 grams | Start Voltage | ~7.5V |
The data in the blue cells is provided by the manufacturer; we measured the data cited in the green cells |
FAN MEASUREMENTS | |||
Voltage | SPL – dBA@1m | RPM | |
free air | mounted | ||
12V | 32 | 34 | 2600 |
10V | 27 | 28 | 2250 |
9V | 22 | 23 | 1900 |
8V | 16 | 16 | 1520 |
7.5V | 11 | 11 | 830 |
Regular readers will notice that the measurement points are not at the usual 12, 9, 7 and 5 volts. This PWM fan, like many others, does not have as linear a relationship with voltage as normal DC fans. The range of speeds controllable via voltage is only about 1400~2600 rpm, and by 7.5V, the fan does not start reliably. Via PWM control in our motherboard BIOS, speeds well below 1400 rpm were seen, and we have no reason to doubt the 300 RPM minimum speed. The fine control of a knob is not available on our motherboard’s BIOS fan controller, so there was no easy way of tagging the SPL at lower speeds. Suffice it to say that 16 dBA@1m is very quiet, and given the smooth quality of this fan, the noise at lower speeds will be essentially inaudble under most conditions.
When mounted on the heatsink, the fan exhibited a bit more noise than in free air, and the difference became smaller as fan speed was reduced. This is normal; it’s the increased turbulence caused by the fan blades’ close proximity to the fins which causes this effect.
Sound Impressions
@ 12V: The noise was mostly broadband turbulence, with some high pitched tonal elements. It was not terribly loud, but too loud to be considered quiet.
@ 10V: The overall level dropped substantially, and the sound stayed smooth, but some tonal traces remained. Still not exactly quiet.
@ 9V: The noise dropped to a point where many users would find it quiet enough to ignore. Some tonality still remained, although inside a decent case with other noise source, it would probably be obscured and blocked.
@ 8V: Very subdued in level, smooth and unobtrusive, probably inaudible in a good case under most conditions for most people. A touch of tonality was still audible from up very close.
The tonal aspects of the noise seem to be caused by the interaction of the fan’s vibrations and airflow, and resonances in the fins. It’s possible that adding damping material between the fan and the heatsink could reduce the tonality, because the fan by itself in free air exhibits little of the tonal traces described above.
Cooling Results
Scythe Katana 3: Stock Cooling Performance | ||||
Fan Voltage | SPL dBA@1m | CPU Temp °C | °C Rise | °C/W |
12V | 34 | 39 | 15 | 0.19 |
10V | 28 | 40 | 16 | 0.20 |
9V | 23 | 41 | 17 | 0.22 |
8V | 16 | 44 | 20 | 0.26 |
Scythe Katana 3 w/ reference 92mm fan | ||||
12V | 17 | 42 | 18 | 0.23 |
9V | 13 | 47 | 23 | 0.30 |
7V | 11 | 50 | 26 | 0.33 |
Load Temp: Prime95x2 for ~10 mins. °C Rise: Temperature rise above ambient (24°C) at load. °C/W: based on the amount of heat dissipated by the CPU (78W); lower is better. |
The Scythe Katana 3 performed quite well, with a temperature rise of 15°C at full fan speed, and dropping by small increments as the fans was slowed. In stock form, at the slowest speed we could test at, with the noise level at just 16 dBA@1m, the rise above ambient was just 20°C. This is excellent performance for a heatsink of this size.
With the quieter, slower Nexus reference fan, the cooling performance was not quite as good, but the noise could be dialed down even further. The cooling performance remained acceptable down to ~13 dBA@1m.
COMPARISONS
When judged on noise vs. cooling, Katana 3 fares well against most other stock heatsink/fan combos of similar size SPCR has tested. Here’s a quick comparison against a few heatsinks with stock fans at around 23~24 dBA@1m.
Comparison: With Stock Fans at 12V | ||
Heatsink | °C rise | SPL |
ZEROtherm CORE92 | 12 | 20 dBA |
Zalman CNPS9300 AT | 15 | 24 dBA |
Scythe Katana 3 | 17 | 23 dBA |
Cooler Master Hyper N520 | 19 | 23 dBA |
Xigmatek HDT-SD964 | 20 | 21 dBA |
Thermaltake MaxOrb | 21 | 24 dBA |
How close is the Katana 3 to the very best silent coolers? Not very close, like most of its direct competitors. But it’s neither as big nor as costly.
°C rise Comparison: The Best of the Best | ||||
Heatsink | Nexus 120 fan voltage / SPL @1m | |||
12V | 9V | 7V | 5V | |
16 dBA | 13 dBA | 12 dBA | 11 dBA | |
Prolimatech Megahalems | 10 | 14 | 17 | 20 |
Thermalright U120E | 12 | 14 | 17 | 24 |
Thermalright HR-01+ | 13 | 15 | 16 | 20 |
Xigmatek HDT-S1283 | 13 | 15 | 18 | 22 |
Noctua NH-U12P | 14 | 16 | 17 | 21 |
Zalman CNPS10X Extreme | 14 | 17 | 21 | 26 |
Zerotherm Zen FZ120 | 15 | 16 | 19 | 24 |
Scythe Ninja 2 | 17 | 18 | 20 | 23 |
Thermolab Baram | 18 | 20 | 22 | 25 |
All results generated with our reference Nexus 120mm fan. |
MP3 SOUND RECORDINGS
These recordings were made with a high
resolution, lab quality, digital recording system inside SPCR’s
own 11 dBA ambient anechoic chamber, then converted to LAME 128kbps
encoded MP3s. We’ve listened long and hard to ensure there is no audible degradation
from the original WAV files to these MP3s. They represent a quick snapshot of
what we heard during the review.
These recordings are intended to give you an idea of how the product sounds
in actual use — one meter is a reasonable typical distance between a computer
or computer component and your ear. The recording contains stretches of ambient
noise that you can use to judge the relative loudness of the subject. Be aware
that very quiet subjects may not be audible — if we couldn’t hear it from
one meter, chances are we couldn’t record it either!
The recordings start with 7~10 second segments of room ambiance, then the fan
at various levels. For the most realistic results, set the volume so that
the starting ambient level is just barely audible, then don’t change the volume
setting again.
- Scythe Katana 3 with stock fan at one meter
— 8V (16 dBA@1m)
— 9V (23 dBA@1m)
— 10V (28 dBA@1m)
— 12V (34 dBA@1m)
- Cooler Master Hyper N520 with stock fans at one meter
— 5V (14 dBA@1m)
— 7V (19 dBA@1m)
— 9V (23 dBA@1m)
— 12V (30 dBA@1m) - ZEROtherm
CORE92 at one meter
— 5V (11 dBA@1m)
— 6V (15 dBA@1m)
— 7V (16 dBA@1m)
— 9V (20 dBA@1m)
— 12V (34 dBA@1m) - Any big heatsink with Nexus 120mm fan at one meter
— 5V (11 dBA@1m)
— 7V (12 dBA@1m)
— 9V (13 dBA@1m)
— 12V (16 dBA@1m)
FINAL THOUGHTS
The Scythe Katana 3 delivers cooling performance that’s good enough for most mid-power CPUs (~65W TDP). Its stock fan is decently quiet, and with a good on-board PWM fan controller, could provide essentially inaudible cooling for a midrange PC. The dreaded plastic push-pins for socket 775/1366 boards are easy to use here, as access to the pins is good. Its low weight makes the push-pins safe; there’s no risk of them popping out or breaking under undue strain. Even though the AMD clip doesn’t allow for heatsink or fan rotation, on most AMD boards, the fan will end up blowing in the right direction, toward the back case exhaust fan.
Whether the down-angled fan helps with VRM or northbridge cooling was not possible for us to assess. With mid and lower power CPUs, the need for such cooling is probably not critical, but our basic perspective is that it certainly can’t hurt.
The support for socket 1366 is an amusing touch, and made possible because it’s just an integral part of the Katana 3 socket 775 mounting bracket. Could the Katana 3 cool a 130W TDP i7 processor effectively? Well, with the fan at or close to full speed, probably, but it would hardly be a quiet solution.
Scythe Katana 3 does not pretend to be a giant killer (or cooler), but does a fine job of quietly cooling a mid-power processor. It is easy to install and use on any current CPU socket. Best of all, its modesty also extends to price: US$26~30.
Scythe Katana 3 | |
PROS * Good cooling for midpower CPU * Quiet PWM fan | CONS * No fan controller supplied? |
Our thanks to Vanvouver retailer Anitec for the Scythe Katana 3 sample.
* * *
Articles of Related Interest
Xigmatek HDT-SD964
Thermaltake BigTyp 14Pro: A Bigger Typhoon
Zalman CNPS9900 LED: The End of the
Nines
ThermalTake SpinQ: Unique Blower-fan
Heatsink
Thermolab Baram Blows In
Scythe Ninja 2: Tweaking a Classic
Zalman 9300AT: Not me too, but
me again
* * *