Scythe Andy Samurai Master CPU heatsink/fan

Want to talk about one of the articles in SPCR? Here's the forum for you.
Post Reply
MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Scythe Andy Samurai Master CPU heatsink/fan

Post by MikeC » Thu Mar 01, 2007 9:55 am


Ryan Norton
Patron of SPCR
Posts: 169
Joined: Wed Sep 21, 2005 5:13 pm
Location: South FL

Post by Ryan Norton » Thu Mar 01, 2007 10:05 am

Always nice to see Scythe products get SPCR treatment as opposed to the "how high can I OC with it" treatment CPU heatsinks get from other sites. That said, every HS review I read makes me believe there's still nothing that much better than a Ninja, assuming your case has room, you're not wildly overclocking, and it's not breaking your bank!

crispyfish
Posts: 222
Joined: Tue Nov 21, 2006 10:53 am

Post by crispyfish » Thu Mar 01, 2007 11:35 am

Great review! It would be nice if you added a table comparing the Andy to the SI-128 and CNPS7700 though, since those are the main competitors in quiet vertical-flow designs on the market right now.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Thu Mar 01, 2007 11:44 am

crispyfish wrote:Great review! It would be nice if you added a table comparing the Andy to the SI-128 and CNPS7700 though, since those are the main competitors in quiet vertical-flow designs on the market right now.
Thanks. :)

The SI-128 data can be added, but our 7700 data is based on an early socket 478 platform which has been retired due to motherboard damage. It's no contest anyway -- even when it was first introduced, the 7700 was not really much better than the much older 7000, and louder. As you can guess just from their relative sizes, the Andy outperforms the SI-128 considerably, by about 5C across the fan speed range.

autoboy
Posts: 1008
Joined: Fri Dec 10, 2004 8:10 pm
Location: San Jose, California

Post by autoboy » Thu Mar 01, 2007 2:49 pm

So without the fan this heatsink could fit in a nsk2400. I wonder if it would be a good passive heatsink for that case?

mrzed
Patron of SPCR
Posts: 281
Joined: Wed Jun 25, 2003 4:01 pm
Location: Victoria, Canada

Post by mrzed » Thu Mar 01, 2007 3:50 pm

I've been wondering the same thing. Ideally, I'm looking to replace my retail AMD HSF, but it runs fine at about 7v, and is surpisingly quiet at that speed. I need to fix other noise sources before getting too concerned with the CPU cooling.

I'm surprised there has not been a top notch HTPC-worthy heatsink produced lately. I think a basic heatpipe equipped top-down design with decent fin spacing couldn't be that hard. Is there not a large enough market? The preponderance of designs with dense fin spacing leads me to think that the performance/OC market must still be heaps larger than the silence/HTPC market. Or perhaps the review sites and forums focus on that, so the designs tend to follow the reviews.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Thu Mar 01, 2007 3:50 pm

passive.... only if the fins are running left/right, ie, parallel with the front of the case. Otherwise, I think the forced air of the fans would just go around the HS fins instead of through. Even with correct orientation, the spacing is pretty tight, but it's probably doable for most <70W max CPUs.

dragmor
Posts: 301
Joined: Sun Jul 10, 2005 7:54 pm
Location: Oz

Post by dragmor » Thu Mar 01, 2007 4:34 pm

That looks like the same fan that came with my RevB Ninja.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Thu Mar 01, 2007 5:10 pm

mrzed wrote:Is there not a large enough market?
Well, Scythe has just now attracted a "flattering competitor" for the Ninja, and decided not to produce the "dwarf Ninja" that it showed recently (a shorter but still 110mm-square and same fin spacing HS). Assuming an efficient market, that indicates a very small market.

Think about it: how many Dell, HP etc computers are produced using Ninjas? None, of course. Ninjas have a small piece of the small replacement market. That's "small" squared! :oops:

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Thu Mar 01, 2007 5:34 pm

Felger Carbon wrote:Well, Scythe has just now attracted a "flattering competitor" for the Ninja, and decided not to produce the "dwarf Ninja" that it showed recently (a shorter but still 110mm-square and same fin spacing HS). Assuming an efficient market, that indicates a very small market.
I asked Scythe to make that small version of the Ninja for a 92mm fan specifically for use in the NSK2400 (and similar "height" HTPC cases) during the tail end of the development of that case. They were pretty lukewarm about it, which I don't really understand. In they end they said the cost was too similar to the Ninja and they didn't think the market would pay the price. Sheesh, why not let the market decide?

There remains no really great HS option for such cases. A short Ninja would be perfect. They could have swept the market. But no, they keep going after the same market over and over again. Surely, increasing the size of the target market has got to be an obvious strategy for business growth/survival.

Right now, I see Arctic Cooling have something they're promoting as a perfect HTPC cooler... but the look of it is not that promising to me. I'll have to see about a sample.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Thu Mar 01, 2007 8:55 pm

By request from Ralf Hutter, this addendum was made:
TOP FLOW ADVANTAGE?

This refers to the claim made in the Andy's marketing blurb, about how the straight down fan airflow helps to cool the motherboard components (like voltage regulators and chips) better than parallel-to-the-motherboard type tower designs. Unless a test is designed specifically for it, this claim is difficult to deny or verify.

What we do know is that if the voltage regulators get too hot, their efficiency drops, and the system ends up drawing more power. We usually monitor the AC power draw of the test system as a matter of routine. We also pay attention to the DC power drawn by the AUX12V (2x12V) socket that powers the motherboard voltage regulators and the CPU. If the power draw under our CPU stress testing varies, usually that's a sign that we should pay attention to whatever is causing this.

The DC power draw with the CPU under high stress measured 78W when we examined it while setting up the test platform. That included tests runs with three tower style heatsinks, the Ninja, Zalman 9500 and Thermalright Ultra 120. The measurement remained at 78W with the Andy, too, so at least for our test setup and conditions, there was no apparent top flow advantage.

Kremmit
Posts: 116
Joined: Fri Dec 15, 2006 11:27 pm

Post by Kremmit » Thu Mar 01, 2007 10:46 pm

Minior quibble: From the caption under the first picture on page 2:
With the fan it stands just shy of 5". It may not be a tower, but it's still pretty tall, and some clearance is still needed above the fan to allow air intake, probably at least one inch. That puts the real operational height on par with the tower giants like Ninja, Thermalright SI-128, Infinity, etc.
Comparisons throughout the review to the SI-128 make sense based on the top-down airflow and basic design shape. But when comparing heights in a list of "tower giants", maybe you meant to list the Ultra-120?

Tzupy
*Lifetime Patron*
Posts: 1561
Joined: Wed Jan 12, 2005 10:47 am
Location: Bucharest, Romania

Post by Tzupy » Fri Mar 02, 2007 2:49 am

I find it odd that the fan sound and performs similarly to the Nexus, since there is a 15-20% difference in rpm, which should be audible and result in a small temperature difference too.
Maybe the Nexus provides better backpressure and is a bit louder at the same rpm? That sounds odd too...
PS. The review on the SI-128 still mentions 'Not compatible with closed-flange fans', which is of course not true.
PS2. The Andy Samurai Master is ugly - compared to my SI-128. :lol:

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Fri Mar 02, 2007 3:57 am

Tzupy wrote:The Andy Samurai Master is ugly - compared to my SI-128. :lol:
5C is 5C. :P

walle
Posts: 605
Joined: Tue Dec 26, 2006 10:52 am

Post by walle » Fri Mar 02, 2007 5:23 pm

[quote]“There was no apparent top flow advantageâ€

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Fri Mar 02, 2007 5:43 pm

Tzupy wrote:I find it odd that the fan sound and performs similarly to the Nexus, since there is a 15-20% difference in rpm, which should be audible and result in a small temperature difference too.
Maybe the Nexus provides better backpressure and is a bit louder at the same rpm? That sounds odd too...
It's nowhere that big a difference in speed -- 1200rpm vs something like 1100. The temperatures were not identical -- but it was within a degree, which is not significant, imo. If the quality of the sound is similar, even a 200rpm difference doesn't really sound that different.

butters
Posts: 64
Joined: Sun Feb 18, 2007 9:48 pm

Post by butters » Sat Mar 03, 2007 7:33 am

[quote="walle"][quote]“There was no apparent top flow advantageâ€

Tzupy
*Lifetime Patron*
Posts: 1561
Joined: Wed Jan 12, 2005 10:47 am
Location: Bucharest, Romania

Post by Tzupy » Sat Mar 03, 2007 8:29 am

@MikeC: I can tell the difference between 1,000 rpm and 800 rpm on a Nexus, but maybe not from 1 m.

@butters: it's not the airflow pattern, but the extra fin area and extra fin-heatpipe contacts that make the difference.
There is a difference in cooling the motherboard chipset: with a Nexus at 450 rpm versus 0 rpm I measure 2-3 degrees difference in idle.
At load the Nexus starts anyway and goes to 600-700 rpm so I can't measure the temperature difference, but it should be more.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Sat Mar 03, 2007 12:21 pm

Tzupy wrote:@MikeC: I can tell the difference between 1,000 rpm and 800 rpm on a Nexus, but maybe not from 1 m.
OK. The gist of what I was saying in the review is that if you use the stock fan with a controller to reduce the speed/noise like most SPCR readers would, the end result would be that you'd get the same cooling at the same noise as with the Nexus. Whether you have the fan set to 9V or 8.5V doesn't matter to anyone.

As for the chipset/VRM cooling, the main indicator for us is NOT the temperature of the chips (or whatever the MB sensors are reading) but the power draw at the 2x12V connector and the AC outlet. If there's no change with those readings, then there's not likely to be any temperature increase either. There was no change from any of the tower HS we tested on this platform before.

This doesn't mean there will never be any advantage, just that there isn't one for our setup. OC the CPU and put it in a case with dual 8800s... and you'd probably see a difference.

mrzed
Patron of SPCR
Posts: 281
Joined: Wed Jun 25, 2003 4:01 pm
Location: Victoria, Canada

Post by mrzed » Sat Mar 03, 2007 1:00 pm

MikeC wrote: This doesn't mean there will never be any advantage, just that there isn't one for our setup. OC the CPU and put it in a case with dual 8800s... and you'd probably see a difference.
This is a really important point. I was thinking about this and wondering if it might be useful to add a single sentence to HSF reviews to that effect. SPCR reviews are IMO, among the very best out there, but they are rarely comparable to other reviews, which tend to use bleeding edge systems, and focus on max overclock etc.

I've noticed that many of the review threads here often have posts questioning the results by comparing them to other reviews where power draw may well be double what SPCR is using.

hravn
Posts: 198
Joined: Sun Sep 25, 2005 10:48 pm
Location: Stockholm, Sweden

Post by hravn » Mon Mar 05, 2007 5:59 am

MikeC wrote:I asked Scythe to make that small version of the Ninja for a 92mm fan specifically for use in the NSK2400 (and similar "height" HTPC cases) during the tail end of the development of that case. They were pretty lukewarm about it, which I don't really understand. In they end they said the cost was too similar to the Ninja and they didn't think the market would pay the price. Sheesh, why not let the market decide?
Yeah, that's really stupid. Scythe seems to have the resources to produce lots of different heatsinks that basically fills the same niche, but can't make a smaller Ninja?

roadie
Posts: 166
Joined: Fri May 12, 2006 2:07 am
Location: Liverpool, UK

Post by roadie » Tue Aug 28, 2007 3:43 am

Would love to see this versus the SI 128 SE.

From my experience of the Scythe fan that comes with most of their new heatsinks, it is silent at ~870 rpm but quite noticeable at full speed.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Tue Aug 28, 2007 10:29 am

roadie wrote:Would love to see this versus the SI 128 SE.
Hartware.de tested both and the 128SE was a clear winner. Regrettably, Hartware has changed its advertising standards and the site is now essentially useless, unless you're looking for phone sex. 8)

colm
Posts: 409
Joined: Tue Jan 31, 2006 8:22 am
Location: maine

Post by colm » Sun Jun 01, 2008 9:04 pm

bump ing an "old" thread

altho just a year.

I am wondering who did the inevitible reverse cooling with this heatsink in the old fatty cases like antecs lanboy, 2600amb and 3something amb etc would deal with ductwork and height. (8.75 inch wide atx tower cases)I am currently running an xp90, and in 80 degree rooms it still brings sytem to 114-116 and cpu to 145 with a 2600 rpm fan on custom ductwork and other perfections leaving me to conclude.... my 2.8e needs yet another try with something bigger, and the 775 is of course my next pc and need somehting to evolve wtih.

also a note about the xp 90: the wings that hold the clips bent, I made new ones to maintain pressure.So far so good. (there is a thread here about that somewhere) It will be trash bound, possibly, after finding this andy samuri whatever it is.
So to sum this up:

anybody reverse cool this andy sink? it is the most realistic I have seen aside form the xp 90 for reverse cooling. I have plenty of room for siize of it, my doubt is the design.

bgavin
Posts: 160
Joined: Sat Apr 12, 2008 5:05 pm
Location: Orangevale, CA

Post by bgavin » Tue Jun 03, 2008 3:36 am

MikeC wrote:As for the chipset/VRM cooling, the main indicator for us is NOT the temperature of the chips (or whatever the MB sensors are reading) but the power draw at the 2x12V connector and the AC outlet.
IMO, this is short sighted. Heat is the enemy of all electronic components. Even if the power draw remains the same, the component life is certainly shorter. This results in a quiet, but early end-of-life product.

As a computer servicer, I do a lot of capacitor replacement (recapping) on VRM modules where the caps have failed. The primary cause is use of cheap Chinese caps instead of premium Japanese types. Heat is the accelerator that makes cheap caps pop much sooner.

Yes, solid polymer caps are much more durable, and the mobo manufacturers have finally seen the light about this. However, their life span is still directly affected by temperature.

Isn't there a simple probe device that allows measurement of surface temps? This would reveal the temp difference between down draft vs. side draft on the VRM.

Post Reply