Ninja Copper: Scythe's 5th Year Celebration

Post by **MikeC** » Tue Mar 04, 2008 6:43 pm

NeilBlanchard wrote:I marked up the photo you posted and I pointed to the round heatpipes that do not appear to be flattened. They are the upper set when it is installed on the CPU, and it would appear that very little heat could be transfered to them. Especially, since there is no solder at all on the Rev Cu, which is more than a little surprising, IMO.

They are flattened, Neil, It's hard to see it from that photo, but they are. Every Ninja's heapipes are flattened similarly. Where they differ is in both the bottom copper base and the upper 2nd Al HS. And yes, I agree with cmthomson, that AL heatsink doesn't really do much; it can't.

Anyway, the NinjaCU is still a good performer, even tho it doesn't match the old sample. W/ Nexus 120/7V only the TRU120E and the old Ninja clearly beat it. My main issue with it is that the clips don't seem reliable given the weight, esp. given the other stories here of tension sag over time with the light standard Ninjas,

In comparison, the Rev B sample isn't even in the same ballpark. That's the one which really has problems. 24C rise at 12V and 31C at 5V?!

FartingBob · Post by **FartingBob** » Wed Mar 05, 2008 4:43 am

From a guy who just got a rev B a few weeks ago without knowing about these small factors which make it less efficient than expected: damn.
Well its good enough for my system, im sure in a year or 2 i'll get a beefy quadcore and then upgrade to whatever is the top HS at the time.

Tzupy · Post by **Tzupy** » Wed Mar 05, 2008 6:34 am

I would bet on the IFX-14 in a contest against the Ninja CU, it certainly has good hetpipe contact, and weighs less than the Ninja CU.
By contest I mean in a low airflow setup. The review at Madshrimps showed it to be *fanless* on par with a fanned Ninja.
Is there any chance to read a SPCR review of it soon?

Scrooge · Post by **Scrooge** » Thu Mar 06, 2008 6:23 am

Let me say that the single, solitary reason I buy Thermalright products is their bolt-through kits. I am very happy that their performance is good as well, but ever since the S370 days, when I ran the Alpha Pal8045 (not the Swiftech, but also bolted through the motherboard) I have only used heatsinks that bolt through. I know it's possible to modify a Ninja to do the same, but frankly I would prefer to encourage manufacturers to do it right themselves. Count me in that camp, and MikeC and co, please feel free to pass that along to Scythe. I'm about to build my mother a new computer, which will be the third heatsink sale for Thermalright to me in the past 12 months due to their mounting mechanism.

cmthomson · Post by **cmthomson** » Thu Mar 06, 2008 1:00 pm

The photos of the new Scythe Orochi seem to indicate that they went back to the Ninja Rev A backplate-and-rail mounting system. The clips are still screw-on (to accommodate both old and new AMD clips; the 775 adaptor is old-style).

I guess the question is whether the clips still provide the same amount of pressure they did with the original Ninja.

KansaKilla · Post by **KansaKilla** » Thu Mar 06, 2008 1:23 pm

Strid wrote:I think you're right on the whole clips vs. bolt through thing. I have a an FX-60 s939 with the old style Ninja. I find that it over time cantilevers itself off the cpu, just slightly. If I leave it too long, the gap between HS base and CPU will increase and the thermal paste will run down ever so slowly, until it leaves the top part completely bare. I was shocked the fist time I took off the and found the base looking like this:

Top part looking burnt with no thermal paste and bottom part looking okay. I was all "Ahh, so that's why the cores were running at more than a 10Â° C difference!" when I saw that. (this was after one year - original thermal paste was Zalman Super Thermal Grease, ZM-STG1. I used the apply a thin layer to IHS surface-method.)

I just experienced the exact same thing on my ninja rev. b. When I took off the heatsink, the top looked the same as your picture. Temps had gradually increased since I put it on a month ago or so. I used the pushpin method and the included thermal grease.

Bah, humbug. Back to the drawing table now for me.

(Edited to display quote properly)

FartingBob · Post by **FartingBob** » Thu Mar 06, 2008 1:25 pm

cmthomson wrote: I guess the question is whether the clips still provide the same amount of pressure they did with the original Ninja.

You'd hope so, the Orochi (which i think should be renamed orca, after the whales) is twice as heavy as other tower heatsinks. Id want the most secure mounting system possible.

ghettojiggalo · Post by **ghettojiggalo** » Thu Mar 06, 2008 8:44 pm

just curious, was the original ninja reviewed w/ the top end caps OFF like in the picture???

i'm building my new rig this weekend, i'll see if theres a temp diff w/ the end caps off/on.

Post by **MikeC** » Thu Mar 06, 2008 9:43 pm

ghettojiggalo wrote:just curious, was the original ninja reviewed w/ the top end caps OFF like in the picture???

No, but that's the state it was in in the past year, and it's continued to perform about the same as before -- maybe a degree or 2 worse than before with some airflow points. Anyway, there's no benefit to pulling the end caps off, trust me... doesn't even reduce the height -- which is why it was tried in the first place.

BillTodd · Post by **BillTodd** » Mon Mar 10, 2008 7:43 am

Mike,

Have you concidered measuring the thermal resistance of these and other suitable heat-sinks (in free and forced air)?

[I've been searching the web to find free-air ratings for the Scythe and other heat-sinks and have found nothing so far]

kaange · Post by **kaange** » Mon Mar 10, 2008 4:41 pm

kaange wrote:Hmmm.
I wonder if a ghetto mod could be done by running a bead of silicone across each side of the base, except in each direction to seal up the base into an open box, then filling the base with water before sealing it all in with more silicone on last side of the base. Just a thought.....

Would be messy if it leaked, of course.

Just wondering if anyone with a rev B Ninja would be willing to try this. I'm betting that performance would improve a LOT.

Bluefront · Post by **Bluefront** » Tue Mar 11, 2008 2:09 am

If you want to try this.....why in Hell would you use water? There are numerous products like "liquid steel". "liquid aluminum", that could be poured/squeezed in the small gaps from both sides, with the excess just being sanded/cut off when dry.

The object is to eliminate small air gaps for better heat transfer. Melted lead could be poured in........but please no water.

kaange · Post by **kaange** » Tue Mar 11, 2008 3:28 pm

Those "liquid steel" type pastes have 2 issues.

1. The liquid in the emulsion may water based (some are solvent based) but also contains binding agents to set the mixture and these are not particularly thermally conductive (think of metallic paint).

2. The viscosity of the "liquid steel" is rather thick so it is unlikely to flow into all the small voids as well as water would (we are talking about some afirly constricted spaces here).

Anyway, what I suggested was a simple, easily reversible test - once "liquid steel" was set, you'd be stuck with the result.

BTW pure water is non conducting.

Post by **NeilBlanchard** » Tue Mar 11, 2008 7:06 pm

Hello,

I'll bet a a mondo 150watt soldering iron, or a plumber's torch would make short work of it. Starting with a little acid flux, and use the heat to draw in the molten solder -- it would do the trick, I think.

kaange · Post by **kaange** » Wed Mar 12, 2008 2:42 pm

You still risk desoldering the lower fins from the heatpipes plus what stops the solder from flowing right out of the other side of the base? Especially when trying to fill the voids around the upper set of heatpipes?

mcv · Post by **mcv** » Mon Mar 24, 2008 8:41 am

Do I understand correctly that the Rev B is a lot worse than the original Ninja? I just bought a brand new PC with one, and this is already the second component that turns out to be a bad buy (the memory was the first one). The store that assembled it claimed that I couldn't use passive cooling with it, and I thought they were crazy, because it was supposed to be the best passive cooler. Now it turns out they were right?

Since they also refused/neglected to use the Thermalright bolt-thru kit, I guess I'd better not risk passive cooling at all, now. I admit I'm a bit bummed about this news.

Post by **NeilBlanchard** » Mon Mar 24, 2008 9:06 am

Hello,

Yes, the Rev. B (the 1100) is not up to the same standard as the original. It suffers from the bog-standard push-pin mounting, and from (apparently) less well designed base and/or inconsistent soldering between the base and the heatpipes.

You can use the Thermalright Socket 775 bolt-through kit ($4-5), and it can certainly work very well -- today's CPU's are more efficient than they were a generation or two ago, so they are not as hard to cool. The overall design of the Ninja (the wide spaced fins, and "omni directional" air flow in particular) is unique, and they contribute to the Ninja's ability to work well with very little air flow. I.E. it works well with a slow spinning (or no) fan.

HammerSandwich · Post by **HammerSandwich** » Wed Mar 26, 2008 2:09 pm

I've experimented with my new setup for a while now, and it has performed well.

Using a Ninja Rev. B with stock pushpins to cool an E3110 (same as 8400) at 3.6GHz & stock voltage, the temperatures are excellent. Full double-Prime load maxes the CPU at 50C with a 500rpm YL. I've run only open-case so far, so the motherboard sits flat on a table and the mounting system hasn't been tested fully. However, adding 3-4 pounds of hardcover books to the Ninja's heatpipes (to increase contact pressure) had zero effect on the temperatures.

More later.

JohnMK · Post by **JohnMK** » Wed May 28, 2008 5:52 pm

I'd love to see a response from Scythe on this issue.

Goldmember · Post by **Goldmember** » Fri Jun 13, 2008 8:38 am

I just noticed the Scythe Ninja 2 SCNJ-2000. Looks like the copper version sans the copper.

And is that a slipstream fan? It still has those stupid push pins though.

http://www.scythe.co.jp/cooler/ninja2.html

winguy · Post by **winguy** » Sun Jun 15, 2008 12:50 pm

More pics of the Scythe Ninja 2 here.

And ComputerBase review.

tehcrazybob · Post by **tehcrazybob** » Tue Jun 24, 2008 8:05 pm

Based on the disappointing performance of the RevB, the Ninja 2 looks like a terrible idea: You'll notice that the slots in the fins have been removed, making it impossible to correctly install a Thermalright bolt-through kit.

I'd really like to see an SPCR review comparing the RevB with bolt-through to the current champions. I know it's not a typical comparison, because you're combining parts from different manufacturers, but it's a combination that many people on these forums use. I'd also be willing to bet the Ninja is the single most common heatsink here, and if a bolt-through kit were shown to improve it up to or beyond the performance of an HR-01 Plus, a lot of users would want to know. It could easily save people money upgrading to a whole new heatsink.

FartingBob · Post by **FartingBob** » Wed Jun 25, 2008 1:47 am

tehcrazybob wrote: I'd really like to see an SPCR review comparing the RevB with bolt-through to the current champions. I know it's not a typical comparison, because you're combining parts from different manufacturers, but it's a combination that many people on these forums use. I'd also be willing to bet the Ninja is the single most common heatsink here, and if a bolt-through kit were shown to improve it up to or beyond the performance of an HR-01 Plus, a lot of users would want to know. It could easily save people money upgrading to a whole new heatsink.

Seconded. Its a very common setup here, would be good if we can get a direct comparison to the other coolers.

Cistron · Post by **Cistron** » Wed Jun 25, 2008 2:22 am

Page6, yellow box.

tehcrazybob · Post by **tehcrazybob** » Wed Jun 25, 2008 4:19 am

Cistron: Thanks for pointing that out. I'm embarrassed I didn't notice it.