Semi-poll of Core 2 Duo Ninja users

hexen · Post by **hexen** » Wed Apr 25, 2007 6:13 pm

Delta_42 wrote:
hexen wrote:
but if you dont use the push pins and replace them with screws/nuts then the problem is rectified no?
http://www.xtremesystems.org/forums/sho ... p?t=117054
But why should you have to pay extra and/or bugger about with mods to get an adequate mounting mechanism? Scythe should provide one out of the box (and used to do so!).

i agree.

they should have, could have, would have

but didnt!

so if u want a ninja, u got to deal with the issue.

sjoukew · Post by **sjoukew** » Thu Apr 26, 2007 2:03 pm

well here are my stats.
I attached the ninja with the mobo out of my case, really making sure the push pins were correctly attached. The feel of the cheap metal parts is something I don't like.

asus p5w dh deluxe
intel c2d e6600
ninja rev b. with original fan running (full speed)
zalman "super thermal grease" zm-stg1 ( I wonder what their next grease will be, hyper, ultra hyper grease ??

)

ambient: 28 degrees
idle: 42 degrees
full load intel TAT : 65 degrees

I expected better temperatures under full load from such a big cooling solution with such heatpipes. It is only slightly better than the zalman cnp7000 which I had first, and that zalman thingie is a real lot smaller than this huge ninja.

Jokoto · Post by **Jokoto** » Wed May 02, 2007 12:59 pm

Asus P5B Deluxe WiFi, stock E6600 with Ninja Rev. B + Nexus @ 600 rpm:

ambient: 23 Â°C
idle: around 30 Â°C
load (Folding@Home 24/7): 47-50 Â°C max

I remember having to mount it several times to get it right when I built this machine last year. If at first you don't succeed...

spacey · Post by **spacey** » Thu May 10, 2007 12:53 pm

Has anyone used the Scythe Universal retention mechanism with the Rev. B? I bought one the other day...

Going to try it out on my E6600 + Ninja Rev. B tomorrow.

spacey · Post by **spacey** » Mon May 14, 2007 9:09 pm

So I overclocked my E6600 to 3 Ghz but left the vcore at stock... whatever that is.

I'm using the Ninja Rev. B with a the Universal Retention plate with the copper spacer... I used a thin layer of AS5 of course. This isn't my first build... however it is my first Core 2 Duo.

Idle temps were around 30 degrees, which was cooler than the ambient temperature of my case (Lian Li V1100 plus with all fans throttled to slowest possible) however as soon as I started Orthos, the temperatures quickly climbed.

I've been letting it run for about an hour now and my temperature according to the Intel Thermal Analysis Tool is 67 and 66 degrees while Speedfan says its 68 and 68.

Did I do something wrong? I am pretty sure I mounted the heatsink on quite well. Back before I overclocked, my idle temperature was 25 degrees with the E6600, I thought that indicated I did things right but these really high load temperatures make me wonder.

If things don't break, I wonder how bad it would be to leave things like this. Perhaps I should see what my load temperatures are at the stock speeds?

mexell · Post by **mexell** » Tue May 15, 2007 1:38 am

I'm happy with my temps on my E6420, Gigabyte 965P-DS3, Ninja Rev.B. But mounting it was a real PITA, once I thought that I've destroyed the clip but I haven't. Idle is about 27~29 (ambient ~22), load was up to 43, with gpu load up to 47 (that's due to my special cooling setup). But I found something interesting on Scythe's homepage (scroll down to Motherboard Compatibility for Ninja PLUS Rev.B). Maybe that could explain a bit...

spacey · Post by **spacey** » Tue May 15, 2007 5:57 am

spacey wrote:So I overclocked my E6600 to 3 Ghz but left the vcore at stock... whatever that is.

I'm using the Ninja Rev. B with a the Universal Retention plate with the copper spacer... I used a thin layer of AS5 of course. This isn't my first build... however it is my first Core 2 Duo.

Idle temps were around 30 degrees, which was cooler than the ambient temperature of my case (Lian Li V1100 plus with all fans throttled to slowest possible) however as soon as I started Orthos, the temperatures quickly climbed.

I've been letting it run for about an hour now and my temperature according to the Intel Thermal Analysis Tool is 67 and 66 degrees while Speedfan says its 68 and 68.

Did I do something wrong? I am pretty sure I mounted the heatsink on quite well. Back before I overclocked, my idle temperature was 25 degrees with the E6600, I thought that indicated I did things right but these really high load temperatures make me wonder.

If things don't break, I wonder how bad it would be to leave things like this. Perhaps I should see what my load temperatures are at the stock speeds?

okay turns out I was not at default voltage. it seems my motherboard likes to adjust the voltage itself when you try and overclock. i was at 1.42v which i think is a fair amount over the stock 1.325v i believe.

i guess given that, the temperature wasn't too bad. i have to find the setting somewhere to disable that to just manually set the voltage - but then i guess i lose the speedstep ability

Jasper · Post by **Jasper** » Sun May 20, 2007 7:03 am

spacey wrote:
spacey wrote:So I overclocked my E6600 to 3 Ghz but left the vcore at stock... whatever that is.

I'm using the Ninja Rev. B with a the Universal Retention plate with the copper spacer... I used a thin layer of AS5 of course. This isn't my first build... however it is my first Core 2 Duo.

Idle temps were around 30 degrees, which was cooler than the ambient temperature of my case (Lian Li V1100 plus with all fans throttled to slowest possible) however as soon as I started Orthos, the temperatures quickly climbed.

I've been letting it run for about an hour now and my temperature according to the Intel Thermal Analysis Tool is 67 and 66 degrees while Speedfan says its 68 and 68.

Did I do something wrong? I am pretty sure I mounted the heatsink on quite well. Back before I overclocked, my idle temperature was 25 degrees with the E6600, I thought that indicated I did things right but these really high load temperatures make me wonder.

If things don't break, I wonder how bad it would be to leave things like this. Perhaps I should see what my load temperatures are at the stock speeds?
okay turns out I was not at default voltage. it seems my motherboard likes to adjust the voltage itself when you try and overclock. i was at 1.42v which i think is a fair amount over the stock 1.325v i believe.

I've got an E4300 with a Ninja revB, currently not so thoroughly installed with some white thermal grease (the setup is still going to move to my real case).

It is specified with a VID of 1.325, and at stock 200FSB/1.8G runs at about 1.1V idle, 1.18 load, if left at auto on my p5b-e plus. Left at auto Vcore a modest overclock to 333/3.0 (66%), it runs 1.42 idle 1.32 load or thereabouts (notice the reversion of the idle/load relation, as well). It's not surprising, then, that it pushes a *lot* more degrees out.

I make it about 1.66 * ( 1.32 / 1.18 )^2 ~= 2.1 times as much dissipation, so that much rise in temps above ambient in the case (and the temporary case is not very well supplied with airflow, at that, which makes case_ambient climb as well).

TonyB. · Post by **TonyB.** » Sun May 20, 2007 10:27 am

I have 46C on both cores of my C2D [email protected] (8x375) @ 1.30Vcore, FSB/DRAM 1:1, on a MSI P6N SLI Platinum Ver. 1.0, using the Rev. B Plus and included fan at 1229rpm in open air. While listening to music, typing this post, and Memtest ver. 1.0.0.1 (testing 882MB memory), according to SpeedFan 4.32.
I'm happy!

TonyB.
_____________________________________________________
P6N SLI Platinum, MSI, MS-7350 Ver. 100
AMI BIOS v. 1.0, 01/02/2007 (a7350nms.100)
E6400 @ 3000.0MHz (8x375),
w/Sythe Ninja plus Rev. B SNJ1100B
2GB Patriot PC-6400/DDR2-800 PDC22G6400LLK
XFX Fatal1ty GeForce 7600GT, passive
2x HDD WDC WD800BB-00CAA1
1x LITE-ON DVDRW SHW-160P6S rev. PS09
Antec TP-430 PSU
Audigy2ZS
Presonus Firebox
M-Audio O2

2666.4 MHz (333.3x8) 1:1 ok
2932.8 MHz (366.6x8) 1:1 ok
3000.0 MHz (375.0x8) 1:1 Alright!
3200.0 MHz (400.0x8) oopso! So far!

Sainty · Post by **Sainty** » Sat Aug 25, 2007 4:57 pm

This is a bit of an old thread, but it would be nice to have those who purchased the revb follow up with any additional thoughts as to what they used to mount the Ninja or if they decided to return it.

The reason why I ask is that I have a revB in the mail as we speak to be used in conjuction with a GIGABYTE GA-P35-DS3P LGA 775 motherboard and Intel Q6600 CPU. Should I purchase additional mounting hardware or just try to RMA this thing ASAP?

SPCR's review of the Rev b was the deciding factor in me purchasing this particular model. If it turns out to be a bust (assuming that people haven't found solutions to this problem) could the administrators consider updating the original review with subsequent feedback? This isn't a blame SPCR post, but I might have thought twice and asked a few more questions before ordering if I had come by this thread earlier.

The last couple of posts were in May. We're in the latter part of August. Do any of you folks using 775 mobos and rev B's have any updates for this first-timer?

Please see this post/thread also:

viewtopic.php?p=364001#364001

Ryan Norton · Post by **Ryan Norton** » Sat Aug 25, 2007 9:24 pm

As I posted in your other thread, I would NOT say the Rev B is a bust on LGA775. It's just not the magic bullet (barring size restrictions etc.) it was for me on my old S939 Athlon 64 system. That said, I haven't seen any other coolers that 100% definitively ARE the magic bullet for LGA775.

whiic · Post by **whiic** » Sun Sep 02, 2007 4:19 am

Like Sainty, I as well plan to use Rev.B with Q6600. I will try it with stock clock, undervolted as low as it runs or as low as motherboard allows to undervolt.

I'm considering various of DIY methods to screw it tightly. I have no retailer around here that carries this Scythe Retention Kit, so one possibility is to just use 4 nuts and 4 bolts and 8 insulating spacers. Theres obvious drawback here as not only do I have to have motherboard out when it's installed, I would need to remove mobo again if I just wanted to tighten even one of the screws later (if I for example had uneven temperature readings from different cores).

Alternatively I could try something like mounting equipment intended for Zalman waterblock: http://www.quietpc.com/nz/images/zmcs2.jpg
I would just use the part that goes behind mobo and the screws, throwing away the rest. I would also need 4 insulating spacers and probably would need to cut the screws shorter (or alternatively use many, many spacers to do the trick). I think even if that is was plastic (at least it looks plastic) it could support Ninja... as long as tightening the screws don't damage the threads on screwholes.

The local retailers also don't carry any mounting sets that have metallic backplates and I don't want to buy another heatsink just to use the backplate supplied with them so I probably pay that 5 eur for the plastic backplate inteded for using with waterblock.

I would like to hear how people have mounted Ninja the DIY way, without all the fancy equipment and springs under the screws (which would CPU/socket damage impossible when overtightened). How tight is it safe to screw it without springs? I understand tightening the screws little by little is of outmost importance but there will still be a point when further tightening should be stopped (especially if the insulating spacers between mounting bracket and motherboard are so thin that the spacer isn't compressed (i.e all the compression force is directly to CPU heatspreader). Well, that would be bad anyway... considering that I would allow any external force to heatsink to easily damage the CPU/socket even after mounting has been done. So the spacers should be compressed and their thickness selected correctly to create correct compression on CPU. Spacer thickness is probably best if it's the same as the spacing with stock Ninja mount minus a fraction of a millimeter (to give slight increase over stock pressure). Then screws could be tightened pretty much as tight as the threads on the backplate can handle, not risking damaging the CPU as spacers would prevent such damage.

Did that sound logical or would you have some improvements on my idea of DIY mounting Ninja Rev.B?

Matija · Post by **Matija** » Sun Sep 02, 2007 4:38 am

I bought the original Ninja yesterday

But it's waiting until I resolve my graphics card problems

nzimmers · Post by **nzimmers** » Sun Sep 02, 2007 6:31 am

well, my Rev B isn't even flat - sure it's polished but I took a real 'flat' - a piece of steel that's milled to a high tolerance and placed it on the contact plate and it was a nightmare, light leaking through a gap in one area, a rise along one side that caused poor contact around the edges...

if you have a revB that works good, it might be because you have a contact plate that's closer to being flat then mine is.

overall, super unhappy with the scythe....won't be buying them again even if the mini ninja looks interesting, I can do without

Bories36 · Post by **Bories36** » Sun Sep 02, 2007 10:00 am

I lapped my Ninja and my E6600 and used AS5 and I get 28c Idle and 39c load, Compared to 40c idle and 54c load before.

grazzhoppa · Post by **grazzhoppa** » Wed Oct 10, 2007 2:27 pm

Ninja Revision B on LGA775 Mobo (Abit IP35 Pro)
Zero problems with installation of the Ninja using the push-pins that it came with.

I used Arctic Ceramique for the TIM. Supposedly this stuff adds 2-3C to the CPU temperature compared to Arctic Silver 5 paste.

Room temperature is 29-30C (85F), according to a wall mounted thermometer in an un-airconditioned building at 5PM.

Intel Core 2 Duo E6750 with the voltage & multipler scaling activated (C1E and EIST, I think are those features' names).

At idle & web-surfing & word processing, the CPU scales down to 2.4GHz and voltage drops to 1.17v.

35-38C, or 9C rise from the room temperature

At CPU load, the CPU scales to 3.2GHz and voltage returns to stock 1.3v.

59-60C, or 30C rise from the room temperature

18 minutes of CPU intensive tests:

StApostol · Post by **StApostol** » Mon Oct 15, 2007 5:03 am

I'm using a Ninja Rev B on an E6750 with AS5. The results are as follows:

Idle:
6x400 = 2400GHz, 1.187V, 23C (motherboard) or 37C (core). Ambient 22-23C.

8h Orthos Load (priority 9):
8x400 = 3200GHz, 1.325V, 55C (motherboard) or 67-68C (core). Ambient 24-26C, but I had two hard disks in the intake path of the Ninja, adding to the heat and reducing airflow.

Edit: I moved the disks (finally) and retested at full load (Orthos, priority 9). The result is an awesome reduction in max temperature by 7-8C, bringing the result in line with grazzhoppa's: 47C (motherboard) / 61C (core).

Normally, I only use the 6x and 7x multipliers (2.4GHz and 2.8GHz at 1.187V and 1.212V respectively), and the load temperature never goes above 45C (motherboard). I only use the 3.2GHz mode for intensive games (Bioshock, UT3), as it significantly increases power consumption.

I believe I would see a 4-5C decrease in load temperatures if I lapped and bolt-mounted the Ninja, but I'm not really concerned to do this right now. I will put a small duct however to capture some of the lost airflow of the Ninja's 120mm fan, which will probably result in a small (1-2C) reduction. I'm also contemplating a different airflow path (bottom to top, instead of front to back), mainly to reduce noise - I don't know if this will increase temperatures.