what are the black ice radiators made of? copper or aluminum?

[chinkgai]

or are they a mix of both? cuz i read that they have copper fins, but are only the fins copper while the rest is aluminum?

[k00lance]

opposite.
Aluminum fins.

[skycrane]

Here you go...

http://www.dangerdenstore.com/product.php?productid=1&cat=13&page=1

double row of copper core w/ copper fins

[MaxxxRacer]

copper fins and copper tubes.. brass tanks and something like tin for the mounting bracket.

but there is NO aluminum in the radiator.

[chinkgai]

opposite.
Aluminum fins.

:slap:

[chinkgai]

copper fins and copper tubes.. brass tanks and something like tin for the mounting bracket.

but there is NO aluminum in the radiator.

thanks maxxx, i was awaiting your response ^^

[Disposibleteen]

copper fins and copper tubes.. brass tanks and something like tin for the mounting bracket.

but there is NO aluminum in the radiator.
If there was aluminum in them Maxxx wouldnt suggest them becuase they would assist in galvanic corrosion ;)

[skycrane]

yes... Maxxxx and his gavanic corrosion.. :toast:

[chinkgai]

yeah thats what i was worried about...

anyways, want to hear an interesting story?

so i had a 4200+ with with my bix3, tdx, and mcp655

i had temps of 44-45c load at about 1.623v and 2751mhz prime stable, this is with water and a couple capfuls of water wetter only

now i have a 4400+ with all else the same, but the difference now is that i've drained my system and added a bottle (2 ounces) of swiftech hydrx to my liter of water in the loop with no water wetter this time

both of the cpu's had their ihs removed, the 4400 had a much bigger core...so i expected it to be hotter than the 4200+

my results? 4400+ priming at 1.586v right now with load of 37-38c only!! its not going to jump 6-7 degrees even if i added more than 1.623v as i did with the 4200+ because i was at 1.65v @ 2925mhz with the 4400+ with full load of 38 that didnt alternate back to 37 at all

so what do u think changed the temps dramatically? was it the hydrx? or did i just had the 4200+ badly seated?

[k00lance]

I got confused with koolance radiators.
They use aluminum

[BillA]

all wrong

brass tubes and headers
copper fins (only)

and I cut them up so I should know

[Psyche911]

anyways, want to hear an interesting story?

so i had a 4200+ with with my bix3, tdx, and mcp655

i had temps of 44-45c load at about 1.623v and 2751mhz prime stable, this is with water and a couple capfuls of water wetter only

now i have a 4400+ with all else the same, but the difference now is that i've drained my system and added a bottle (2 ounces) of swiftech hydrx to my liter of water in the loop with no water wetter this time

both of the cpu's had their ihs removed, the 4400 had a much bigger core...so i expected it to be hotter than the 4200+

my results? 4400+ priming at 1.586v right now with load of 37-38c only!! its not going to jump 6-7 degrees even if i added more than 1.623v as i did with the 4200+ because i was at 1.65v @ 2925mhz with the 4400+ with full load of 38 that didnt alternate back to 37 at all

so what do u think changed the temps dramatically? was it the hydrx? or did i just had the 4200+ badly seated?

Well, I can't remember if the cores (ie. Manchester, etc.) on the 4200 and 4400 are the same or not. That would be one reason. I bet a big part of it is because of the larger core. It means more surface area which results in greater contact with the waterblock and better heat transfer.

I've longed for CPUs to be thinner and larger in area. Think if they had 4-5 square inches of surface area and a block to match. :D

[chinkgai]

i'd have thot bigger is hotter, weird.

as for cores, i've read at anandtech that some 4200+ are toledos that had too many manufacturing defects in the cache and were thus downgraded while others are purely manchester cores. duno if that makes a difference tho.

[Psyche911]

Well, sort of.
Older processors (130nm+) result in larger cores and use more electricity, therefor running hotter. That's not strictly because of their overall size though.

With the same manufacturing processes and tricks, the more surface area the better, as far as heat transfer. The problem is with wafers, size matters. The bigger the cores, the fewer per wafer, and higher cost.

I think I know why the size is different though. Aren't the cache sizes between the two processors you've had different (512KB vs 1MB)? I believe cache takes up about 50% of the processor. So that would explain why one is quite a bit bigger than the other.

[HiJon89]

It could just be you're getting better contact now.

[MaxxxRacer]

hmm.. bill.. i need to have another chat with our friend over at armelin widget co...

he says his radiators have copper tubes.. when i cut it up i thought it was brass... didnt look like brass..

[aMp]

There can be a lot of variation between cores. You probably could've seen the same temp difference between two 4200's, from different batches, anyway. My Venice 0516 EPGW overclocks poorly (2.7 under H20) and gets ridiculously hot over 1.6V, while better steppings run cooler and O/C farther.

Also, just a theory that Maxx or someone else might be able to evaluate: If the 4400's die is larger than the 4200, then you got a Manchester instead of a rebadged 1MB Toledo. The Manchester has a TDP of 95W to the Toledo's 110W, so the 4200 theoretically puts out less heat. But the Toledo's heat is spread out over a larger area -- a 199mm2 die, some 26 percent larger than the Manchester's 147mm2. Particularly with the IHS removed, this should make for more efficient heat transfer from a Toledo core than from a Manchester.

I've got no idea if that actually works in the real world. Has anybody else run both Manchester and Toledo cores in the same system?

-aMp-

[saratoga]

The bigger core is going to make a noticable difference for watercooling. With water, its all about w/m^2.