MaxxxRacers Test Rig WorkLog

[ExtraPickles]

I talked it over with the epoxy and he felt it was best to go el Naturelle.

Nudist talking epoxy FTW!!

-pickles

[MaxxxRacer]

OCme, I am using Omega T-type Bead probe thermocouples. (Part # 5SC-TT-J-36-36 )

The epoxy used to hold them on is Omega bond OB-101, which is what intel reccomends for this. In intels newer documents they say to use OB-100, but OB-100 does not conduct heat as well and does not creat such a strong bond, like the OB-101

Gocchin, I am using an Olympus C-8080WZ and no, I am not using natural light. I have two movable work lights that I attach to the table, which use 125watt (equivalent) flourescent bulbs. I then use the cameras built in color correction to make it seem like the flourescents are producing perfectly white light. Generally the coloring is just a wee bit off after I am done with the pictures, so I go into photohop and clean it up with the auto-color correction and adjust the brightness.

[creidiki]

More effort than im willing to do fer some PC pics, thats sure ^^"

[Bloody_Sorcerer]

that CPU looks like hell... give her a good lap dance-errr, job.

[creidiki]

Is that your secret OC-fu then? Giving your CPUs lap dances? ;p

[MaxxxRacer]

Bloody, the cpu actually already went through a 2000grit lap job.. it just looks wierd because i didnt lapp it far enough to completely get rid of the crappy casting job that AMD does on those IHS'. (actually Foxconn makes them)

[KoHaN69]

UPDATE: I tested the 3000 winchester and everything looks good. At 2.25ghz ~1.6v I get idle temp of 22.7C and load temp of 25.8C. Thats an increase of 3.1C from idle to load... Ambient was 17.8C with fluctuations between 17.8 and 17.9 during the brief test.. To be honest I am quite shocked at the incredibly low temperatures. 8C over ambient for full load temps for as close to actual core temp as your gonna get is pretty dam good in my book..

Erm, question!
I thought the temperatures don't have a linear ratio.
What if ambient is 30C? and idle is cpu is 25C?
Would the flunctuation increase linearly?

Just wondering...

Very impressive results nonetheless though :thumbsup:

[MaxxxRacer]

CPU cant idle at 25C if the ambient is 30C... just not possible..

and what do you mean by "linear ratio" and "fluctuation increase linearly"

[skaktuss]

linear:non-linear:

[MaxxxRacer]

lol.. no comments skaktuss. I was asking what he was refering to with the linear comments.

[skaktuss]

he wanted to say, that if u will increase ambient temp with 1C steps cpu temperature function wont be linear!

[MaxxxRacer]

well of course a 1C increase in ambient wont net an exact 1C increase in cpu temp..

[KoHaN69]

he wanted to say, that if u will increase ambient temp with 1C steps cpu temperature function wont be linear!

exactly.
meaning, that the difference in load/idle temperature might be different at a higher ambient temperature

[OCme]

I wonder if altitude & or barometric pressure would also have an effect on load/idle temperature?

I can see where other variables would come into play...

[Erasmus354]

exactly.
meaning, that the difference in load/idle temperature might be different at a higher ambient temperature

That doesn't necessarily mean that it isn't a linear ratio. It could still be linear, just not at a perfect 1:1 ratio.

[MaxxxRacer]

OCme the barometric pressure will have a very minor effect.. denser air carries more energy, and thus better cooling..

humidity also has an effect... higher humidity = more cooling ability.

[KoHaN69]

OCme the barometric pressure will have a very minor effect.. denser air carries more energy, and thus better cooling..

humidity also has an effect... higher humidity = more cooling ability.

But... My question remains unanswered

At different ambient temp, would the load/idle difference be the same?

[creidiki]

I bet a slice of cake the delta will increase, for a given thermal load, as ambient increases, as long as pressure and humidity and everything else are the same.

[MaxxxRacer]

Tell ya what.. the next time i have the test rig up, I will test it out..

but frankly as of now I just dont know.. never seen any tests done on it and dont have any of my own..

What I can say is within 5C, I would imagine (no proof of course) that the delta will be the same.. its when you start getting to the extremes, do the thermal characteristics start to change.. BTW, by extremes i mean about 15C in either direction from 25C.

[creidiki]

Btw, whats the eta on the 1st batch of tests? :p

[MaxxxRacer]

Still waiting on the flowmeter + fittings from the UK, and the chiller is borked for the moment.. it tries to set its temperatuer to a continuingly dropping temp.. i think the rheostat is busted. I will be desoldering it and checking that tommorow..

But as soon as the flowmeter comes in and the chiller are working, testing will begin..

Sorry I dont have better news.

[creidiki]

Someone posted a pressure drop curve for the swissflow over on the O-CuK forum... 'ere..

Thing is, i cant read that graph. Is that good, or bad, or just decent?

[MaxxxRacer]

the pressure drop on the swissflow is pretty rediculous.. but its what i can afford for the time being.. at 2gpm the thing soaks up an amazing 16ft of head. An RD30 at full power wouldnt stand a chance of getting 2gpm with this thing i dont think.

[creidiki]

Uuuuuhhhh.. ok. I'll file under "dont get one" then.

[MaxxxRacer]

ya.. its only useful for testing. I suppose I could create a pressure graph vs flow so that i can just lookup the flow value, but that would require that i buy another dP meter and create the setup for that, which i just cant picture myself doing financially or sensibly.