Idea for evap.What do u think?

[clonning]

Hi guys, I have no experience with it but I have idea of physics and thermodynamics laws. i think that might be stepper design , but I tried to enlarge gas-copper contact surface. Looks like tiny radiator. i am not still sure if the refrigerant will evaporate at these tight spaces between the "ribs". i will use it on 775 socket. What do you think about the dimensions, are they accurate? I was amazed of the many different designs at that forum.... But I noticed that simple things might work better ( i mean about the evap design)!

[sdumper]

Looks like a classic stepper whats new the measurements?

[quintus]

Too thin, i use 36mm evap and Piotres has a 54mm:P

[Ian[Frozen]]

Which works better, the top down method or the bottom up like this one??

[clonning]

The bottom up is for CPU i think. I am not sure if the gravity affects on the gas flow. And the up- bottom ones are for GPUs as long as I know, most of the GPU mountings are upside down

[-cyclone-]

Too thin

totally agree with that. Evap's mass would not be enough to hold big loads properly.

[Coyote]

totally agree with that. Evap's mass would not be enough to hold big loads properly.

mass doesnt hold load, heat transfer does. and an efficient evap with great heat transfer very close to the heat source will have a very small temp variation from the source to the evap.

@itor evaps had really thin bases, and those are extremely efficient. you can see it on my avatar.

[n00b 0f l337]

Yeap, mass just can give the illusion by reducing temperature swing, it will still swing, just at a slower rate.

[teyber]

heres my larger version of ssilencers evap.
ssilencers the first one who came up with that design...

[-cyclone-]

mass doesnt hold load, heat transfer does. and an efficient evap with great heat transfer very close to the heat source will have a very small temp variation from the source to the evap.

but.. how do you imagine good heat transfer from the evap upstairs via quite thin but surfaceable zone?
I've tried and created several excellent LN2 pot designs and I stay that thermal conduction and 'cold accumulation' really matters in xtreme cooling. Anyway I still have some lack of experience in Ph-Ch, especially evap designs, so appreciate all the info you may provide here -)

[Coyote]

actually the thinner the base, the better its the heat transfer into the refrigerant. of course there is a limit of how thin you can go, cause the base starts to bend. but thats a mechanical issue, not a thermal issue. plus you must have a good disipation surface as close to the source as possible.

on an evap, you have a refrigerant imput, and a refrigerant output. so basically you just concentrate all of the cooling power, as close to the heat source as you can. of course, with a good disipation surface too, thats why we have those extral layers. but the bulk of the heat transfer is done close to the base.

pots are a different story, cause you cant never reach the focused efficiency you can have on a evaporator. thats why you cool down big chunks of metal in order to hold that temp as much as you can. plus theres always the "directional" issue on the pots.
on a pot, you cant do this focused cooling. specially for LN2. gaseous N2 "wants" to go in the oposite direction to liquid N2. plus there is no flow, but you just store the coolant. so in that case, having a big chunk of cold metal is the only real option.

but on a evaporator, the less copper you have between the heat source and the refrigerant, the better. if you have too much copper, it end up acting as an insulator, instead of a conductor. and thats not what you want, cause it will cause an increase of temperature on the heat source.

[sdumper]

actually the thinner the base, the better its the heat transfer into the refrigerant. of course there is a limit of how thin you can go, cause the base starts to bend. but thats a mechanical issue, not a thermal issue. plus you must have a good disipation surface as close to the source as possible.

on an evap, you have a refrigerant imput, and a refrigerant output. so basically you just concentrate all of the cooling power, as close to the heat source as you can. of course, with a good disipation surface too, thats why we have those extral layers. but the bulk of the heat transfer is done close to the base.

pots are a different story, cause you cant never reach the focused efficiency you can have on a evaporator. thats why you cool down big chunks of metal in order to hold that temp as much as you can. plus theres always the "directional" issue on the pots.
on a pot, you cant do this focused cooling. specially for LN2. gaseous N2 "wants" to go in the oposite direction to liquid N2. plus there is no flow, but you just store the coolant. so in that case, having a big chunk of cold metal is the only real option.

but on a evaporator, the less copper you have between the heat source and the refrigerant, the better. if you have too much copper, it end up acting as an insulator, instead of a conductor. and thats not what you want, cause it will cause an increase of temperature on the heat source.

How much of a base thickness are you finding optimal.

[Coyote]

im using @itor's evaps (same as ssilencer used to use), and if my memory is good, the base is about 2mm. maybe a bit less than that.

keep in mind, that since all mother cpus have a IHS, you have almost 3mm of metal between the heat source and the refrigerant, so thats quite a bit.

many years ago, when using this evap over an a64 without IHS, the delta between cpu temp, and evap temp was really small (less than 10c). while testing, the same system using a chilly1 spiral, had way over 20c of delta, and that evap desing had a big chunk of copper over the core.

[sdumper]

What are the holes in the evap picture in your sig used for?

[Coyote]

those where from the earlier model, just for the looks basically.
you have those holes on the last plate basically. just to get the evaporated refrigerant to the cap, and then the suction line.

@itor could have just do the same thing as the other layers have, and it would serve the same purpose. but he has this obscesion with the way things look, even if he is the only one that gets to see it (cause once the evap is brazed, you cant really appreciate it). old man shenanigans i guess :P

[-cyclone-]

Thanks Coyote, another piece of good theory-like practice was quite useful -)
As I see evap is much closer to the waterblock than to the LN2 pot. Will have to refresh my friend's recent investigation data in mind