Base mass and thickness?

OverShocked · 10-24-2009, 10:03 PM

Hey guys,

Working on a new pot for dice/ln2, and wondering what kind of base mass do you guys use for an i7 sized heatload?

What kind of thickness would be suitable?

Ive already asked one person here, but wanted to get a few extra opinions.

der8auer · 10-25-2009, 04:03 AM

Hey

i prefer a thickness of ~15mm at the ground of the base and a mass of ~1,5kg. If you use more mass it will be good for LN2 but too much for DICE and you need long time to cool it down.

OverShocked · 10-25-2009, 10:07 PM

Thanks alot dude, ill mainly be using it for ln2 so should i look at a base mass of 1.7-1.8?

OverShocked · 10-25-2009, 10:57 PM

Hmmm, it seems that im missing something.

here is a shot of what solidworks estimates the mass. It says it 1.37kg. But the pot is 40mm thick and has a 70mm diameter. Keep in mind there are no holes in it yet, so that will dramatically reduce the mass.

How would i get any more mass in that w/o adding a center pole?

s e t h · 10-26-2009, 02:43 AM

in my opinion go for a 50mm high piece to start with...then drill your holes say 20mm into that....the more smaller holes the better as you get higher surface area with more holes that will therefor pull temps down faster.

OverShocked · 10-26-2009, 07:45 PM

Alright, ill try 50mm high tonight, What do you guys think about pins, instead of holes?

lesstutrey · 10-28-2009, 02:34 PM

Pins good for dice, not so much ln2... and i say, the more mass the merrier, if you wanna cool it down you can use a blowtorch or heat gun, it cools anything fast (ok it does waste a bit of coolant, but whatever, pour better

). I'd make that base thicker, like say 20mm-30mm more and then do staggered hole depth, it keeps your mass up after drilling, and might even allow for a mount low on the pot, which can be nice when cooling down, so you aren't worried about melting the mount

Though maybe i'm just a mass fan because i have 3,65kg pot.

OverShocked · 10-28-2009, 04:27 PM

That must be that custom duniek pot i saw. 3.65 kg is f'ing massive!

What do you mean by "staggered" hole depth?

Are you reffering to more depth on the sides and less in the middle?

voklskier4452 · 10-29-2009, 01:46 PM

not duniek his pot is made by otternase.. and his version is light version of the pot

lesstutrey · 10-29-2009, 06:58 PM

Quote:

Originally Posted by OverShocked

That must be that custom duniek pot i saw. 3.65 kg is f'ing massive!

What do you mean by "staggered" hole depth?

Are you reffering to more depth on the sides and less in the middle?

Exactly what i mean

, and even the holes next to one another could have ya know..5-10mm variance on the outside edge.. so it would have alternating hole depth around the circumference like this pattern: 20mm deep then 15mm deep then 10mm deep, then 15mm deep, then 20mm deep, repeat.
oh, and my pot is done by otternase, it's linked in my signature, it's in his pot thread too.. The full copper one weighs like 4.35kg.. somewhere around there

OverShocked · 10-30-2009, 11:47 AM

Hmm ok, ill try what you just said and see how much mass i end up with.

OverShocked · 10-31-2009, 01:58 AM

Alright, check this out. What do you think?

The top is supposed to be alu, but there is no way to do that in the current version of solid works.

Base weighs in at 1.851 kg.
The whole pot wieghs about 3.5kg.

^cutaway.

^complete pot.

Constructive criticism welcomed.

s e t h · 11-01-2009, 11:49 PM

more holes in my opinion....the heavier the base the more surface area you need if you want to be able to lower temps while say running full tilt on an i7.
otherwise i think its going to be good...you have the primary mass right over the core where it make the most difference.
if you have too little surface area it will become nearly impossible to lower temps while benching under full load.

OverShocked · 11-02-2009, 12:18 AM

Alrighty, thanks for the reply. I threw that old design out a few days ago and came up with this.

I think this does alot better, because it has about 1.5x as much surface area with still 1.7kg of mass.

What do you think? Should i add a little more mass to it?

s e t h · 11-02-2009, 12:55 AM

more mass is almost always a good thing yes...basically your temperature swings to the upside will be slower the more mass you have.
say the nitrogen has evaporated from the last pour then only the mass is preventing the cpu temp from rising....the more mass the slower that rise.
any hole drilled into the base doesn't need to be much wider than say 6mm...bigger holes just reduce mass and add a less surface area.

the only time mass works against you is when you coldbug a cpu...then you want the block to warm up as fast as possible to boot again.
more important in my opinion is to avoid coldbugging the chip and have as much mass as possible so that temp swings are reduced.
but as you know more mass takes longer to cool unless you have a really large amount of surface area.

OverShocked · 11-02-2009, 01:03 AM

Yeah...

Do you mean holes 6mm in diameter or radius?

s e t h · 11-02-2009, 01:05 AM

diameter
i've used 4mm diameter with 12mm depth which worked perfectly and can really pull town temps...but then i had about 100 holes.
yours appear deaper which is why i say 6mm...you want the expanded gas to be able to leave the hole without blowing all the liquid out with it.

also using drill bits for my holes resulted in many broken bits...6mm diameter bits would be stronger if drilling in that manner

OverShocked · 11-02-2009, 01:13 AM

Alrighty, i made a couple of edits... what do you think?

By the way, thanks for all the help! Ive been studying up on heat transfer but alot of it doesnt apply....

Ive seen pots with a larger center hole and i seems to work well.

s e t h · 11-02-2009, 01:28 AM

i think its a promising design...if it was my block i would reduce the size of the larger holes and then drill more holes overall.
my last block had about 300cm2 of surface area (total area contacted by nitrogen excluding any side walls above the base)...if i was you i'd be aiming for something like that as well...particularly since you have more mass than my last design.

i also think i preffered your earlier drawings where the vitrical edges appeared talled which to me means more mass...ie post#12 'cutaway'

OverShocked · 11-02-2009, 01:48 AM

I didnt like those earlier pics because they only have a little less mass and a hell of alot less surface area.

Specs:

Post #12:

Base mass:1.851 kg

Base surface area: 28496.79 millimeters^2

Post #18:

Base Mass = 2153.31 grams

Volume = 241944.44 cubic millimeters

Base surface area:42279.98 millimeters^2

Post #18 is 78mm square, so its 8mm larger than post #12.

post #18's base thickness is larger aswell.

OverShocked · 11-02-2009, 01:49 AM

Ill work on a design that is like post #18, but is 70mm square like post #12.

s e t h · 11-02-2009, 02:07 AM

yeah...70mm is better...once you get any bigger than that it often becomes difficult to insulate the block where it meets the northbridge 'sink.
your specs are looking good though..2kg's with over 400cm2 of surface area...your block is gonna rock

OverShocked · 11-02-2009, 02:07 AM

Here, How about this?

Mass = 2025.90 grams

Volume = 227629.52 cubic millimeters

Surface area = 35107.27 millimeters^2

OverShocked · 11-02-2009, 02:09 AM

Quote:

Originally Posted by s e t h

yeah...70mm is better...once you get any bigger than that it often becomes difficult to insulate the block where it meets the northbridge 'sink.
your specs are looking good though..2kg's with over 400cm2 of surface area...your block is gonna rock

Yeah lol.

The 78mm monster was for an x58, the mounting for that is 80x80.

s e t h · 11-02-2009, 02:14 AM

latest image looks good...cut of corners will help with insulation as the edge of the block won't be compressing the insulation too hard at the corner.
if you can somehow get more surface area you are unlikely to regret it....but i think the 340cm2 you have will work fine...it'll certainly allow temps to drop even under load.