• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Milling my own pot

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

General_Kerr

Registered
Joined
Mar 1, 2012
Sempron 145, it defaults to 14. 150 is 14.5, 140 is 13.5 I believe.

You can use lower multipliers however. Which is useful to ensure what limit you are hitting - CPU or mobo.

Ah, I had the Sempron 140.


I've actually got a new dry ice pot design I'm going to machine tomorrow(if all goes well) and hope do do another run.

(I'm going to have to make a new set of mounting brackets though... so the run might not happen for another week or 2)

My original made it down to -94F (According to my thermocouple), and my new design has some improvements. (+ I wouldn't trust the reading because of how it was oriented)
EDIT:
Uhm.. actually the shop I do my machining in is closed, the supervisor needs some shots in his eyes... or he'll go blind.
Well... that's not the best answer when I ask if I can come over tomorrow :s


EDIT EDIT:

I.M.O.G Split this from the frosty pics thread

I swear, creating a thread was on today's Todo list. :p
 
Last edited:
Not bad, -70C for the pot, should be able to get that down to the -78C mark though. Or was that under load at -94F/-70C?
 
Not bad, -70C for the pot, should be able to get that down to the -78C mark though. Or was that under load at -94F/-70C?

That was under load.

1.65 volts, dualcore amd sempron @ 4.7 GHZ (second core was an unlock)

But, I accidentally dropped some toilet paper (off of that screwdriver) into it fairly early, and couldn't get it out... it screwed wit the temps.

My new pot design should be able to get even better temps.

Oh, and also... someone put kerosene in the acetone container... :cry:

I've scheduled next weekend to machine the pot
 
awesome, can't wait to see some pictures/internals. Any idea on the weight/size/materials/etc?
 
awesome, can't wait to see some pictures/internals. Any idea on the weight/size/materials/etc?

Any idea on the weight/size/materials/etc?

Weight: Derp, I'm redoing the design...

Size: Derp, 2.5" Diameter, Going to see what thickness will use least material without compromising the pot's function.
(It was the easiest to lift* from the college)

Materials:
-Main base: 5052 Aluminum (~ 75% the conductivity of your typical alloy copper, at 1/100th the material cost.)

-Holding Tube: 6061(Stronger alloy) 2.5 Dia, .125 walls

-Mounting Brackets: 1/4" thick 6061 Aluminum + 6-32 Althread



* With permission... :rolleyes:
(I'm buying a replacement piece too)
 
Weight: .3786 LB (Base Only) (Intentionally low)

Internal surface area (less walls): 9.3657 in^2

I'm still playing around with the design though.
 
From what I've been reading, if you can do a plastic-type (type of plastic escapes me) for the hold-down bracket [if you don't already have/use one] it is supposed to be better for frost/etc than aluminum.
 
From what I've been reading, if you can do a plastic-type (type of plastic escapes me) for the hold-down bracket [if you don't already have/use one] it is supposed to be better for frost/etc than aluminum.

Acetal

aluminum is easier to get.
(doesnt sweat when covered with petrolium jelly :p )

you'll see once i get pics.
The "Bracket" will be part of the pot, not able to be removed. the other one will have a piece of insulation between it and the mobo.
 
online metals will sell small amounts of copper, like 2"-4" cuts of it for a decent price, Yes its a more expensive metal and its different to work with, but it is possible to get small amounts of it just as easily. The copper in these situations makes a huge difference.

You really only need the base to be copper, the reservoir just needs to capable of sealing to the base in some for, (threads generally work quite well for this). Generally its bad practice if your looking for effiency to use a highly thermally conductive material for the reservoir tube. Atleast for pots aimed a LN2 useage. Dice pots are a balance between a moderate amount of mass and as much surface area as you can possibly machine into it.

Mounting mechanisms tend to work the best when they are high on the pots rather than near the base, this can help minimize the amount of icing that will occur around them. it also helps prevent clearance issues that you run into on a lot of the more modern high end designs.
 
Erm... you just basically took every point on my pot that is different from a typical design, and said it's bad.

(I'm not trying to step on your toes, I'm just stating my POV and reasons. I appreciate input, even if it directly contredicts my entire design. :) )

1. Aluminum instead of copper
2. Not using threads to seal. (I'm using a shrink-fit., It's permanent but will never leak)
3. Aluminum tube (Not much of an issue though, not intended for LN2, only dice dust should contact the tube)
4. "Moderate Mass" as for as ln2 pots go... I'm making an extremely light pot... with a ton of surface area for it's size)
5. The mounting mechanism is uhh... going to be about 1.5" above the processor. There's no other way to do it with the materials I've got on my desk.



So, Let me expound:

1.

Aluminum is ~1/3 the density of copper, and has ~60-80% the thermal conductivity of copper (For typical alloys)

The material cost for my dry ice pot will be about $3 for each base. If I purchased copper from online metals it would be $50 for each base. (Less Shipping!)

I can get copper, In fact I have about 25 pounds of it on my desk. I can even get it from the same place I'm buying my aluminum from. (Likely for less than online metals' prices)

I'm not convinced that an aluminum dry ice pot will be significantly outperformed by a copper one.

I reached -70C on my first try. (And my new design is totally different, for the better I think)

Also the thermocouple probe had about 1 1/4" of aluminum between it and the cavity, with just plain ice(i mean H2O) it registered ~5 C or something.(no pics) (No load)

(I had to work my way up to dice, plus I had to wait for someone to come back from the store with acetone :p )




2.
I Don't have a lathe, only a CNC mill. I would have to threadmill my parts to make threads (or get a 3" tap and die... no)

A Srink fit isn't too hard to assemble, just heat/cool the parts and slide them together. + my 1/8" tube wall isn't an issue (otherwise I need bigger material and thicker walls)




3.
It's intended for dice.

I would consider Acetal for the brackets/tube If I was considering making a copper pot for ln2.




4.
At .3 LBS this is an extremely light base, and it will be lighter when I finish the design.

It is my intention that this pot never run out of dice, the main cavity is designed to be able ot mix well with dice and acetone, the temp should reach it's point and stay there as long as the pot is topped off.

(Dice, It's not LN2 where you add X ammount every so often, you just keep it full.)




5.
My tube OD is the same as my stock OD, the only way to get a bracket on is to put it between the two... Saves $2 on allthread?

If I make an acetal top, I'll probably do the bracketing differently.

I don't see how frost is such an issue though, I ran for 3-4 hours on my dice run and the worse of it was the frost at the very top, there wasn't anything inside of my insulation, or near the processor socket. (but since everything is coated with jelly, It wouldn't matter if there was.)
 
Here is what my internals look like atm.

those holes are going to take forever to drill, but I can't get an endmill long enough to reach.

EDIT:

New Surface Area: ~16 IN^2,
Old Surface Area : 9.3657 in^2
(Surface area does not include walls, only the "Internals")

New Weight: .3578
Old Weight : .3786
(Base Only)


InternalsFirstPic.jpg
 
Last edited:
The first pot that i built was a few copper scraps that i cut on a lathe and shrink fitted them together. I eventually soldered them, but the shrink fit did work. I think the aluminum should get you good results, especially since you are designing it with plenty of surface area and using smaller cpu that doesnt have a real big load. Good work!
 
thanks guys!

I've had to put this project on the backburner, I got bored recently so i started cultivating numerous opportunities and projects... Apparently I've gotten better at convincing people! (I'm naturally an intravert, but i'm forcing myself out)

I've got 6 new website projects (one is for $2K, three are gratis, the other two still have pricing in the details!)

three machining projects, besides those at college. (I finished one on Friday though, for $50 for three hours of work.)

that and the family business (Injection molding) has a new customer and two new jobs that I need to be involved in. (Design/Mold making)

soo.... maybe this weekend, or maybe the weekend after that. there are still a few other projects that I volunteered for that im going to hear about tomorrow.
 
I'm looking forward to it. I actually quite like the idea. The number of holes in the floor of the internals might be a little much, but it should work very well with DI. :)
 
I'm looking forward to it. I actually quite like the idea. The number of holes in the floor of the internals might be a little much, but it should work very well with DI. :)

They aren't very deep though, but they add a TON of surface area.

I just hope they don't form gas bubbles.

I would prefer complex slots, but I don't have the $ freedom atm to be buying special Xtra long endmills.
 
The number of holes hes got shouldnt be a problem and should be good for DICE. Dice pots are mostly about surface area.

I think he said his base size is 2.5-3" OD, which is similar in size to the KV2. Kv2 also used a complex nozzle pattern its part of what makes it such a good dice pot.
 
I know it will be good with DIce (check the last part of my post ;) ). I was just mentioning it will be a little much if he ever moves to LN2. As I'm sure you know, the V2 is a PITA to control on LN2. ;)
 
with the few conversations i have had with General_kerr he seems to be focused on dice for now. I don't think he has any plans of running this pot on ln2.

But yes the kv2 can be a pain on ln2 due to the lack of mass. Though it works quite well for full pot runs on on amd chips. It doesn;t spit like the F1ee does at times. Plus it has more Storage volume. For those of us that tunnel vision during benching :D.
 
1. It's 2.5" Diameter. I would have used 3" dia but there was some 2.5" dia right handy!

2. Yep, focused on dice only. I don't have the $ or the hardware worth testing to do ln2 yet.



Anyone have an idea about what to do with that center slug?
 
Back