CPU pots for Dice

[Brolloks]

So I have been using the Koolance V2 pot for Dice for a while now and wanted to see how it compares against some other CPU pots....note this is for dry ice only.

I did not do any analytical tests or charts and stuff, just want to share some of my experience using the various pots.

Pots Used:

Koolance V2

K|ngp|ns's F1EE

K|ngp|ns's Evo 2.4

CPU - 980X on RIIIE


COMMENTS and EXPERIENCE

Koolance V2

Pro's:

The Koolance pot from a temperature perceptive managed to hold the temps much better than the other two, between -72C and -75C. It depth and width of the pot seems to be a prefect balance to avoid acetone from blowing out, it also keeps the dice/acetone in slurry form which is imperative to keep the temps stable and low. Frosting of the outside of the pot was no problem at all. Also the pot has a nice capacity so you can really load up on the dice which reduces dice addition frequencies. The spacing and size of the holes at the base of the pot works also well to balance optimum cooling keeping the mixture slurry.
The pot is nickel plated which does not tarnish like the copper pots

Cons:

The width of the base makes it challenging to fit especially on boards with caps close to the socket.
The bolts that came with the pot is completely unpractical, I like the back plate and bolt configuration that the Kingpin pots has, simple and effective. First thing you do after getting the pot is to ditch the mounting bolts and get simple 5 or 6 “ threaded bolts.
Seeing that the acetal mounting bracket comes down to a third above the socket frost seems to collect just above the socket, so be careful, I just wrapped absorbent towel around and over the bracket joint to collect any moisture.


K|ngp|ns's F1EE

Pro’s:

Pot has a good surface bas that fits nicely on the CPU HIS. It is full copper and heave so no problem with making poor contact and being copper it is great heatsink (i.e. for drawing heat away from the CPU).
Bolt, back plate and rod configuration makes it really easy to mount and un-mount.

Cons:

The F1E took a long time to get the temps close to -70C and did not stay constant either. Also without the AL adaptor it is near impossible to get the acetone to stay in the pot, especially in the cooling down phase, it would blow out like a volcano, even with the AL adaptor I still had to put a rag over the mouth, the rag ended up soaking with acetone.
The main problem with the F1E was that I could not get the dice/acetone mixture, the dice tend to clog up the holes and then the temps would just shoot up as no acetone was cooling.
Also temps would hover around -65C and would rise quickly if I do not pay attention.
It was really difficult to use with dice and my poor OC results reflected that.

Not a major issue but the copper base tends to tarnish up rather quickly so that requires some maintenance every now and again.


K|ngp|ns's Evo 2.4

Pro’s.

Easy to mount as it has the same mounting hardware as the FiEE. Pot base area that makes contact with the CPU’s HIS is also nice and small so no interference with the caps around the CPU socket area. Pot is long enough to prevent any acetone blowout.
Pot base is copper which ensures good heat transfer.

Cons

Although I did not have any acetone blowout the Evo could not get as low in temps as the Koolance, also it would not hold the temps stable for a good period of time. The lack of holes in the base also reduced the heat transfer and it was evident in the performance. Keeping the dice/acetone a slurry was not as challenging as with the F1 but unless I kept stirring the mixture it would lock up and temps would rise as a result. The width of the pot is on the small side so that also contributed to the lack of slurry formation. Also I could not load the pot as that would immediately result in the dice to pack on and loose the slurry mixture one needs.


My Conclusions:

All in all I very much favor the Koolance pot from ease of use and effectiveness when using dry ice. It kept temps low, maintained good slurry and allowed me to focus on benching rather than keep the “pot going”.

Hope this helps all would be dice users out there.

[dejo]

Nice work Ed, this should really help some that are pondering getting into sub zero cooling.

the Koolance is the only pot that I have ever used but have been really happy with it once you get it properly seated on the cpu ihs. It seems to be really effective at the job it was designed for. I do wish that there was a bit of a taper on the base just outside of where the ihs resides when mounted as that would alleviate the problems of many boards with the close caps. For the price and ability this pot seems to be a great value.

[Bobnova]

Nice work Ed, this should really help some that are pondering getting into sub zero cooling.

the Koolance is the only pot that I have ever used but have been really happy with it once you get it properly seated on the cpu ihs. It seems to be really effective at the job it was designed for. I do wish that there was a bit of a taper on the base just outside of where the ihs resides when mounted as that would alleviate the problems of many boards with the close caps. For the price and ability this pot seems to be a great value.

My thoughts exactly.

Thanks for testing, it's nice to have a direct comparison finally.



EDIT:
Also, ZOMG GREEN TEXT IS ONLY FOR MODS!!!!!!1!1111!1!(......... :D)

[Brolloks]

^Darn, you noticed:chair:...there are only three bright colors and lime green happens to be one of them, and my favorite:bang head

[Bobnova]

QUICK somebody promote our Glorious Leaders to moderator status!

[ghost_recon88]

Koolance Rev.2 for sale over @ XS for $90 shipped, FYI.

[Khemikal]

Seems to mimic the other comparisons of the Koolance vs Kingpin pots on DICE I've read. Koolance is better temp wise but needs work in the mounting department, plus the base can cause problems with mounting on some boards.

[Bobnova]

Damn that's a good price. Beats the hell out of any other $90 shipped pots, that's for sure!

[GoD_tattoo]

Nice job with the info...I know making the leap into sub zero is a jump and it is nice to give people an idea on which pots are better for dice and which are better for LN2...Thanks! Gotta throw in the Ryba cpu pot I got recently...It works great... I like it better for dice than the F1ee the team has...Seems to get colder and stay colder...Never measured it, but it seems that way..

Pictures would make this thread uber awesome BTW.....

[EarthDog]

Great info Brollo! :)

[dejo]

if they would just turn a little of the outer edge of the bottom the Koolance V.2 would be the ultimate dice pot and many would be hard pressed to do any better. the mounting isnt as bad as some say, its usable (a pain, but usable). My only real gripe is the fitment on the boards that I have, as I have to make sure that I dont get the eraser very thick over the caps or I struggle to get a good mount. When I benched one of my procs, I barely had any contact at all, about .75 inch circle on the paste afterwards. thats scary

[Bobnova]

I would like to see a bit more shaved off the edge like Dejo said, i was tempted to shave it myself, really.
I'd also like to see straight threaded rod, their mounting is a pain on 775. On 1366 it'd be a lot nicer as there is more room.

On my P5Q Pro i finally just started putting a .25* slab of copper under it to raise it up. Works great, waaaaay better clocks in 32m and wp1024.

[dejo]

if you did it yourself, it would compromise the plating.

[Bobnova]

Yup, that it would. That's part of why I didn't.
Mainly though I didn't want to risk making a hole into the interior and killing it.

[El<(')>Maxi]

The Koolance pot from a temperature perceptive managed to hold the temps much better than the other two, between -72C and -75C.

These numbers are at idle right? What about when you hit it with a load like Vantage CPU test? What CPU frequencies are we talking about? For sure the F1 is not a DI container, it's designed for LN2. I'm surprised to see a container with a base like the koolance doing so well on DI actually, conventional wisdom for a good DI container is a flat base. I expected to see the Evo doing better quite honestly among these three units. But hey I trust your results so I'm interested in hearing more about the Koolance under load.

[dejo]

another possible variation that I havent seen discussed is the dry ice itself. I usually use the penguin brand, they have some in a blue bag and some in a red bag. the red for me seems to beat down to powder way better and may be part of the reason that I managed to get 5400 out of the 920 and was 2 or 3 bclk lower with the blue. I have had the red 2 different times and just seemed to get better results with it.

[87dtna]

Koolance Rev.2 for sale over @ XS for $90 shipped, FYI.

Wow thats a sweet price.

Ed I have the cheapest simple BartX pot if you want to test that too I'll ship it to you. I'd like to know just how much better the V2 is. Although it doesn't have 1366 mount though, only 775/1156 and AM2/3.

[Brolloks]

These numbers are at idle right? What about when you hit it with a load like Vantage CPU test? What CPU frequencies are we talking about? For sure the F1 is not a DI container, it's designed for LN2. I'm surprised to see a container with a base like the koolance doing so well on DI actually, conventional wisdom for a good DI container is a flat base. I expected to see the Evo doing better quite honestly among these three units. But hey I trust your results so I'm interested in hearing more about the Koolance under load.


This is with a 980X at 5.4 ghz running Wprime 1024M, temps do NOT go above -70C, at idle it creeps up to -76 C, I have had it sit at -78 on idle if the mount is good, having a good mount is a challenge due to the ridiculously wide base.
Given I did not log the temps analytically for each pot , as I said it is just my experience.
In both the EVO and F1 the problem was keeping the dice/acetone a slurry, it kept "solidifying" which we all know is the achilles heel for effective heat transfer. The acetone is the transfer medium, if the dice solidifies on top of it the acetone in contact with the base do not get cooled effectively.

[El<(')>Maxi]

What kind of temp measuring gear are you using? Never seen a container that cold even @ idle, -73° or so is the norm with the system off on a FLUKE 51II with k-type themocouple. I've never benched the Koolance but looking at the base I'm baffled how you can maintain a better mix with it vs the Evo, looks to me like those holes will plug just like the F1. But all that doesn't matter really...and temperature is relative. What does matter is performance. Can one container hold a load better than another and in doing so maintain higher frequencies? If the Koolance is only swinging around 4° idle to load in wPrime that's pretty good. R2 Mousepot with a flat base swings about 5-6° under similar benching conditions.

[Brolloks]

I have a recently calibrated Fluke 50D as well as a brand new TENMA with all new K-Type thermocouples, used the Fluke on all three pots. The holes on the Koolance are not as deep as on the F1 and narrower, so it does not clog up at all. I'm sure the F1 is a fantastic pot on LN2 but for me it sucks on dice, period.

[Bobnova]

I haven't found the holes to plug, I don't know why not.
My guess is that the expanding gas bubbling from the bottom (closest to the heat source, and hence the hottest) clears the them.

[Gautam]

I thought the Evo would do a bit better, F1EE is no surprise...you can tell just by looking at it that it won't do well with DI. You should be comparing raw overclocks though. I never liked the idea of compare base temp a) vs base temp b).

[Brolloks]

That is just it, with the Koolance I could get to 5.6-5.7 without too much trouble, in fact I had a saved profile for 5.5 for all my 3D03/05 and AM benchies, with the F1 and Evo I could not run those at 5.5, in fact I could barely get into windows.
Anyways as always YMMV so not saying the F1 and Evo are bad pots, it just did not work for me.

[ChanceCoats123]

Other than the white text being atrocious, great write up. I want to get a pot and this helps me decide on the DIce pot so it'll not only be a bit easier to get the lower temps, but keep them there. :thup:

[Gautam]

That is just it, with the Koolance I could get to 5.6-5.7 without too much trouble, in fact I had a saved profile for 5.5 for all my 3D03/05 and AM benchies, with the F1 and Evo I could not run those at 5.5, in fact I could barely get into windows.
Anyways as always YMMV so not saying the F1 and Evo are bad pots, it just did not work for me.
Ah. Well, once again I'm not that surprised. The Evo was originally designed for use on a single core AMD Athlon FX57 after all...5 years ago. I too noticed it really struggling even on a Celeron Cedar Mill at ~6.2-6.4GHz a couple of years ago, so I guess it's to be expected that it won't be able to handle a Nehalem or Gulftown too well. Vince is coming out with a new dry ice base just for this reason.

The Koolance is a solid performer for sure from when I've seen in it action. They did a good job, though tbh they drew upon a lot of past ideas from the community, and it's basically the newest design out atm.

[Brolloks]

I'm looking forward to use the new Gemini from Vince, looks like a very versatile pot.

[ChanceCoats123]

Just wondering Ed, but approximately how many lbs of DIce did you use and for how much time? I'm just doing some mental math, but 980x is a 140watt cpu at stock with my 702 being a 95. So I would get roughly 3/2 times more cold out of the same amount of DIce, or since you can't really get much colder than the temp it sublimates at, I would get roughly 3/2 more time benching out of the same amount. I'm really getting the itch to get my own pot too. ;)

[Brolloks]

20 lbs of dice last me a good 4-6 hrs continious benching, I normally use 10lbs in the evening for about 2-3 hrs and keep the rest for the next morning and can still do a good 1-1/2 hrs, some of it sublimates off course over night.

[ChanceCoats123]

Oh very nice. So with my mad math, roughly 13 lbs will get me the same 4-6 hours, or I could buy the 20 and get 6-9 hours. Sounds fun! :evil laugh:

[Brolloks]

Depends on a lot of factors, ambient, OC level, insu;lation, dice particle size , air cirulation in room etc, but yea you should get 20-30% more on the Phenom

[ChanceCoats123]

:clap: Thanks Ed. :thup: Great write up by the way.

Ps. Did I mention the White is atrocious?

Here's what I get. :rolleyes:

85595

[Brolloks]

What do you mean, do you see a white background? Strange...

[ChanceCoats123]

No, I have the forum skin set to white because it brightens my overly dark basement, and your white-gray text is blending in.

I read it simply by ctrl+a'ing that hoe.

85596

[Brolloks]

I wanted the text to be all green but that would aggrevate the mod colony ;), LOL.

[ChanceCoats123]

LOL good stuff! Anyway goodnight all.

[Gomeler]

You'll have even more fun with the pot under LN2 if you can find a chip without a coldbug. This pot will coldbug CPUs that normally don't coldbug on some other pots out there.

These numbers are at idle right? What about when you hit it with a load like Vantage CPU test? What CPU frequencies are we talking about? For sure the F1 is not a DI container, it's designed for LN2. I'm surprised to see a container with a base like the koolance doing so well on DI actually, conventional wisdom for a good DI container is a flat base. I expected to see the Evo doing better quite honestly among these three units. But hey I trust your results so I'm interested in hearing more about the Koolance under load.

That is primarily due to the price point dry ice pots target I believe. Why drill expensive bore holes when you are making a cheap ass dry ice pot? Heat transfer is heat transfer, flat bases have worked moderately well with dry ice due to the acetone well.. making perfect contact with the base giving you typically a couple square inches of contact patch. The koolance is a piece of copper swiss cheese. The temps mentioned for this pot are in line with what I saw with my few experiences under dry ice. Makes sense given the stupid amounts of surface area.

[ChanceCoats123]

You'll have even more fun with the pot under LN2 if you can find a chip without a coldbug. This pot will coldbug CPUs that normally don't coldbug on some other pots out there.



That is primarily due to the price point dry ice pots target I believe. Why drill expensive bore holes when you are making a cheap ass dry ice pot? Heat transfer is heat transfer, flat bases have worked moderately well with dry ice due to the acetone well.. making perfect contact with the base giving you typically a couple square inches of contact patch. The koolance is a piece of copper swiss cheese. The temps mentioned for this pot are in line with what I saw with my few experiences under dry ice. Makes sense given the stupid amounts of surface area.

You're talking of the Koolance correct?

Also, you gotta love those Amd chips. ;)

[Brolloks]

Moved this into the appropriate section :)

[87dtna]

Proud new owner of Gomeler's Koolance V2 :)

[ChanceCoats123]

Proud new owner of Ed's Mini pot!

[xoke]

I just found this post tonight, but I recently picked up the Koolance V2 on my own accord. It is my first DICE pot, so I'm really learning with it...learning how DICE cools, how much you need, how to keep the slurry, etc. I gotta tell you, my first time out, first time on DICE ever the Koolance pot stayed between -77 and -74 under extreme heat all night. It takes about 2 minutes to get cold, only because with DICE and acetone you have to take your time with a room temperature pot, but the darn thing stays cold forever. I'm in the habit of keeping my bench rig running until the pot and CPU are back to room temperature...it takes 30 minutes with a 1090t to get that thing back into the 30c range. I love it. Best $180 I ever spent. (Agreed, the mounting hardware is a joke, go get some threaded stock or long bolts and be done with their silly 2 piece mounting system. :thup:

[Brolloks]

Bump for some of the new guys going dice

[ChanceCoats123]

Not to contradict Brolloks (too badly at least), it is definitely F1EE > Dragon Evo 2.anything on DIce. The F1EE has more surface area for more cold, it has wayyyy more mass to help avoid temperature swings. To be honest the only pot better on DIce (in this thread) would be the V2 since it has sooo much surface area.

[Janus67]

Awesome thread here, tons of great info.

Any recommendations on choosing between the Koolance V1 and the Koolance V2? From what I understand the V2 is a bit better but could have capacitor-hitting issues. Should that make the V1 the safer buy (assuming it can still fit all current sockets (at least through 1366?)

[Seebs]

Awesome thread here, tons of great info.

Any recommendations on choosing between the Koolance V1 and the Koolance V2? From what I understand the V2 is a bit better but could have capacitor-hitting issues. Should that make the V1 the safer buy (assuming it can still fit all current sockets (at least through 1366?)

Your bigger problem will be finding a V1 for sale... Unless you find someone selling a used one. I own a V2 and can't speak highly enough of it; granted, I've only tried it with DICE. As far as the cap hitting issue... The base on the V2 is quite wide, but so far I've ran it on a couple Gigabyte boards for AM3 (890FXA-UD5 and 790FXT-UD5P) and a couple Asus boards for LGA775 (P5Q-E and Rampage Extreme) and it did not give me any clearance issues on either of the boards...

[ChanceCoats123]

The V2 wins in everything. Don't listen to any of the "the V1 is better for Dice and the V2 is better for Ln2" crap. It's all shenanigans, and here's why:

V2 > V1 on Dice:
1. V2 has much more surface area. With Dice, surface area is key to lower temperatures because it allows more heat transfer.
2. V2 has much more mass. While this can make pull down time slower, the pot will swing much less when put under load.

V2 > V1 on Ln2:
1. Again, the surface area comes into play. This time, the V1 is the issue. WAY too little surface area means the pot is unresponsive. The V2 is probably not much better since it will be over responsive though.
2. Mass, also gets to bat twice. More mass = slower pull down, but much steadier temperatures under load.

Now everyone who says the V1 is better than the V2 in any way from here on out will be shunned.

[Bobnova]

It's better in out the door cost :P

[87dtna]

Now everyone who says the V1 is better than the V2 in any way from here on out will be shunned.

LOL, well I've owned both and I'll have to say something you probably can't disagree with.

The V1 is better on DICE *with* dual cores. I noticed nothing better about the V2 if you are just benching dual cores (or single). The V1 is easier to mount. You use 2-3 times as much acetone with the V2 because it's so wide, it's a total waste (again with a dual/single core CPU). Also the V1 has a lot more mass than you think....the V2's base is short and wide, the V1 is not as wide but twice as tall(the copper part before the acetal). But of course, with a quad or more, the V2 is the way to go so temps stay more even.

I have not tested the V1 on LN2, but let me tell you the V2 SUCKS on LN2 with intel chips. Torin3 would confirm with me that it is an insanely fast pot. While benching an E6600 and E6400 at the mini bencing party I was trying to keep temps between -90c and -110c. I would pour a itty bit of LN2 in and temps would get to -106c or so, 20 to 30 seconds later (no exageration) I'd be pouring again because temps were at -95c or warmer. And thats with no load on the CPU. With the F1EE even on the I5 760 I was only pouring ~every 2 minutes.


Also I've noted that the V2 compared to the F1EE, when using it on LN2 for AMD chips, the V2 uses WAY more LN2 to maintain -19x temps as well.

[Janus67]

Nice info there guys. I would be looking to buy a used one either way (doubt my wife will let me spend $200-ish on a part to cool the CPU, but she would let $100-ish fly [as I've seen from past sales here]).