PDA

View Full Version : Chilly1 DI container w/ fx-55 setup



k|ngp|n
11-27-2004, 09:58 AM
Just got some time to test the container, make a mount, and get everything ready to go. 1st boot with neo2 went great. Booted @ -20 and fed the tube dry ice at a slow pace until it it -70c @ load in bios. I think I had a little too much acetone in the container, and it boiled over on the first try and rained smoking, freezing acetone all over my mobo and card while mobo was on! :(. Shut it down, let it all dry, and it was good to go again as I got it to post on second try. I wasnt too worried cuz i was testing with 3500 nc :/.

I booted again 2nd time @ -35c and all was going great and it took about 8 minutes of adding dry ice slowly to get it to -70c in bios with 1.6v on cpu. Then around -70, this neo2 went completely berserk for ten minutes, then wouldnt post. I never made it out of bios after that point.

Pretty sure I can get around this by booting with more cpu speed and volts. Maybe more of a load from the cpu will help the board at those temps?? I had 0 problems with the vapo and this setup, but then again I was only getting -42c or so at the evaporator. Will give it another go this weekend when I have more time.
I think this fx will go 3.5g+ @ -70c. :slobber:

Here are some pics from my adventure:

I tested the crap out of the container and the whole DI process, before I mounted everything up to prevent accidents. LOL, guess I didn't practice enough.
http://pic8.picturetrail.com/VOL256/1324330/5822313/74716910.jpg

It was very easy to mount mobo and cpu to container. I used the mount I made for my cascade and an50r, which basically sandwiches foam between mobo and wooden mount. All I had to do was rotate plate 90 deg for neo2. #6-32 rods and stock bottom motherboard mounting plate work like a charm. The bracket is 1/4 plexi and the springs are from vapo, I used brass thumb screws for tightening.
http://pic8.picturetrail.com/VOL256/1324330/5822313/74716874.jpg
http://pic8.picturetrail.com/VOL256/1324330/5822313/74711668.jpg
http://pic8.picturetrail.com/VOL256/1324330/5822313/74717564.jpg
http://pic8.picturetrail.com/VOL256/1324330/5822313/74716903.jpg

More too come I hope :/. Depends if I can make this board work at -70c
The container is definitely nice though. It held temps for quite awhile after I stopped adding dry ice and there was always about a 3-4c temp difference from bottom of block to the liquid temps. I used 1.5 day old dry ice too so maybe i'll get a few more degrees next time..

RaVeN155
11-27-2004, 12:38 PM
I love your insulation :slobber:

twilius_basic
11-27-2004, 06:51 PM
Good stuff m8 :D I too also like your insulation ;-)

GL with getting the probs sorted and some good clocks from that FX :cool: I like dryice cooling, performance = or > than normal single-phase for a lot less money ;)

Bennah
11-28-2004, 07:55 AM
Very nice setup man. Looking forward to some overclocking and benching screenshots ;)

BTW, where did you get the UEI temp probe from? You know where I could possibly pick a couple up?

gclg2000
11-28-2004, 08:20 AM
Simple question. I'm almost done with a water chiller. I'm using the famous "2 gallon igloo from walmart" for my res. Couldn't one just just start dropping some dry ice into the res??? and just maintain it with DICE. I mean, it will evaporate into thin air, and as long as the liquid (isopropanol) doesnt get to high, it shouldn't "boil over" right? Isn't this a decent way to have soem fun with a completed chiller...just add DICE and viola????

jjcom
11-28-2004, 10:23 AM
Dry ice=frozen air
so what it "melts" its only air. It doesn't cool the water. All you'll get is a ton up bubles...good if you want a nice show, just put some lights on the buttom..:)

jjcom

bh2k
11-28-2004, 11:46 AM
ummm, no, jjcom dry ice is frozen C02.

Paxi
11-28-2004, 12:43 PM
n1 setup
tell us if you are ready and when you know the first results @ overclocking ;)

k|ngp|n
11-28-2004, 02:49 PM
Very nice setup man. Looking forward to some overclocking and benching screenshots ;)

BTW, where did you get the UEI temp probe from? You know where I could possibly pick a couple up?

I got them from chilly. I saw them at a johnstone supply store though here in the U.S.

I might give this neo2 another (3rd) shot on monday when I can get more dry ice. So far, no success on this mobo @ anything colder than -50c or so at the die. On vapo (-41c at die) everything works fine so it has to be neo2 cold bug. I have checked my contact a million times. This board is one of the originals when they first came out, so I might just have to grab a new one.

k|ngp|n
11-29-2004, 03:36 PM
Finally got it to go! I made a new thicker mount to get a better crush on the core and gave it a little bit of a load at startup and that was all it needed :/

I almost managed 3.5ghz benchable(superpi/3d) @-70c(high ghz screenshots are boring and useless so i don't bother) :/

Here are some more pics and screenshots of my best benches from today. This container really is kick ass BTW.... I only lost about 1.5 degrees in one whole 3d mark run and almost nothing in 1 run superpi 1m:

http://pic8.picturetrail.com/VOL256/1324330/2517856/75030302.jpg

Here is one @ midddle of 3do1 with temps taken right from the base of the block:
http://pic8.picturetrail.com/VOL256/1324330/2517856/75030288.jpg

and some nice benches with bh-5(using 183 divider) and x800xt on air and only @ 584/584 for 3d :) :
http://pic8.picturetrail.com/VOL256/1324330/2517856/75028457.jpg
http://pic8.picturetrail.com/VOL256/1324330/2517856/75028447.jpg

Chilly is sending me another container with his spiral block for gpu. It should get the card about -15 to -20c colder than my gpu chiller I hope.

DI so far has definitly beat my vapoLS pretty good.(-42c max evap loaded temps). I got my cpu about 150mhz faster and it was way more stable for like 3-4 straight of benching. I just kept it at 3.4+ for whole time lol.
I can't help but wonder what this chip is going to do @ -100c ?? :slobber:

Nazaar
11-30-2004, 12:33 AM
so your score is 38299 with an xt-pe on air ??


hmm, i'd say, get some DI on that one !!

k|ngp|n
11-30-2004, 05:05 AM
so your score is 38299 with an xt-pe on air ??

hmm, i'd say, get some DI on that one !!

Yep. I want to see how high I can get my system scores before I add the subzero card clocks(584/584) . With this DI, I think 39k may be possible with same card clocks. I have a little bit more headroom left to boost my lobby scores, then thats it. I'm gonna try for 23s superpi 1m next.

charlie
11-30-2004, 01:34 PM
incredible results....
nearly 40k on AIR cooled GPU....

Kingpin,
Few DI questions....
Did you use block or pelleted DI?
I've found my rig just EATS the DI, hardly leaving time to bench!
Acetone right?
PLUS! How long did you have to wait for the base to cool down to -70C?

k|ngp|n
11-30-2004, 04:04 PM
incredible results....
nearly 40k on AIR cooled GPU....


Thanx charlie. I think 40k with gpu on air will happen for sure.
I might go for number 1 spot with chilly1 DI container on gpu too. The acetone is so nasty though and I might need to take a break from it for abit.
This is my first go at it with DI.
I used the DI pellets and acetone. I found a supplier 10 minutes away from my house. So far, what works best is to fill container about 1/4 full, add pellets until -20c then boot, set vcore in bios, and get into os where I keep adding DI until about -65c. It takes around 10 minutes. Once my temps hit -72c or so, it's go time. Start of 3dmark 01' -72c...about 1/2 way thru bench I add a few pellets(enough to get temps going down again) and at end of bench temps are around -70 to -71c.

charlie
11-30-2004, 09:05 PM
Thanx charlie. I think 40k with gpu on air will happen for sure.
I might go for number 1 spot with chilly1 DI container on gpu too. The acetone is so nasty though and I might need to take a break from it for abit.
This is my first go at it with DI.
I used the DI pellets and acetone. I found a supplier 10 minutes away from my house. So far, what works best is to fill container about 1/4 full, add pellets until -20c then boot, set vcore in bios, and get into os where I keep adding DI until about -65c. It takes around 10 minutes. Once my temps hit -72c or so, it's go time. Start of 3dmark 01' -72c...about 1/2 way thru bench I add a few pellets(enough to get temps going down again) and at end of bench temps are around -70 to -71c.

The pellets are REALLY better! Does the acetone vapor give you a headache? Wanna little tip? Set Vcore in BIOS to MAX and then in the desktop, use VID manipulation in Clockgen to bring the Vcore down temporarily. Temps will drop REAL fast then!

C

k|ngp|n
12-01-2004, 05:09 AM
The pellets are REALLY better! Does the acetone vapor give you a headache? Wanna little tip? Set Vcore in BIOS to MAX and then in the desktop, use VID manipulation in Clockgen to bring the Vcore down temporarily. Temps will drop REAL fast then!

C

Great idea! I never even thought of that. This will be much better than having to go back into bios and gradually raising vcore as temps drops. I will do this for sure on my next run.

as far as the acetone vapor, I have taken every precaution with proper ventilation and I still get sick on the fumes.

jjcom
12-01-2004, 02:09 PM
maybe set the rig up out side? That might help out, but depending on where you live, it could cause more problems then good.

jjcom

k|ngp|n
12-12-2004, 07:49 AM
update on DI process:

I can now get the base temps of the container to -72c in less than 20 seconds...Container gets filled about 1/2 way with pellets or broken up DI. Pour in about a shot or less of acetone, then fill container to top with DI and within 20 seconds literally, temps @ load are -around -70c.

Many thanx to jason aka mickeymouse for the DI tips...

twilius_basic
12-12-2004, 08:32 AM
update on DI process:

I can now get the base temps of the container to -72c in less than 20 seconds...Container gets filled about 1/2 way with pellets or broken up DI. Pour in about a shot or less of acetone, then fill container to top with DI and within 20 seconds literally, temps @ load are -around -70c.

Many thanx to jason aka mickeymouse for the DI tips...

That is awesome performance there m8. Really making the most of what you have: loaded temps of -70C are insane! Probably the top-notch design of the container that helps so much. If you don't mind my asking, how much did it cost you? :)

Also, how did you get ahold of Jason? I've been meaning to ask him some DI q's, but he is no longer around here or FM :(

Thanks k|ngp|n :)

Firefoxx
12-12-2004, 12:01 PM
Great results and 38k on aircooled X800, holy :banana::banana::banana::banana:, don't want to know, what this gpu will do on DI 43k sounds possible...

k|ngp|n
12-13-2004, 07:56 PM
That is awesome performance there m8. Really making the most of what you have: loaded temps of -70C are insane! Probably the top-notch design of the container that helps so much. If you don't mind my asking, how much did it cost you? :)

Also, how did you get ahold of Jason? I've been meaning to ask him some DI q's, but he is no longer around here or FM :(

Thanks k|ngp|n :)

hey...thanx :P The DI is really fun and more involved than with cascade(when your benching)....unless you love turning valves. It also helped me learn alot on how this processor scales. With the DI, you can basically set yout load temps to whatever you want.

I have since tried to run DI on the gpu, and my whole card froze up and stopped working...but it's running fine again now. I need to work on my insulation a little bit better. I was thinking about making a "taco" to seal whole card in with 4-6 mil plastic and seal string. I need go grab a 9600 or something to test it on first.

You probably dont see jas post here or anywhere else for that matter because of the drama the ensues. He is active at other forums...look around.

twilius_basic
12-13-2004, 09:17 PM
hey...thanx :P The DI is really fun and more involved than with cascade(when your benching)....unless you love turning valves. It also helped me learn alot on how this processor scales. With the DI, you can basically set yout load temps to whatever you want.

I have since tried to run DI on the gpu, and my whole card froze up and stopped working...but it's running fine again now. I need to work on my insulation a little bit better. I was thinking about making a "taco" to seal whole card in with 4-6 mil plastic and seal string. I need go grab a 9600 or something to test it on first.

You probably dont see jas post here or anywhere else for that matter because of the drama the ensues. He is active at other forums...look around.

Sounds like fun :D I just got my DI tube soldered today, so if I ever get a working A64 rig again, and get up the guts, I'm going to have some fun too :cool:

GL with getting DI working on your GFX :)

Also, where might I look for J? I saw him at akiba-pc.com and vr-zone.com a while ago, but no where recently....

Keep on running m8! :D

charlie
12-15-2004, 09:42 AM
Lol.... Let me tell you about a new trick I've been using! Modular DI tube....

I have several baseplates to fit different boards, A64, Dothan DFI, etc... lately I've been using Elmer's School Glue Gel to attach tube to baseplate.... don't laugh! With DI, it works fine. With LN2 though, it cracks instantly!
This way it's real easy to swap out baseplates...

k|ngp|n
12-17-2004, 12:48 PM
Elmers GLue? LOL
How does that work? Doesnt the acetone seep thru? OR do you not use it with your new process?

bam~bam
12-17-2004, 12:58 PM
very nice oc =D

Hilppa
12-17-2004, 01:18 PM
Lol.... Let me tell you about a new trick I've been using! Modular DI tube....

I have several baseplates to fit different boards, A64, Dothan DFI, etc... lately I've been using Elmer's School Glue Gel to attach tube to baseplate.... don't laugh! With DI, it works fine. With LN2 though, it cracks instantly!
This way it's real easy to swap out baseplates...

LOL .. :D Had to read it twice to make sure I didnt misunderstand :D Elmer's glue.. that is the craziest thing I've heard for a while...

enzoR
12-19-2004, 04:32 AM
LOL i had to read it 3 times. how do u get a good thermal transfer from the tube and the baseplate? Elmers glue... tss :D

Firefoxx
12-19-2004, 07:25 AM
hehe...that sounds funny, lol, but it's a good idea to be able to switch baseplates...

charlie
12-19-2004, 10:25 AM
hehe...that sounds funny, lol, but it's a good idea to be able to switch baseplates...

yeah, alcohol doesn't eat it.... and my baseplate is 5/8" so most of my cooling comes from the baseplate. Of course I'm not saying it's better than brazed, but this allows me to do a lot of experimenting with tubes and plates :D

C

Aldus
12-19-2004, 11:20 AM
Where do you guys get the copper for the baseplates ? Also is there a thickness for the pipe that works the best or will any old copper pipe thats decently (1/8") thick work ?

gclg2000
12-24-2004, 08:31 AM
bump.

We DEMAND another run and pics when you get the GPU cooler.

Awesome stuff man, merry christmas.


Where do you guys get the copper for the baseplates ? Also is there a thickness for the pipe that works the best or will any old copper pipe thats decently (1/8") thick work ?

www.onlinemetals.com

nCrusader
01-04-2005, 10:39 AM
I noticed when crotale and I used Dry Ice (mobile 1400 @ 3754MHz) that the acetone boils very intensivly, quickly boiling over the tube, in the begining before the temps had tropped. What steps do you do when you start? We first poured a few centimeters of acetone and then added just a very few pellets and later on added more. Is there any way to get around the intense boiling?

When using dryice on the graphiccard, is there any differences compared to cooling the cpu?

sorry for the bad english

charlie
01-05-2005, 10:00 PM
I noticed when crotale and I used Dry Ice (mobile 1400 @ 3754MHz) that the acetone boils very intensivly, quickly boiling over the tube, in the begining before the temps had tropped. What steps do you do when you start? We first poured a few centimeters of acetone and then added just a very few pellets and later on added more. Is there any way to get around the intense boiling?

When using dryice on the graphiccard, is there any differences compared to cooling the cpu?

sorry for the bad english


No, no, no!!!
You've got it backwards! Add the dry ice pellets and THEN pour in a LITTLE fluid....maybe 30-40ml.

Jort
01-06-2005, 11:47 AM
is there now solution to mill a circel in the baseplate and let the container fit into the baseplate?

crotale
01-07-2005, 07:21 AM
No, no, no!!!
You've got it backwards! Add the dry ice pellets and THEN pour in a LITTLE fluid....maybe 30-40ml.

Any special reason for doing it that way?

How much pellets do you have in the container to begin with?

k|ngp|n
01-07-2005, 07:59 AM
Any special reason for doing it that way?

How much pellets do you have in the container to begin with?

IT freezes the acetone as it pours through the DI and minimizes it bubbling up like when you fill container with fluid first and start dropping DI into fluid and it start bubbling like mad initially.

Fill the container about 3/4 way with DI, then pour in your fluid, and then fill to top with DI while your benching. You will see the DI "sinking" in the container, just keep it full to top during session and you'll be golden :P

This is what worked best for me and was great advice from someone who definitely knows his sh**. :)

G H Z
05-02-2005, 02:47 PM
100% Ethel Alchol freezes @ -90, Acetone @ -94

I wonder if the fumes would be less disturbing.