PDA

View Full Version : Sally gets Cold



1Day
08-24-2007, 10:02 AM
http://img180.imageshack.us/img180/9089/boshuterqy0.jpg

When I saw this post I had an very strong sense of déjà vu - the compelling need to try the extreme world of sub-zero benching that overwhelmed me and for the first (and so far only) time I tried sub-zero benching I recognized. I thought that since I am very new to this forum I might share that first time with you all.

This is my story of that Dry Ice roller coaster ride.

Sally gets really cold

Where I live (South Africa) there are not a whole heap of people who are particularly interested in either overclocking hardware or benchmarking hardware. I guess it is mainly because the technology is just so expensive here. But there are a few - and the one thing that seems universal among geeks who freeze computer equipment is the generosity and willingness to share what they have.

A fellow bencher was kind enough to lend me one of his benching pots. It might not look like much when compared to the works of art that are available in the US and Europe but it did the job and surely demonstrates that looks are not that important when it comes to benching.


http://img19.imageshack.us/img19/1804/fig2dq1.jpg

First thing I did was read everything I could find about keeping my beloved Sally in one piece. Then I went shopping. And let me tell you I love shopping. On the list were nail varnish, acetone and absorbent kitchen cloths all easily found at the local shopping mall. But more difficult for me was finding the insulation materials, in the end I was successful in getting the goodies I was looking for thanks to a local contractor who drove especially (for me) to the supplier to fetch the stuff.


http://img507.imageshack.us/img507/3199/fig3ls4.jpg

Then began the singular most painful task I have had to do – removing the Tuniq Tower 120 air cooler from Sally.


http://img256.imageshack.us/img256/9216/fig4wa0.jpg

This is Sally without the TT120 and after a bit of a clean. For those who have not used LiquidPro thermal solution for fear of staining the HIS of their CPU, do not worry it comes off quite easily and has never stained any of the processors I have used it on. And this is the prototype test bench (ver 2) that was put together for this Dry Ice session.


http://img265.imageshack.us/img265/1644/fig5ol2.jpg

Prepping the MOBO
I began by painting on a few layers of clear nail varnish onto the motherboard around the CPU area, both front and back. I have outlined the area I coated.


http://img412.imageshack.us/img412/8171/fig6mj6.jpg

A good tip for any new painters of motherboards is to cover the CPU socket with something to protect it. I cut a business card in half and that did the trick.
Then taking a square of the insulation that I had bought I pressed it firmly against the socket and the other bits (not so hard guys – you just want an impression) and proceeded to cut out the outlines. Not very well or neatly done I might add. Next time I will buy a scalpel from the local chemist and not use my vegetable knife from the kitchen.

http://img220.imageshack.us/img220/7780/fig7yg3.jpg

Then taking an absorbent drying mat which I found in the local supermarket I cut out a duplicate gasket. This was to be the second layer – and boy was I lucky I did.


http://img265.imageshack.us/img265/7995/fig8ti3.jpg

An additional gasket was made from the black insulation; this one has a circle with sloped walls to provide a final seal between the dough nut of the pipe and the insulation protecting the motherboard and CPU from condensation. This is what it all looks like after being mounted.


http://img220.imageshack.us/img220/3527/fig9ul3.jpg

(Now before someone says “but why not use Vaseline - or some other kind of dielectric grease”; let me just say that I am aware of the practice and I am sure that it works just fine, but it is not for me.)


Prepping the PIPE
Ralph was in a huge rush when he brought the pipe up to Pretoria for me. So as a result there were a few things that were needed to complete the pipe for use. The first thing I did was lap the block. Now I do not subscribe to a mirror finish as I feel that is a waste and offers no benefit in temperature performance at all. So I use for my final polish a 600 grit water paper. The old dough nut was a little sad and needed to be removed and buried which I promptly did. A nice new 25mm diameter wall thickness dough nut was cut and slipped onto the pipe bottom.


http://img256.imageshack.us/img256/7756/fig10bs3.jpg

Fasteners were acquired and tested with a bit of wood as the backing plate as you can see. Speaking of backing plate I found the perfect short term solution at the local electrical wholesaler (looking for a temp gauge – no luck sadly). A PVC blank socket cover did the trick after four holes were drilled and counter sunk for the M4 screw heads. Here another gasket was made and fitted to protect the underside of the motherboard from condensation too.

Dry Ice
Dry Ice International provided at a reasonable cost the 3mm rice pellets dry ice that I used. The product comes in 10KG packs and fits in a normal family sized cooler box. I bought a Styrofoam box from them as I shamefully do not have a cooler box at all.


http://img507.imageshack.us/img507/8976/fig11pi1.jpg

The only evil thing about the acquiring dry ice process was that the collection point happened to be an ice cream shop. Temptation was just too strong – will have to spend a few extra hours on the exercise bike in the days to come – the most yummy double caramel ice cream I have had in ages was included in the purchase as well as 20KG of dry ice pellets. Acetone (as pure a grade as I was able to find) was bought from Clicks. I did not want to use my nail polish remover that I have even if it is acetone based.

All ready to go…NOT
A small dollop of thermal paste (ceramique) was added to the block of the copper pipe and mounted. Here it is all ready for my first ever attempt at Dry Ice. Just have to put the mother board onto the tray and I would be ready to go. Well yes except that the backing plate and gasket was too thick and did not fit. Ten minutes later and a bit of angle grinder usage the motherboard tray was modified to fit. Two minutes later acetone added to the pot and dry ice followed. Switch on and ….beeperpeep. Nothing, Sally would not boot to bios.


One hour later – I seat the graphic card correctly and all is good. Oops. :o


http://img220.imageshack.us/img220/4182/fig1yg4.jpg

Sally is finally cold.

Here are my temps after the first boot up.

http://img19.imageshack.us/img19/597/tempswt5.jpg

But like so many things in life the first time is always fraught and filled with “complications” this first attempt on Saturday night was no different.


I had no idea what I was doing to be honest – all the reading I had done and all the advice that people had given me did not include some basic stuff. For example this is the bit I was told: the acetone that you are using to aid in the cooling process is added first about two or three cap-fulls is to be added (15ml) – now the bit that was not mentioned – those three cap-fulls would last two to three hours of benching. So instead I got paranoid and kept adding acetone till my first 100ml bottle was finished within the first hour. :o :lol:

Here are some of the results I got that first night.

http://img75.imageshack.us/img75/2605/3d200172976pr7.jpg

http://img518.imageshack.us/img518/4510/3dmark0305gu8.jpg

http://img262.imageshack.us/img262/2627/3dmark03a05hy5.jpg

http://img265.imageshack.us/img265/9601/3dmark0613243tx8.jpg

I did not think to take screen shots of the later runs as I was a bit carried away with my main aim - setting a personal best at Super PI 1mil - and using the Futuremark benches as stability tests.

Then the second tip I was given that I did not understand correctly. Was this - If you need to unseat your pot in a hurry just pour in some boiling water.

Now here is the scene I have very tall and thin pot, I have no misty vapor all over the show like I am expecting. My temp gauge on Speed-fan says 5’C and Guardian has all its temps in red. My first reaction is that Sally is in mortal danger and I have to save her. Boiling water is poured into the top of the pipe but a small tentative amount. A bit of vapor spews out and that is about it, I so I pour more in, and again nothing happened now I pour quite a lot more in. This time something does happen. The ice blockage that I guess had formed was finally melted and a lot of boiling hot water hit the dry ice. Geyser time was the result. Water everywhere, all over my desk. :o :yikes:


This was when all my preparation beforehand paid dividends. After I had stopped swearing I mopped up all the water on the desk and took the pot off. I dried the motherboard as best I could. And with some trepidation I inspected the gaskets I had made for the pot. The top one was wet around the edges and middle (absorbent fabric) gasket was decidedly damp around where it touched the pot tube – but not the bottom one. There was not a drop of water on or near the CPU area. I gave up benching for the night and put an industrial strength blower on full and aimed the gale force wind at the motherboard. :thumbs:


Sunday morning I had my second opportunity to do some over-clocking and benching. This time there was no dry ice and water shower.


http://i19.tinypic.com/5y9g3up.jpg

http://i18.tinypic.com/66w6ofn.jpg

And that’s all I was able to do as my two helpers needed their mom to pay attention to them. But I was able to set a personal best in super PI so I was quite content to stop and play with my two daughters.


http://i206.photobucket.com/albums/bb227/1Day2007/DSC04854.jpg

Here is the link (http://valid.x86-secret.com/show_oc.php?id=215473) to my highest super PI 1 mil stable over clock.

Dry Ice benching was fun but I still do not have the hang of it completely. I certainly had fun and next time I will focus more on 3DMark benching as I do not think that there is much head room left in my E6420 for a higher clock. And drivers for my HD2900XT should be more mature by then.:beer:

Tauvo
08-24-2007, 10:35 AM
Sounds like you had a great time and your patience in preparation paid off with the Super PI times you wanted! Great job! :rock: :rock: :rock:

Just Learnin'
08-24-2007, 10:36 AM
:thumbs: :beer: Thanks for sharing, brave or crazy? :lol:

harleybro
08-24-2007, 11:06 AM
Nice write up thanks alot for sharing this with us. Looks like you had fun for your first go around. :thumb:

NINaudio
08-25-2007, 06:51 AM
Very cool 1Day!

nutcase
08-25-2007, 08:10 AM
Looks good 1Day :thumb:

just words of info here.

when doing dry Ice, you only need a small amount of acetone. I normally use a shot glass worth of acetone. since it is at such a low temp, the acetone will stop evaporating and you should not need to add anymore.

Luckily you had a long tube. with a short tube like mine, you could get a volcano as I did when I first started using dry ice :o

1Day
08-25-2007, 08:43 AM
Nutcase sad to say I had the volcano but only after I added a tumbler of boiling hot water to the tube. And a shot glass has different measures in different countries. So what is that in fluid ounces or milliliters?

nutcase
08-25-2007, 08:50 AM
Nutcase sad to say I had the volcano but only after I added a tumbler of boiling hot water to the tube. And a shot glass has different measures in different countries. So what is that in fluid ounces or milliliters?

Most shot glasses are 2 oz glasses that I use. usually just enough acetone is needed to have a small pudding at the bottom of the tube.

and yep, water and dry ice makes for a Violent reaction :yikes:

harleybro
08-25-2007, 08:58 AM
Most shot glasses are 2 oz glasses that I use. usually just enough acetone is needed to have a small pudding at the bottom of the tube.

and yep, water and dry ice makes for a Violent reaction :yikes:

You should see water an LN2. When I used to break down my rig each time after benching the tube was heated with warm water. :mischeif: :rotflmao: As for the acetone I also would use just enough to give a covering on the bottom of the tube. Maybe 1/4" deep or so. :yup:

nutcase
08-25-2007, 09:01 AM
I forgot another thing.

do not fall asleep when dry ice benching. I did that once and when I woke up an hour later, I had fried an FX-57 :hide:

harleybro
08-25-2007, 09:46 AM
I forgot another thing.

do not fall asleep when dry ice benching. I did that once and when I woke up an hour later, I had fried an FX-57 :hide:

:yikes: :rotflmao: :rotflmao: ok nutcase ya got me on that I haven't done that one before. I might have been close when I had 3 gallons of denatured alochol mixed with 2 gals. of water and throwing dice in it for some extreme liquid cooling.

boshuter
08-25-2007, 05:50 PM
Nice write up 1day... welcome to BleedinEdge cooling .. :thumbs:






Luckily you had a long tube. with a short tube like mine, you could get a volcano as I did when I first started using dry ice :o

Hehehe... I think everyone has done that, my first time I had a little too much acetone and when I dumped a bunch of powdered dice in it started spewing all over the place :yikes: ..... I didn't know what was happening, but I was pretty sure I didn't want that stuff all over my board :chit: I just clamped my hand over the top of the tube until it calmed down...... then after I calmed down I added more ice (slowly) and went on benching ;)

I found that if you fill the pot about half full with dice, then slowly trickle the acetone in that it helps the spewing.