Q6600 on Dry Ice -- FSB: 480, CPU: 4.32GHz -- Updated with pics.

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
Here's the deal. I won't give all the specifics except my goal is to overclock a G0 Q6600.


Currently I can't justify the means so I need your help to guarantee a 4.3GHz-4.5GHz overclock.

So in reality, the biggest concern is how much is this G0 Q6600 going to hold me back? My dream would be to make this a reality with a Q9650, but we all know that's a ridiculous price to pay (after reviewing my pay check and bank accounts). So let's not talk about the Penryn :>

I have actually dug up a lot of information and for those who are interested in what I'm attempting...here's a decent forum thread from another site I found...

http://forums.legitreviews.com/about6289.html



FINAL OVERCLOCK:

FSB-480
CPU-4320MHz, 1.74V



finaloverclockyp6.jpg



dsc00347cv5.jpg



dsc00352ny8.jpg



dsc00356ew9.jpg



dsc00360cj8.jpg



dsc00361ze5.jpg



dsc00365ru2.jpg



dsc00362dp5.jpg



dsc00367mg1.jpg



dsc00366wo0.jpg



dsc00364ke7.jpg



dsc00374vh1.jpg



dsc00375qb0.jpg



dsc00376tt7.jpg



dsc00373st6.jpg
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
Process:

1) Buy the crap...gigabyte DQ6? I'm leaning towards the P35. I like DDR2 and the performance difference vs. DDR3 seems marginal at best.

--I purchased a...

Q6600; L734B275 w/ VID of 1.275
Gigabyte X38-DS4
Crucial Ballistix PC2-8500 @ 2.2V

2) Build the heatsink that holds the Acetone and Dry Ice (refer to link in original post).

--Type L Hardened Copper @ 12" w/copper cap. Found it at local plumbing supply and Brazed by local HVAC company. I placed a 4"x4" conduit cover plate on the top of the copper pipe and secured it to a bottom plate using wing nuts and 8-32 threaded rod cut to about 13".

3) Use a chemical to seal the front and back of the mobo. but what to use?

--Used Duct seal putty~$2. Any gel would probably do fine but I was a little paranoid as it was my first try at DIY extreme cooling. The putty was placed as shown in above pics...around the CPU socket and surrounding areas. **NOTE** It doesn't remove very well from the board ;)

4) Apply neoprene cut outs to socket and surrounding areas as needed.

--Used one neoprene square on the underside of the mobo as shown in the photos above. I utilized a back plate from a Thermalright Ultima 90 to secure it.

5) Bring it all together :)

6) Post pictures, results, etc.


Feel free to PM me or for amendments as I'm going to see about creating a concise but detailed report as to my experience IF I choose to go through with this.
 

Evilonigiri

Splendid
Jun 8, 2007
4,381
0
22,780
Are you trying to set a record or something? Or you're doing this for fun?

Whatever it may be, you may need to go through more than one cpu, since they all Oc differently.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
Well...I guess you could say both.

It started out with noticing the gigabyte competition. Suddenly my dreams of peltiers were flashing in front of me. Then I realized that the TDP of the Q6600 was too high for the 226W Peltier I found.

Doing the math shows that the Peltier would become far too inefficient and that water cooling without the Peltier would probably turn out better.

Then I realized that at my local wally world we had dry ice. So I began thumbing through the internetz for ideas as to how to assemble the heatsink.

Now I'm at the conclusion you brought up. This could turn out to be a bit pricey as I might have to go through several CPU's. The only CPU that seems to qualify for this request would be the Q9650. However, that's not really an option if we compare it with my bank accounts n' such. :>

So bottom line. I want to blow the hell out of a Q6600 with a little help from some voltage and dry ice...but not unless I can reach some mid-level 4+GHz overclocks. But after reviewing information for the Q6600 it seems that I might be running into a 'switching' wall. I'm not so sure the 65nm Quad core process is binning well enough to give me a guaranteed 4+ GHz using Dry ice.
 

Raviolissimo

Distinguished
Apr 29, 2006
357
0
18,780


things i've used at work to keep electronic parts dry -
* epoxy soldermask
* conformal coat, silicone & non-silicone
* potting. another word for encapsulation

if you can keep condensation off the motherboard
using neoprene & other sealing materials (another
example - a connector made by a company named
Airborne that i've used at work had a rubber gasket
that went between the male & female parts. kept
water out real good. very important at 12,000 volts ! :pt1cable:



to keep the GF happy, minimizing the appearance of
expenditures is important. or maybe spring on a nice
piece of jewelry, so that when you buy a second G0
and a second motherboard, and maybe then some
different RAM, she won't be so upset. once my GF
got real upset because i paid $600 for an o'scope
on the spur of the moment.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780



That would be stellar.

However...I was looking at those and they go for about $900 or $600 USD used. That's not something I could justify. It seems like a lot of kids are able to hit about 3.8 with their Q6600 on air coolers though. I might just take the plunge.

Maybe I'll look around these forums and see if I can find a DD 4401 kit. Then I'll see if I can find a higher rated peltier since they're cheap. The 226W was about 38 USD. I have a 750W PCP&C that isn't doing much with my current configuration so maybe I should stretch its legs.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
HOLY CRAP!!!...I got some ideas for sealing the motherboard against condensation.

I could use hair spray, egg whites w/sugar (to make it like a hard candy so it's easier to clean up)....or *drum roll*...submerge it in cooking oil?

What do you guys think? I know they're cheap but looking at Ravio's suggestion these things came to mind.

I was hoping I could find a some type of spray on lacker coating or something rather, but the cooking oil sounds easiest and full proof without having to worry about covering up sockets or slots.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
The more I think about it...the more I believe I'm getting carried away.

I bet insulating the underside of the motherboard with neoprene surrounding the socket should suffice the need. I'm also going to throw in some non-conductive gel inside the actual socket.

I think this project is gonna be a go.
 

Antinomy

Distinguished
Jan 7, 2008
4
0
18,510
As a MoBo of choice I would suggest DFI LP P35-T2R, Asus Maximus Formula, Foxconn MARS.
About the preserve from condensation there are a lot of threads at vr-zone.com and xtremesystems.org ;)
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
I was kinda hoping for something cool and inspirational such as...

I can't think of any good one liners so you fill in the blank. I found a website I'm purchasing some of that from.

I had planned to use my thermalright backplate filled with that Gel and add some into the socket for good measure. I was also going to apply it around the exposed base of the fab-O-lous d'ice I was putting together. I'll give screen shots as I prep it for some silver soldering. I got it for about 22 bucks at a plumbing store, but I couldn't find type K...only M-copper.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
alright...all done with the project kids. I'll update the info as it's convenient for me. I'm a little disappointed.

Final Overclock:

FSB-484
CPU-4.32GHz
Voltage-1.74
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
Alright...I'll see what I can do to post pictures for everybody...I'll try to have it up shortly. I don't know if anybody is interested or not but I figured I could at least share my experience.
 

cnumartyr

Distinguished
Nov 3, 2007
2,287
0
19,780



I take it it went ok though. No damage.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780


nah...no damage. I was pissed though.

I'm confused about my temperatures. If my temps in windows are correct, between 2C and -8C, then it's a good overclock. If they're wrong...then I suck :)

Because temps should have been around -30 or -40. If that were the case I should be able to hit 4.5GHz.

Sorry about the pictures...I'm working on those now.

http://service.futuremark.com/compare?3dm06=4675104

Obviously my 3DMark isn't patched correctly as the score isn't representative of the calculated numbers.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
I believe where I made my error(huge mistake) was not realizing how quickly the acetone would dissipate...so I added way too much ;)

You can see that in one of the photos as bubbling acetone nearly reaches the top of the cylinder due to the gas attempting to escape from the melting dry ice.
 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780
I'm wondering if my batch didn't bin so well as I needed so much voltage for such an awful overclock. That's why I'm leary about my true temperatures. I might try it again later in the week with a lot less acetone this time. We'll see if I can get my hands on some better equipment too. It'd be nice to have a multimeter and something to read temps.

You also have to throw some error into this project based on the d'ice pot that I made.

Either way...my assumption is that with 1.6V and proper cooling I shouldn't have had such a hard time hitting 4.5GHz. Therefore, my temps under XP were most likely correct.

I could get up to 484FSB and that was my wall. I don't know if it was CPU or board bound but I definitely could not hit 485 without problems. I even tried frying my mobo and CPU to get it higher.

 

I800C0LLECT

Distinguished
May 16, 2007
262
0
18,780



Certainly...the CPU score looks about right, I didn't care about overclocking the GPU.