Fx 5900 U full w/c (pushing man & machine to their limits) - Page 6

Augustus2 · 01-25-2004, 06:09 PM

OK, Bladerunner: you win.

Let's hope it doesn't limit the cooling capacity for the core.

With 6 mm ID tube you have (only) 28.26 mm². You want the main flow over the videochip, say 18 mm². With those two channels each 2 mm wide, they should be 4.5 mm deep. (How deep are they?)
You can also try limiting the flow in the back channel with an inset(? plug?) inside the channel, to adjust the width to (in this situation) 10/3.2=3.2 mm, forcing the main flow to the front channel.

You almost wish you had a (very expensive) flow simulation program available in these situations. (In this Dutch topic http://gathering.tweakers.net/forum/...sages/478776/1 you'll find a lot of beautiful flow simulation pictures).

BladeRunner · 01-25-2004, 06:58 PM

No you do make a good point, and that area of the block design is technically a poor show on my part ....... hang on I'll smack myself about a bit

The real truth I suppose is I just lost patience with all the milling and snapping milling bits at the time.

The other thing is that cooler is designed for my set-up so will have sub ambient coolant, (12C to 14C), so assuming the flow is reasonable it should be fine. As you say it might be a good idea to try to percentage the flow more in favour of the frontside block. The main thing in making it was simply to test a concept, and I'd be the first to admit needs development. It would certainly be much easier to make on a proper CNC machine. My goal in all this is to come up with the best single unit block to cool the complete card, with one inflow, one outflow, while being a single slot solution.

The only thing I'm still finding hard to solve is the weight factor using all copper designs, but this is always going to be hard. I still rather avoid plastics if possible for the usual reasons.

*EDIT* Channel depth is > <4.2mm, best I can check now......... did have it written down somewhere :shrug:

acuba · 01-27-2004, 02:38 PM

I can only say... GOD DAMN, that´s nice

guandi · 01-27-2004, 06:09 PM

Wow BladeRunner, you own

I remember emailing you a while ago about custom Gf FX 5950 blocks, (the offer is still open btw, name your price

) but i think i will give it a go myself now, as i have found a place just round the corner from me that does CNC machining and my dad is freindly with the owner, so if i get too stuck with the old mill (circa 1960!!!) at work i know where to go

pierreg60 · 09-15-2005, 03:38 AM

Hello all!!
very good job Blade for your 5900

which are the thickness of copper used?
I think 2 and 6 mm? :shrug:

killernoodle · 09-15-2005, 02:05 PM

The last post of this thread is almost 2 years old....

I dont think Bladerunner posts anymore.

Brians256 · 09-15-2005, 08:26 PM

Last I heard, he was involved in some startup business, but I never heard from him after the setup phases. Bladerunner was one of my early heroes in this hobby. I hope he's doing OK.

Nugit · 09-16-2005, 01:01 AM

Well zfz got a new forum during the summer, that might be the place to catch him if any is: http://zfzforums.com/
He has pretty recent posts over there.

pierreg60 · 09-20-2005, 10:11 AM

ok thanks for the answer

01-25-2004, 06:58 PM	#127
BladeRunner Cooling Savant Join Date: Jul 2001 Location: Chesterfield Uk Posts: 459	No you do make a good point, and that area of the block design is technically a poor show on my part ....... hang on I'll smack myself about a bit The real truth I suppose is I just lost patience with all the milling and snapping milling bits at the time. The other thing is that cooler is designed for my set-up so will have sub ambient coolant, (12C to 14C), so assuming the flow is reasonable it should be fine. As you say it might be a good idea to try to percentage the flow more in favour of the frontside block. The main thing in making it was simply to test a concept, and I'd be the first to admit needs development. It would certainly be much easier to make on a proper CNC machine. My goal in all this is to come up with the best single unit block to cool the complete card, with one inflow, one outflow, while being a single slot solution. The only thing I'm still finding hard to solve is the weight factor using all copper designs, but this is always going to be hard. I still rather avoid plastics if possible for the usual reasons. *EDIT* Channel depth is > <4.2mm, best I can check now......... did have it written down somewhere :shrug: __________________ Zero Fan Zone Last edited by BladeRunner; 01-26-2004 at 10:50 AM.

01-27-2004, 06:09 PM	#129
guandi Cooling Savant Join Date: Dec 2003 Location: UK Posts: 174	Wow BladeRunner, you own I remember emailing you a while ago about custom Gf FX 5950 blocks, (the offer is still open btw, name your price ) but i think i will give it a go myself now, as i have found a place just round the corner from me that does CNC machining and my dad is freindly with the owner, so if i get too stuck with the old mill (circa 1960!!!) at work i know where to go __________________ A8N-SLI Deluxe FX55@3ghz 2x TwinX1024-3200XL 2x BFG6800Ultra Audigy2 ZS + Gigaworks S750 2x120gb/4x80gb lots of water/copper/pumpy goodness.

09-15-2005, 03:38 AM	#130
pierreg60 Cooling Neophyte Join Date: Aug 2004 Location: france Posts: 17	Hello all!! very good job Blade for your 5900 which are the thickness of copper used? I think 2 and 6 mm? :shrug: __________________ Sorry for my english ........ i am frenchy

09-15-2005, 02:05 PM	#131
killernoodle Thermophile Join Date: Sep 2002 Location: Charlotte, NC Posts: 1,014	The last post of this thread is almost 2 years old.... I dont think Bladerunner posts anymore. __________________ I have a nice computer.

09-20-2005, 10:11 AM	#134
pierreg60 Cooling Neophyte Join Date: Aug 2004 Location: france Posts: 17	ok thanks for the answer __________________ Sorry for my english ........ i am frenchy

01-25-2004, 06:09 PM	#126
Augustus2 Cooling Neophyte Join Date: Jan 2004 Location: Netherlands Posts: 2	OK, Bladerunner: you win. Let's hope it doesn't limit the cooling capacity for the core. With 6 mm ID tube you have (only) 28.26 mm². You want the main flow over the videochip, say 18 mm². With those two channels each 2 mm wide, they should be 4.5 mm deep. (How deep are they?) You can also try limiting the flow in the back channel with an inset(? plug?) inside the channel, to adjust the width to (in this situation) 10/3.2=3.2 mm, forcing the main flow to the front channel. You almost wish you had a (very expensive) flow simulation program available in these situations. (In this Dutch topic http://gathering.tweakers.net/forum/...sages/478776/1 you'll find a lot of beautiful flow simulation pictures).

01-27-2004, 02:38 PM	#128
acuba Cooling Neophyte Join Date: Sep 2003 Location: Sweden Posts: 8	I can only say... GOD DAMN, that´s nice

09-15-2005, 08:26 PM	#132
Brians256 Pro/Staff Join Date: Oct 2001 Location: Klamath Falls, OR Posts: 1,439	Last I heard, he was involved in some startup business, but I never heard from him after the setup phases. Bladerunner was one of my early heroes in this hobby. I hope he's doing OK.

09-16-2005, 01:01 AM	#133
Nugit Cooling Neophyte Join Date: Jun 2003 Location: Aalborg, Denmark Posts: 85	Well zfz got a new forum during the summer, that might be the place to catch him if any is: http://zfzforums.com/ He has pretty recent posts over there.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)