Advise about distribution of jets - Page 3

Long Haired Git · 03-04-2005, 05:06 AM

Continuing my hijack of this thread - a couple of worked formulae for those looking to make a jet block:

Head = ( 1.434.Flow^2 ) / ( count^2.radius^4 )
where
Head is meters of water, Flow is in Litres per minute, count is the number of jets (so whole number), and radius is in millimeters.

Quick example of a single, 5mm diameter jet @ 5 LPM we get:

Head = ( 1.434 * 5 * 5 ) / ( 1 * 1 * 2.5 * 2.5 * 2.5 * 2.5 )
Head = 35.85 / 39.0625
Head = 0.918 meters.

Now this does NOT take into consideration the friction in the jets or the cups or the barbs etc, and is just the waterblock resistance. But its a start.

Oh, for the anal, 1.434 is:
(1250 * density of water in Kg/m3 * conversion factor from Newton/m2 to meters H20) / ( 9 * Pi^2 ).

Next, velocity of the jet is:
v = (100 * Flow ) / ( 6 * Pi * count * radius^2 )
which is
v = ( 5.305 * Flow ) / ( count * radius^2 )

So, from the example,
v = ( 5.305 * 5 ) / ( 1 * 2.5 * 2.5 )
v = 26.525 / 6.25
v = 4.25 meters per second

Les · 03-04-2005, 06:38 AM

Many (inc Cathar,I think) short cut P/Q/Velocity/Reynolds calculations by using Pressure Drop 5.0.It also contains useful information on fluid properties vs Temperature (Viscosity.Density)
Find have to take as many short cuts as possible on the road from flow to die temperature.
My preferred route for Jets is :-
Flowrate --PD Drop---> velocity --Flomerics--> h(conv) --Kryotherm--> h(eff)--Waterloo--> "Die Temp"(isothermal coolant).

Unfortunately Kryotherm only does channels, so for "Cups" you have to look elsewhere(dunno where) for conversion of h(convection) to h(effective).

Hoot · 03-06-2005, 02:33 AM

After many starts and restarts, I finally got the telescopic step down jets done. Too tired to reassemble tonight. Here's a look at the 31 Jets:

Way too much work for what looks to me like not enough volume, but I'll be happy to be wrong tomorrow when I get time to test it. Oh yeah, i plugged the excess holes with homebrew SS set screws.

Hoot

Long Haired Git · 03-06-2005, 07:29 PM

Interested to find out, but I will continue to caveat my advice with the need to know what your blockless flow rate is and what the pump you're using is.
Without these two items, we don't know what P-D you can cope with and what flow rate to aim for.

Hoot · 03-06-2005, 10:54 PM

As my signature says, I'm using an Eheim 1250 with .13m head. As for actual flow rate, it's sure a heck of a lot slower with this latest revision block in there. I put a piece of my daughters glitter in the water and watch how fast it goes around the loop of my three blocks, radiator and pump. It used to go around in a second or less with my Whitewater clone. Now it takes almost two seconds. I don't have access to a flow meter and really don't want to drag the box upstairs to the sink to see how long it takes to fill a 2 liter bottle, but I could. It's tough when your development rig is your main rig. The new revision block seems to be doing well, but it still only bests my WhiteWater clone by 1C at load at best. That's on an A64 running approximately 124W, not a Tbred running 70W. Once these A64s have been out long enough, I would expect to see more test results, whether scientific or anecdotal, to emerge from owners of mainstream commercial blocks. Heck, I'd love to see how a real Storm G5 delivers running in a rig similar to mine. I emailed Stew over a week ago inquiring when the preorder for the next batch will begin, using the email address at the web site, but I have yet to get a response. So far, the three blocks I have with track records back to Socket A and the new one all fall somewhere in the .16-.17c/w range on this 90nm core at approximately 124W. It's frustrating to not be able to see any rave deviation outside that range with the three or four variations I've tried with this new block. I keep ending up within the margin of error for the MAX6657 reader as compared to the MCW6002-64, WhiteWater clone and my old faithful PinFin block. My rig needs an enema...

Edit: I'm using this Wattage Calculator to arrive at 124W.

Hoot

Cathar · 03-07-2005, 01:01 AM

Quote:

Originally Posted by Hoot

I emailed Stew over a week ago inquiring when the preorder for the next batch will begin, using the email address at the web site, but I have yet to get a response.

You're not alone. What with real job (tm), and having the flu, I haven't responded to many people yet. Am slowly working through the lot.

Hoot · 03-07-2005, 07:32 AM

Sorry to hear that Stew. Our family has been working through some creeping-crud as well. Luckily, I've dodged the bullet so far. Both the wife and daughter are down with it as I write this. I'm just waiting for the other shoe to drop, as it's rare that when one of us gets sick that the other two don't. Hope you're feeling better soon.

Hoot

Hoot · 03-09-2005, 07:57 AM

Well, the AS5 has had time to settle and still no joy. At load I'm seeing 22C water yielding 43-44C die temperature, which is still in the same range as the other three blocks. I'm starting to suspect that using an ambient system, it's going to be hard to break through that 22C differential barrier. I'm burned out on this block. About the only thing left to try, at a lot of effort, is to pull the NB and GPU blocks out of the loop, running the 1250 to the CPU block only. My gut feeling is that this is not a flow issue. I sincerely feel that I'm down to the source impedance of the high power-density core and the series resistance of the TIM being the limiting factors. I can't stress enough that "this is not your fathers Tbred" we're talking about.

Hoot

Long Haired Git · 03-09-2005, 03:06 PM

Ambient + 22 degrees is what my passive rig ran at too....
It'll be interesting to see what the Storm does for you.
If you do pull that rig about, please do that flow test I was talking about...

bigben2k · 03-09-2005, 04:49 PM

Quote:

Originally Posted by Hoot

... I'm burned out on this block....

I purposely stayed out of this thread, because there's so many more things to try, I figured you'd get tired with it before too long.

This particular design is very sensitive to a number of things. At your stage, you'd want to try different spacings between "cups", within the baseplate, assuming that you've already tried various depths for the cups. Yeah, it means starting over, again and again...

Otherwise I'd point out that your jet tubes are too thick.

Nifty idea with the screws though...

Hoot · 03-09-2005, 09:22 PM

Quote:

Originally Posted by bigben2k

I purposely stayed out of this thread, because there's so many more things to try, I figured you'd get tired with it before too long.

This particular design is very sensitive to a number of things. At your stage, you'd want to try different spacings between "cups", within the baseplate, assuming that you've already tried various depths for the cups. Yeah, it means starting over, again and again...

Otherwise I'd point out that your jet tubes are too thick.

Nifty idea with the screws though...

The cup ID is 3mm following Stews philosophy of lots of return path room. They are spaced 3.5mm on center. I tried 2mm ID cups on 2.5mm centers with the same result. I've tried different insertion depths for the pipes, but kept the cup depth as needed to provide 1mm of copper between the bottom of the cup and the face of the plate.

I will be ordering a G5 when the next batch is available. I'll sure feel silly if it also yields 22C differential as I expect it to exceed MCW6002-64 performance for 2.5x the price.

Hoot

bigben2k · 03-09-2005, 11:47 PM

The G5 will perform best.

You can shave that bottom lower than 1 mm; Cathar brought it down to ~0.1 mm (!). Square ended endmill, in a drill press. You might be closer than you think...

Cathar · 03-10-2005, 12:21 AM

Quote:

Originally Posted by bigben2k

You can shave that bottom lower than 1 mm; Cathar brought it down to ~0.1 mm (!). Square ended endmill, in a drill press. You might be closer than you think...

Not quite that thin.

Cascade/White Water had 0.7mm thick base-plates but this was also matched by the fine-ness of the internal structure. WW had 1mm wide fins/channels, while the Cascade had 2.0mm diameter cups.

Really the Cascade's base-plate should have been at least 1.0mm thick though in hind-sight for better low-mid range performance and would've made it more suitable for pumps like the DDC, and should've been about 2.0mm thick for pumps like the Eheim 1046. As it was, I designed both the Cascade and the White Water primarily for operation in the 6-10LPM flow rate range assuming pumps like an Eheim 1250 or better, and with 4LPM/1GPM as being merely "acceptable".

With 3mm wide cups I would personally select a base-plate thickness in a Cascade-like design to something in the 1.5-2.0mm range.

A 3.5mm center-to-center separation for 3.0mm cups seems about right to me too.

Oh, and 13/15 is just such a magical fraction of convenience.

Hoot · 03-12-2005, 01:58 AM

Well, I dug out my old Eheim 1048 and spare Chevette heater core. The NB and GPU blocks, along with 4) 12.7mm ID polished copper elbows in their paths are out of the loop. That leaves my Eheim 1250 w/12.7mm ID barbs, 12.7mm ID hoses, Camaro Heater Core w/ 12.7mm ID barbs (w/180x50mm Patriot fan) and the block. Head was maintained at .13m. Sad to say, despite hopeful wishes, still holding a 22C differential. This evening that means Water = 21c and Core = 43c running Prime95 Torture, In-Place Large FFTs fixed at 896k, Priority 10. I never mentioned it, but the coolant is distilled water and water wetter, standard ratio.

The temperature difference between these .88mm ID jets and the 1.67mm ID jets is less than 1c, but the visible flow rate is significantly slower, so I may as well go back to the larger jets. I guess I should be tickled that it edges the other blocks slightly. I can't wait until I can order the Storm G5 so I can see a significant improvement.

Oh yes, the GPU is now 2c cooler, not getting its water from the CPU block.

Hoot

Long Haired Git · 03-12-2005, 08:07 PM

Sad to see so little improvement. Flllowwww rraaatteesss pppllleeeeaaasssseeeee

Hoot · 03-13-2005, 12:12 AM

Quote:

Originally Posted by Long Haired Git

Sad to see so little improvement. Flllowwww rraaatteesss pppllleeeeaaasssseeeee

Imprecise Science Time.

Lacking a flow meter, I was compelled to resort to using a time-to-fill process. Actually, it was kind of interesting. Not the time-to-fill as much as the back pressure.

Here goes:

I have a 12.7mm ID burping tee between the output my radiator and the inlet to the block very near the block, .13m above the pump. I attached a 1m length of 12.7mm ID Tygon tubing to the tap on it, going straight up in the air. Using a 1L beaker, I timed how long it took to fill it in an open loop (no push-pull benefit). The pump was situated so that it did not need to lift the water more than a few cm on the intake. The original system head of .13m was maintained going to the radiator, then up through the block. The outlet tube from the block, going to the beaker was as close to level with the block as possible. I live at 314m ASL and the water temperature was 20c with no water wetter in it. I turned on the pump and let it run long enough to burp out most of the air in the system. I then diverted the outlet tube from the resevoir (sink) to the beaker. Ten passes yielded an average time-to-fill of 12 seconds, or 5LPM. The back pressure from the block pushed the 12.7mm diameter column of water 790mm up the tubing while filling the beaker. I had the whole family helping. My wife ran the stopwatch and my 11 year old daughter marked the column height while filling the beaker. "Smokie", my Dalmation just sat and looked at us with his head cocked. Without the block, the average time-to-fill the beaker after ten passes was 5 seconds or 12LPM. There was no water pushed up the column without the block. Many towels ended up in the washer and Smokie ran off eventually after getting a squirt out of mischief.

Flow rate w/o block = 12LPM
Flow rate w/block = 5LPM

Pressure = (((1.27cm * .5)^2)*Pi)*79cm = 100cc = ~100g of water, then a gap here because I'm not certain what units to use. Meters of water is mysterious because I don't know what diameter a column of water that is based upon, or is it a constant, independent upon volume?

Anyway, do those values allow for determining whether I'm in the vicinity of the sweet spot for the type of impingement I'm trying to hit? I'm a darn good Engineering Technician, but I'm not an Engineer.

Thanks for everyones input on this Jet process. Now it's time to clean up the mess.

Hoot

Straker · 03-13-2005, 02:25 AM

(could have sworn I registered here already, but...)

Pressure is constant, doesn't depend on volume per se. That's what makes hydraulic car lifts and such work - pressure is force divided by area, so you can lift the car (albeit slowly) by blowing in water through a tiny hose. If that seems funny (and it does, not trying to be patronizing), think of a teapot with a really long spout; when it's full of water, there's no reason you'd expect the water in the body of the pot to be anything but level with the water in the spout.

Hoot · 03-13-2005, 08:55 AM

That's why I'm not an Engineer. In my minds eye, it would take more pressure to lift a column of water 12mm in diameter 79cm than say it would to lift a column of water 1cm in diameter 79cm, simply because the first amount of water weighs more. In my minds eye, describing both of those as having the same pressure (.79m of H2O) gives my mind a hinky. That's why I'm not an Engineer...

Hoot

03-04-2005, 05:06 AM	#51
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Continuing my hijack of this thread - a couple of worked formulae for those looking to make a jet block: Head = ( 1.434.Flow^2 ) / ( count^2.radius^4 ) where Head is meters of water, Flow is in Litres per minute, count is the number of jets (so whole number), and radius is in millimeters. Quick example of a single, 5mm diameter jet @ 5 LPM we get: Head = ( 1.434 * 5 * 5 ) / ( 1 * 1 * 2.5 * 2.5 * 2.5 * 2.5 ) Head = 35.85 / 39.0625 Head = 0.918 meters. Now this does NOT take into consideration the friction in the jets or the cups or the barbs etc, and is just the waterblock resistance. But its a start. Oh, for the anal, 1.434 is: (1250 * density of water in Kg/m3 * conversion factor from Newton/m2 to meters H20) / ( 9 * Pi^2 ). Next, velocity of the jet is: v = (100 * Flow ) / ( 6 * Pi * count * radius^2 ) which is v = ( 5.305 * Flow ) / ( count * radius^2 ) So, from the example, v = ( 5.305 * 5 ) / ( 1 * 2.5 * 2.5 ) v = 26.525 / 6.25 v = 4.25 meters per second __________________ Long Haired Git "Securing an environment of Windows platforms from abuse - external or internal - is akin to trying to install sprinklers in a fireworks factory where smoking on the job is permitted." (Prof. Gene Spafford) My Rig, in all its glory, can be seen best here AMD XP1600 @ 1530 Mhz \| Soyo Dragon + \| 256 Mb PC2700 DDRAM \| 2 x 40 Gb 7200rpm in Raid-0 \| Maze 2, eheim 1250, dual heater cores! \| Full specifications (PCDB)

03-04-2005, 06:38 AM	#52
Les Cooling Savant Join Date: Oct 2001 Location: Wigan UK Posts: 929	Many (inc Cathar,I think) short cut P/Q/Velocity/Reynolds calculations by using Pressure Drop 5.0.It also contains useful information on fluid properties vs Temperature (Viscosity.Density) Find have to take as many short cuts as possible on the road from flow to die temperature. My preferred route for Jets is :- Flowrate --PD Drop---> velocity --Flomerics--> h(conv) --Kryotherm--> h(eff)--Waterloo--> "Die Temp"(isothermal coolant). Unfortunately Kryotherm only does channels, so for "Cups" you have to look elsewhere(dunno where) for conversion of h(convection) to h(effective). Last edited by Les; 03-04-2005 at 06:54 AM.

03-06-2005, 02:33 AM	#53
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	After many starts and restarts, I finally got the telescopic step down jets done. Too tired to reassemble tonight. Here's a look at the 31 Jets: Way too much work for what looks to me like not enough volume, but I'll be happy to be wrong tomorrow when I get time to test it. Oh yeah, i plugged the excess holes with homebrew SS set screws. Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

03-06-2005, 07:29 PM	#54
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Interested to find out, but I will continue to caveat my advice with the need to know what your blockless flow rate is and what the pump you're using is. Without these two items, we don't know what P-D you can cope with and what flow rate to aim for. __________________ Long Haired Git "Securing an environment of Windows platforms from abuse - external or internal - is akin to trying to install sprinklers in a fireworks factory where smoking on the job is permitted." (Prof. Gene Spafford) My Rig, in all its glory, can be seen best here AMD XP1600 @ 1530 Mhz \| Soyo Dragon + \| 256 Mb PC2700 DDRAM \| 2 x 40 Gb 7200rpm in Raid-0 \| Maze 2, eheim 1250, dual heater cores! \| Full specifications (PCDB)

03-06-2005, 10:54 PM	#55
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	As my signature says, I'm using an Eheim 1250 with .13m head. As for actual flow rate, it's sure a heck of a lot slower with this latest revision block in there. I put a piece of my daughters glitter in the water and watch how fast it goes around the loop of my three blocks, radiator and pump. It used to go around in a second or less with my Whitewater clone. Now it takes almost two seconds. I don't have access to a flow meter and really don't want to drag the box upstairs to the sink to see how long it takes to fill a 2 liter bottle, but I could. It's tough when your development rig is your main rig. The new revision block seems to be doing well, but it still only bests my WhiteWater clone by 1C at load at best. That's on an A64 running approximately 124W, not a Tbred running 70W. Once these A64s have been out long enough, I would expect to see more test results, whether scientific or anecdotal, to emerge from owners of mainstream commercial blocks. Heck, I'd love to see how a real Storm G5 delivers running in a rig similar to mine. I emailed Stew over a week ago inquiring when the preorder for the next batch will begin, using the email address at the web site, but I have yet to get a response. So far, the three blocks I have with track records back to Socket A and the new one all fall somewhere in the .16-.17c/w range on this 90nm core at approximately 124W. It's frustrating to not be able to see any rave deviation outside that range with the three or four variations I've tried with this new block. I keep ending up within the margin of error for the MAX6657 reader as compared to the MCW6002-64, WhiteWater clone and my old faithful PinFin block. My rig needs an enema... Edit: I'm using this Wattage Calculator to arrive at 124W. Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums Last edited by Hoot; 03-06-2005 at 11:05 PM.

03-07-2005, 07:32 AM	#57
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	Sorry to hear that Stew. Our family has been working through some creeping-crud as well. Luckily, I've dodged the bullet so far. Both the wife and daughter are down with it as I write this. I'm just waiting for the other shoe to drop, as it's rare that when one of us gets sick that the other two don't. Hope you're feeling better soon. Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

03-09-2005, 07:57 AM	#58
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	Well, the AS5 has had time to settle and still no joy. At load I'm seeing 22C water yielding 43-44C die temperature, which is still in the same range as the other three blocks. I'm starting to suspect that using an ambient system, it's going to be hard to break through that 22C differential barrier. I'm burned out on this block. About the only thing left to try, at a lot of effort, is to pull the NB and GPU blocks out of the loop, running the 1250 to the CPU block only. My gut feeling is that this is not a flow issue. I sincerely feel that I'm down to the source impedance of the high power-density core and the series resistance of the TIM being the limiting factors. I can't stress enough that "this is not your fathers Tbred" we're talking about. Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

03-09-2005, 03:06 PM	#59
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Ambient + 22 degrees is what my passive rig ran at too.... It'll be interesting to see what the Storm does for you. If you do pull that rig about, please do that flow test I was talking about... __________________ Long Haired Git "Securing an environment of Windows platforms from abuse - external or internal - is akin to trying to install sprinklers in a fireworks factory where smoking on the job is permitted." (Prof. Gene Spafford) My Rig, in all its glory, can be seen best here AMD XP1600 @ 1530 Mhz \| Soyo Dragon + \| 256 Mb PC2700 DDRAM \| 2 x 40 Gb 7200rpm in Raid-0 \| Maze 2, eheim 1250, dual heater cores! \| Full specifications (PCDB)

03-09-2005, 11:47 PM	#62
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	The G5 will perform best. You can shave that bottom lower than 1 mm; Cathar brought it down to ~0.1 mm (!). Square ended endmill, in a drill press. You might be closer than you think... __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

03-12-2005, 01:58 AM	#64
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	Well, I dug out my old Eheim 1048 and spare Chevette heater core. The NB and GPU blocks, along with 4) 12.7mm ID polished copper elbows in their paths are out of the loop. That leaves my Eheim 1250 w/12.7mm ID barbs, 12.7mm ID hoses, Camaro Heater Core w/ 12.7mm ID barbs (w/180x50mm Patriot fan) and the block. Head was maintained at .13m. Sad to say, despite hopeful wishes, still holding a 22C differential. This evening that means Water = 21c and Core = 43c running Prime95 Torture, In-Place Large FFTs fixed at 896k, Priority 10. I never mentioned it, but the coolant is distilled water and water wetter, standard ratio. The temperature difference between these .88mm ID jets and the 1.67mm ID jets is less than 1c, but the visible flow rate is significantly slower, so I may as well go back to the larger jets. I guess I should be tickled that it edges the other blocks slightly. I can't wait until I can order the Storm G5 so I can see a significant improvement. Oh yes, the GPU is now 2c cooler, not getting its water from the CPU block. Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

03-12-2005, 08:07 PM	#65
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Sad to see so little improvement. Flllowwww rraaatteesss pppllleeeeaaasssseeeee __________________ Long Haired Git "Securing an environment of Windows platforms from abuse - external or internal - is akin to trying to install sprinklers in a fireworks factory where smoking on the job is permitted." (Prof. Gene Spafford) My Rig, in all its glory, can be seen best here AMD XP1600 @ 1530 Mhz \| Soyo Dragon + \| 256 Mb PC2700 DDRAM \| 2 x 40 Gb 7200rpm in Raid-0 \| Maze 2, eheim 1250, dual heater cores! \| Full specifications (PCDB)

03-13-2005, 02:25 AM	#67
Straker Cooling Neophyte Join Date: Mar 2005 Location: Edmonton Posts: 1	(could have sworn I registered here already, but...) Pressure is constant, doesn't depend on volume per se. That's what makes hydraulic car lifts and such work - pressure is force divided by area, so you can lift the car (albeit slowly) by blowing in water through a tiny hose. If that seems funny (and it does, not trying to be patronizing), think of a teapot with a really long spout; when it's full of water, there's no reason you'd expect the water in the body of the pot to be anything but level with the water in the spout.

03-13-2005, 08:55 AM	#68
Hoot Cooling Neophyte Join Date: Dec 2002 Location: Twin Cities Posts: 67	That's why I'm not an Engineer. In my minds eye, it would take more pressure to lift a column of water 12mm in diameter 79cm than say it would to lift a column of water 1cm in diameter 79cm, simply because the first amount of water weighs more. In my minds eye, describing both of those as having the same pressure (.79m of H2O) gives my mind a hinky. That's why I'm not an Engineer... Hoot __________________ DFI LP UT NF4 Ultra-D 7.04-3 Bios A64 3700+ San Diego 11x257 (2827 Mhz) @ 1.57Vcore 2x1Gb Crucial Ballistix 7:6 (300 Mhz) 3-3-3-8-1T @ 2.77V Connect 3D x850xt 620/625 2x WD 36GB Raptors in Raid0 DIY Jet Impingement WB on CPU / Pipecap WB on GPU Shrouded Camaro Heater Core / Swiftech MCP655 Pump Antec P-160 Tower / Seasonic S12 600W PSU MAXIM 6658 Thermal Diode Reader Moderator at OCForums

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