Citation de "Stefan"

The prob is not the pump, maybe the rad .. and sure you can name the components, just make abbreviations.

most interesting part .. how big is the radiator and at what voltage the fans are running ?

I think there's a slight misunderstanding. I'm not blaming the pump. All I'm saying is that the sum of the parts is responsible for the eventual performance of a system. I agree that flow isn't as critical as many people appear to claim it is. However, you have to meet a certain minimal threshold. Once you go too far below about 1 l/m it's going to have a greater affect on the overall temp than just 1-2C.

My system consists of a WC HTSF Triple radiator, 3x Papst 4412 2GL fans, HK 2.5 CPU, HK 2.1 NB, HK GPU-X² 8800, Aquastream pump, Aquaero controller, AC flow meter (get 87 l/h on 74hz and run 65 hz at 71 l/h), 5.25" Aquabox Plexiglas, 10/8 tubing through the whole system.

Case is Lian Li 343B with the right hand side completely for the cooling of the system. Fresh air is sucked in at the front and pumped out at the back and the top. The other half of the case basically isn't affected by the radiator heat. Temps on the mobo side are only 1-2C above room temperature at all times.

The fans are controlled by the Aquaero unit. When stress testing they are running 100% at around 1530 rpm. During warm above 25C ambient conditions CPU temps can reach 65C with Prime 95 or OCCT. During heavy gaming on the other hand my CPU only reaches around 53C and the GPU doesn't often go above 50C in the same conditions.

All the above running at: QX6700 3.3Ghz @ 1.35V, FSB/NB 333Mhz @ 1.55V, RAM 1000Mhz 4-4-4-12 at 2.12V, GPU 643/2106. Mobo is a D975XBX2.

(BTW, contact with the CPU and HK 2.5 is good. The individual CPU cores only differ 1-2C at maximum.)

The GPU just uses up a lot of the cooling capacity of a system, therefore my recommendation of the above for an SLI system, Quad CPU, and mosfet cooling blocks.

Here's a pic to verify my build so you can see it's not some crap set-up

http://www.netgeek.net/wc.jpg

If you use a different flow meter you should be able to go up to about 150-160 LPH at your current Aquastream pump setting. The one you need starts with Inn...

Alternatively you can also drill out the stock AC flow meter, but then you would need to calibrate the flow yourself. The stock AC flow meter artificially limits the flow to about 80-90 LPH due to the tiny 3.3 mm hole in the intake side. If you are looking for the best flow then this flow meter mod is the way to go as the Inn... unit also has limitations on flow due to inlet size restrictions.

http://s12.photobucket.com/albums/a225/T…erAssembly3.jpg

http://s12.photobucket.com/albums/a225/TopNurse/AC/?action=view&current=ACFlowmeterAssembly4.jpg


BTW, nice rig. ;D

Thanks Nurse.

Here's an OCCT graph of a test I ran yesterday. Air intake temperature was 24.7C, water temp 32.1C at full load.

Citation de "Top_Nurse"

Alternatively you can also drill out the stock AC flow meter, but then you would need to calibrate the flow yourself.  The stock AC flow meter artificially limits the flow to about 80-90 LPH due to the tiny 3.3 mm hole in the intake side.  If you are looking for the best flow then this flow meter mod is the way to go as the Inn... unit also has limitations on flow due to inlet size restrictions.

I'll try that when I have to take my stuff apart again for the next gen GPU in Nov/Dec.. For easy calibrating I'll get another AC flow meter and stick it in the same loop as the one I have. Both the modded and stock version should essentially give the same readout after pulse correction. My guess is though that a bigger passage hole will increases the minimum flow reading threshold. So, instead of being able to read all the way down to lets say 8l/h it won't respond below 20-30l/h.

Citation de "divve"

I'll try that when I have to take my stuff apart again for the next gen GPU in Nov/Dec.. For easy calibrating I'll get another AC flow meter and stick it in the same loop as the one I have. Both the modded and stock version should essentially give the same readout after pulse correction. My guess is though that a bigger passage hole will increases the minimum flow reading threshold. So, instead of being able to read all the way down to lets say 8l/h it won't respond below 20-30l/h.

You only want the bigger hole flow meter in your loop. The only way to calibrate is by doing a series of 5-10 tests where you measure the amount of water pumped over a period of one minute. Toss the high and low figures and average the rest. This figure is the LPM. You then adjust the Aquaero impulses per liter till the display reads what you got in your tests.

Since the stock Aquastream will run considerably more than 20-30 LPH at 46 Hz, you really don't need to worry about that.

My plan is to only use two flow meters in a test loop for the calibration process. When both stock and mod version have the same readout I'll know its calibrated properly...at least as good as the original that is...

Citation de "divve"

My plan is to only use two flow meters in a test loop for the calibration process. When both stock and mod version have the same readout I'll know its calibrated properly...at least as good as the original that is...

That isn't going to work properly as the stock unit will impede the flow of the the modded unit.

It should work. You only can't use this test outside the flow range achievable with only the modded flow meter. However, when things match up between lets say 50-100 l/h, it'll be reasonable to assume that accuracy will be close enough beyond that.