Pump Selection Sanity Check

effnish · 02-07-2006, 07:10 PM

I'm building a new setup for my A64 3000+ and I'm still trying to nail down which pump to get. I've already decided on a 12v pump and I'm building the system more around quiet cooling than absolute performance. With that in mind I ran a few numbers to find out what the performance trade-offs were for each of 3 pumps. This is my first attempt at the exercise so let me know if you see anything blatently wrong.

Based on what I calculated no pump is a bad choice but I'm leaning towards the CSP-Mag for Price, quietness, and performance

Equipment list so far:
BIP III
Apogee
MCW60

Using the numbers I could find for the parts above, I did the analysis to see what my expected Water-Air temp delta would be for a: MCP350, MCP655 (#1-#5 setting), and CSP-Mag.

System resistance curve fit: PSI Drop = 0.3324x3 - 0.21x2 + 2.1156x - 0.704 where x is the flow rate in GPM. Number includes an additional 10% resistance for tubing

Rough estimate of 100W for the CPU and 50W for the GPU.

MCP350: 1.33 GPM, .037 C/W CPU, .039 C/W GPU, .032 C/W BIP III, 8W pump
CSP-MAG: 1.07 GPM, .039 C/W CPU, .040 C/W GPU, .034 C/W BIP III, 8W pump
MCP655 #5: 1.61 GPM, .036 C/W CPU, .038 C/W GPU, .031 C/W BIPIII, 15W pump
MCP655 #4: 1.41 GPM, .037 C/W CPU, .038 C/W GPU, .032 C/W BIPIII, 14W pump
MCP655 #3: 1.15 GPM, .039 C/W CPU, .039 C/W GPU, .033 C/W BIPIII, 6.5W pump
MCP655 #2: 0.89 GPM, .042 C/W CPU, .042 C/W GPU, .035 C/W BIPIII, 4W pump
MCP655 #1: 0.66 GPM, .046 C/W CPU, .046 C/W GPU, .036 C/W BIPIII, 2.5W pump

bigben2k · 02-07-2006, 07:23 PM

Interesting. Nice work!

Long Haired Git · 02-07-2006, 09:05 PM

Check out the review where they measure how much heat dump each pump does.

Where did you get the data for the BIP three?

My approximator is being updated as we speak...
Will post URL soon.

Dryadsoul · 02-07-2006, 09:14 PM

Great JoB! I see your data confirms my choice of the Laing D5 as fair one then!
Thanks and quiet wasn't my concern.

effnish · 02-08-2006, 10:00 PM

I pulled the BIPIII numbers off of the first page of the "How much Pump is enough" post. The numbers are actually for a thermochill 120.3 but they work in a pinch.

My pump heat dump numbers probably aren't too acurate as I did them last and wasn't concentrating due to my wife talking trash about me being a nerd.

ElMoIsEviL · 02-11-2006, 10:38 AM

Courtesy of Coolercases.co.uk

Eheim 1046
1046 = 6 LPM
(didn't have correct fittings to carry on with this one)

Eheim 1048
1048 = 8.7 LPM
1048 + TDX (nozzle1) = 5.9 LPM
1048 + TDX + Maze4 GPU = 5.2 LPM
1048 + TDX + Maze4 Chipset = 5.4 LPM
1048 + TDX + Maze4 GPU + Maze4 Chipset = 4.7 LPM
1048 + TDX + EVO1 = 4.3 LPM
1048 + RBX (nozzle1) = 5 LPM
1048 + White Water = 5.4 LPM

Eheim 1250
1250 = 16.4 LPM
1250 + TDX (nozzle1) = 8.3 LPM
1250 + TDX + Maze GPU = 6.9 LPM
1250 + TDX + Maze4 Chipset = 7.2 LPM
1250 + TDX + Maze4 GPU + Maze4 Chipset = 6.3 LPM
1250 + TDX + EVO1 = 3.15 LPM
1250 + RBX (nozzle1) = 6.9 LPM
1250 + White Water = 7.6 LPM
1250 + Maze4 (copper top) = 11.1 LPM
1250 + Maze3 = 8.7 LPM

Eheim 1260
1260 = 18.9 LPM
1260 + TDX (nozzle1) = 12.6 LPM
1260 + TDX + Maze4 GPU = 8.8 LPM
1260 + TDX + Maze4 Chipset = 9.7 LPM
1260 + TDX + Maze4 GPU + Maze4 Chipset = 8.9 LPM
1260 + TDX + EVO1 = 4.5 LPM
1260 + TDX + RBX (nozzle1) = 9.5 LPM
1260 + TDX + White Water = 10.5 LPM

Laing D4
Laing D4 = 16.4 LPM
Laing D4 + TDX (nozzle1) = 9.5 LPM
Laing D4 + TDX + Maze4 GPU = 8.4 LPM
Laing D4 + TDX + Maze4 Chipset = 8.9 LPM
Laing D4 + TDX + Maze4 GPU + Maze4 Chipset = 7.8 LPM
Laing D4 + TDX + EVO1 = 4.2 LPM
Laing D4 + TDX + RBX (nozzle1) = 8.4 LPM
Laing D4 + TDX + White Water = 9.2 LPM

C-system Pump (single)
C-system = 5.5 LPM
C-system + TDX (nozzle1) = 4.7 LPM
C-system + TDX + Maze4 Chipset = 4.4 LPM
C-system + TDX + Maze4 Chipset + Maze4 GPU = 3.9 LPM

C-system Pump (2 Pumps)
2x C-system = 8.6 LPM
2x C-system + TDX (nozzle1) = 5.8 LPM
2x C-system + TDX + Maze4 Chipset = 5.2 LPM
2x C-system + TDX + Maze4 Chipset + Maze4 GPU = 4.7 LPM

Laing DDC-12v (test carried out with 3/8" to 1/2" step up barbs)
Laing DDC = 7.2 LPM
Laing DDC + TDX (nozzle1) = 6.3 LPM
Laing DDC + TDX + Maze4 Chipset = 6.0 LPM

Laing DDC-12v (test carried out with modded 1/2" inlet )
Laing DDC = 12.2 LPM
Laing DDC + TDX (nozzle1) = 9.7 LPM
Laing DDC + TDX + Maze4 Chipset = 8.6 LPM

Note: Alphacool has a new revision of the pump out ( Lang DDC-1 Plus) which again has an increased flow rate that once arrives will be tested in the same manner. (cheers Leeum)

Amendment 25/01/06
Laing DDC-1Plus With Alphacool Plexi Top ½” Barbs (side mount inlet & outlet)
Laing DDC-1Plus = 9.7 LPM
Laing DDC-1Plus + TDX (nozzle1) = 8.4 LPM

Laing DDC-1Plus With Alphacool Plexi Top ½” Barbs (top mount inlet - side mount outlet)
Laing DDC-1Plus = 14.5 LPM
Laing DDC-1Plus + TDX (nozzle1) = 10.5 LPM

Added 29/01/06
Laing DDC-1Plus With Alphacool Plexi Top ½” Barbs (top mount inlet - side mount outlet)
Laing DDC-1Plus + NV-78 = 11.5 LPM
Laing DDC-1Plus + NV-78 + TDX (nozzle1) = 9.2 LPM

Laing DDC-1Plus + Alphacool NVXP-3 6800/7800 = 6.9 LPM
Laing DDC-1Plus + Alphacool NVXP-3 6800/7800 + TDX (nozzle1) = 6.9 LPM

Laing D5
Laing D5 (set @ no 5) = 16.4 LPM
Laing D5 (set @ no 1) = 4.4 LPM

Here's a look at the pumps used which gives a good insight into physical size.......from left to right

Eheim 1260>>>Eheim 1250>>>Eheim 1048>>>Eheim 1046>>>Laing D4

Here's a Laing D4 Pump along side the tiny Laing DDC-12v

This is the new Laing D5 which is exactly the same physical size and shape to the Laing D4 apart from offering an inbuilt rheostat to control flow rates and being black in colour.
Set on No5 (highest setting) the pump is very very quite though still gives a little noise where as set on No1 (lowest setting) i would say the pump is virtually silent.
Set on No4 it would seem noise level to be similar to the Laing DDC

And one of the D5 Rheostat

Long Haired Git · 02-12-2006, 06:29 PM

Cheers ElMoIsEviL !!!!

Whilst I prefer the pump review from that French site as it included watts added to the coolant, I can use these figures to derive PQ charts for the Maze4 chipset and GPU coolers.
Which will have to do until I can convince RoboTech to do a proper test of them.

02-11-2006, 10:38 AM	#6
ElMoIsEviL Cooling Neophyte Join Date: Jan 2006 Location: Ottawa, Ontario Posts: 71	Re: Pump Selection Sanity Check Courtesy of Coolercases.co.uk Eheim 1046 1046 = 6 LPM (didn't have correct fittings to carry on with this one) Eheim 1048 1048 = 8.7 LPM 1048 + TDX (nozzle1) = 5.9 LPM 1048 + TDX + Maze4 GPU = 5.2 LPM 1048 + TDX + Maze4 Chipset = 5.4 LPM 1048 + TDX + Maze4 GPU + Maze4 Chipset = 4.7 LPM 1048 + TDX + EVO1 = 4.3 LPM 1048 + RBX (nozzle1) = 5 LPM 1048 + White Water = 5.4 LPM Eheim 1250 1250 = 16.4 LPM 1250 + TDX (nozzle1) = 8.3 LPM 1250 + TDX + Maze GPU = 6.9 LPM 1250 + TDX + Maze4 Chipset = 7.2 LPM 1250 + TDX + Maze4 GPU + Maze4 Chipset = 6.3 LPM 1250 + TDX + EVO1 = 3.15 LPM 1250 + RBX (nozzle1) = 6.9 LPM 1250 + White Water = 7.6 LPM 1250 + Maze4 (copper top) = 11.1 LPM 1250 + Maze3 = 8.7 LPM Eheim 1260 1260 = 18.9 LPM 1260 + TDX (nozzle1) = 12.6 LPM 1260 + TDX + Maze4 GPU = 8.8 LPM 1260 + TDX + Maze4 Chipset = 9.7 LPM 1260 + TDX + Maze4 GPU + Maze4 Chipset = 8.9 LPM 1260 + TDX + EVO1 = 4.5 LPM 1260 + TDX + RBX (nozzle1) = 9.5 LPM 1260 + TDX + White Water = 10.5 LPM Laing D4 Laing D4 = 16.4 LPM Laing D4 + TDX (nozzle1) = 9.5 LPM Laing D4 + TDX + Maze4 GPU = 8.4 LPM Laing D4 + TDX + Maze4 Chipset = 8.9 LPM Laing D4 + TDX + Maze4 GPU + Maze4 Chipset = 7.8 LPM Laing D4 + TDX + EVO1 = 4.2 LPM Laing D4 + TDX + RBX (nozzle1) = 8.4 LPM Laing D4 + TDX + White Water = 9.2 LPM C-system Pump (single) C-system = 5.5 LPM C-system + TDX (nozzle1) = 4.7 LPM C-system + TDX + Maze4 Chipset = 4.4 LPM C-system + TDX + Maze4 Chipset + Maze4 GPU = 3.9 LPM C-system Pump (2 Pumps) 2x C-system = 8.6 LPM 2x C-system + TDX (nozzle1) = 5.8 LPM 2x C-system + TDX + Maze4 Chipset = 5.2 LPM 2x C-system + TDX + Maze4 Chipset + Maze4 GPU = 4.7 LPM Laing DDC-12v (test carried out with 3/8" to 1/2" step up barbs) Laing DDC = 7.2 LPM Laing DDC + TDX (nozzle1) = 6.3 LPM Laing DDC + TDX + Maze4 Chipset = 6.0 LPM Laing DDC-12v (test carried out with modded 1/2" inlet ) Laing DDC = 12.2 LPM Laing DDC + TDX (nozzle1) = 9.7 LPM Laing DDC + TDX + Maze4 Chipset = 8.6 LPM Note: Alphacool has a new revision of the pump out ( Lang DDC-1 Plus) which again has an increased flow rate that once arrives will be tested in the same manner. (cheers Leeum) Amendment 25/01/06 Laing DDC-1Plus With Alphacool Plexi Top ½” Barbs (side mount inlet & outlet) Laing DDC-1Plus = 9.7 LPM Laing DDC-1Plus + TDX (nozzle1) = 8.4 LPM Laing DDC-1Plus With Alphacool Plexi Top ½” Barbs (top mount inlet - side mount outlet) Laing DDC-1Plus = 14.5 LPM Laing DDC-1Plus + TDX (nozzle1) = 10.5 LPM Added 29/01/06 Laing DDC-1Plus With Alphacool Plexi Top ½” Barbs (top mount inlet - side mount outlet) Laing DDC-1Plus + NV-78 = 11.5 LPM Laing DDC-1Plus + NV-78 + TDX (nozzle1) = 9.2 LPM Laing DDC-1Plus + Alphacool NVXP-3 6800/7800 = 6.9 LPM Laing DDC-1Plus + Alphacool NVXP-3 6800/7800 + TDX (nozzle1) = 6.9 LPM Laing D5 Laing D5 (set @ no 5) = 16.4 LPM Laing D5 (set @ no 1) = 4.4 LPM Here's a look at the pumps used which gives a good insight into physical size.......from left to right Eheim 1260>>>Eheim 1250>>>Eheim 1048>>>Eheim 1046>>>Laing D4 Here's a Laing D4 Pump along side the tiny Laing DDC-12v This is the new Laing D5 which is exactly the same physical size and shape to the Laing D4 apart from offering an inbuilt rheostat to control flow rates and being black in colour. Set on No5 (highest setting) the pump is very very quite though still gives a little noise where as set on No1 (lowest setting) i would say the pump is virtually silent. Set on No4 it would seem noise level to be similar to the Laing DDC And one of the D5 Rheostat __________________ Asus P5N32-E SLI \| Intel Core 2 Extreme QX6700 @ 4.23GHz (Swiftech MCW5002-775T 226W Pelt) \| 4x1GB OCZ Special Ops DC PC2-6400 DDR2 XTC \| 2xBFG TECH GeForce 8800 GTX OC 768MB Water Cooled \| Sound Blaster X-Fi FaT@l1Ty FPS Sound Card \| 2x74GB WD Raptor 10,000 RPM RAID0 \| 2x320GB Seagate 7200.10 7,200 RPM RAID0 \| SuperCooled: Enterprise Custom Cooling

02-12-2006, 06:29 PM	#7
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Re: Pump Selection Sanity Check Cheers ElMoIsEviL !!!! Whilst I prefer the pump review from that French site as it included watts added to the coolant, I can use these figures to derive PQ charts for the Maze4 chipset and GPU coolers. Which will have to do until I can convince RoboTech to do a proper test of them. __________________ 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)

02-07-2006, 07:10 PM	#1
effnish Cooling Neophyte Join Date: Jun 2004 Location: Colorado Posts: 4	Pump Selection Sanity Check I'm building a new setup for my A64 3000+ and I'm still trying to nail down which pump to get. I've already decided on a 12v pump and I'm building the system more around quiet cooling than absolute performance. With that in mind I ran a few numbers to find out what the performance trade-offs were for each of 3 pumps. This is my first attempt at the exercise so let me know if you see anything blatently wrong. Based on what I calculated no pump is a bad choice but I'm leaning towards the CSP-Mag for Price, quietness, and performance Equipment list so far: BIP III Apogee MCW60 Using the numbers I could find for the parts above, I did the analysis to see what my expected Water-Air temp delta would be for a: MCP350, MCP655 (#1-#5 setting), and CSP-Mag. System resistance curve fit: PSI Drop = 0.3324x3 - 0.21x2 + 2.1156x - 0.704 where x is the flow rate in GPM. Number includes an additional 10% resistance for tubing Rough estimate of 100W for the CPU and 50W for the GPU. MCP350: 1.33 GPM, .037 C/W CPU, .039 C/W GPU, .032 C/W BIP III, 8W pump CSP-MAG: 1.07 GPM, .039 C/W CPU, .040 C/W GPU, .034 C/W BIP III, 8W pump MCP655 #5: 1.61 GPM, .036 C/W CPU, .038 C/W GPU, .031 C/W BIPIII, 15W pump MCP655 #4: 1.41 GPM, .037 C/W CPU, .038 C/W GPU, .032 C/W BIPIII, 14W pump MCP655 #3: 1.15 GPM, .039 C/W CPU, .039 C/W GPU, .033 C/W BIPIII, 6.5W pump MCP655 #2: 0.89 GPM, .042 C/W CPU, .042 C/W GPU, .035 C/W BIPIII, 4W pump MCP655 #1: 0.66 GPM, .046 C/W CPU, .046 C/W GPU, .036 C/W BIPIII, 2.5W pump

02-07-2006, 07:23 PM	#2
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Re: Pump Selection Sanity Check Interesting. Nice work! __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

02-07-2006, 09:05 PM	#3
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Re: Pump Selection Sanity Check Check out the review where they measure how much heat dump each pump does. Where did you get the data for the BIP three? My approximator is being updated as we speak... Will post URL soon. __________________ 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)

02-07-2006, 09:14 PM	#4
Dryadsoul Cooling Neophyte Join Date: Feb 2006 Location: Western face of Mt. St. Helens Posts: 25	Re: Pump Selection Sanity Check Great JoB! I see your data confirms my choice of the Laing D5 as fair one then! Thanks and quiet wasn't my concern.

02-08-2006, 10:00 PM	#5
effnish Cooling Neophyte Join Date: Jun 2004 Location: Colorado Posts: 4	Re: Pump Selection Sanity Check I pulled the BIPIII numbers off of the first page of the "How much Pump is enough" post. The numbers are actually for a thermochill 120.3 but they work in a pinch. My pump heat dump numbers probably aren't too acurate as I did them last and wasn't concentrating due to my wife talking trash about me being a nerd.

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