|
|
Testing and Benchmarking Discuss, design, and debate ways to evaluate the performace of he goods out there. |
Thread Tools |
11-08-2005, 01:05 PM | #1 |
Cooling Neophyte
Join Date: Jan 2003
Location: Spain
Posts: 25
|
Test bench.
Hi.
I expect to write in the suitable place, because I wanted to do a few commentaries to them. Time ago, I measured a low difference of temperature between entry and exit of the water block, when de water block is mounted over my processor. I supposed that it would be due to the losses in the Socket, a bad measurement of my system, or that really the consumption of the processor is minor enough that the indicated by some programs. Recently I finish a test bench, except necesary modifications and for the present time in the first tests, I have confirmed this low consumption. Here, I suppose that you will have done similar tests and I wanted to know, if really they have measured something similar and in what proportion to be able to compare my measurements Any commentary will be grateful. |
11-08-2005, 02:43 PM | #2 |
Cooling Savant
Join Date: Jan 2003
Location: Sydney, Oz
Posts: 336
|
Water has a very high specific heat.
Every gram can soak up approx four joules of energy for only a one degree rise. Take five litres per minute of flow, and some maths shows this is approx 80 grams per second whizzing through the block. Watts is also a "per second" measurement, being joules per second. So, a CPU dumping 80W is dumping 80 joules per second. So, each gram of water is responsible for soaking up one joule of energy. Flip back to the opening sentence, and you'll see an increase of only approx 1/4 of one degree. This is within the loop. The operating temperature of the CPU is set by the operating temperature of the coolant. This is determined by how hot the coolant has to be relative to the ambient air for the radiator (bong, tec, fridge, water cooler, whatever) to be able to loose the same energy as entering the system.
__________________
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) |
11-08-2005, 04:54 PM | #3 | |
Cooling Savant
Join Date: Sep 2003
Location: Vallentuna, Sweden
Posts: 410
|
Quote:
What sort of numbers are you getting dnkroz? At what flow rates, power levels? Example from one of my measurements attached. Flowrate calculated from pressure drop and MCW6000 K-factor 1.843 (lpm/sqrt(mBars)) |
|
11-09-2005, 09:51 AM | #4 |
Cooling Neophyte
Join Date: Jan 2003
Location: Spain
Posts: 25
|
Sorry, definitively I have not explained well.
Thank you for you response Long Haired Git, I know the functioning and this information, but I was referring to concrete information of measurements. Thank you for you also Incoherent, for the graph. Better, I will put the concrete information. I have a XP Thoroughbred, some programs indicate me, that this processor must have a consumption in power about 68W, but I have measured a low power in W across the waterblock, lower was expecting for me. Concretly I have measured between the entry and exit of the block an average of 0.11ºC for a flow of 313.4l/h (82.79 gal/h), aprox 40W. The heats losses I think that they are too high in relation to the calculation of the program and I thought, between other things, that I was measuring badly. Recently I have finished a bank of test whit a...heat die? and I have repeated the test by means of a resistance with a consumption of 41W measuring in water a difference across the block, an average of 0.12ºC and average also flow of 309.6l/h (81.78gal/h), 43.1W. The reason of commenting all that, it is because I wanted to confirm my information in similar cases comparing power generated by the processor. |
11-09-2005, 11:52 AM | #5 | |
Cooling Savant
Join Date: Sep 2003
Location: Vallentuna, Sweden
Posts: 410
|
Quote:
An old thread might be relevant for you. I got 49W measuring power using the old fluxblock concept on a Barton 2500. Your measurement doesn't seem too far out. Cant remember the Thoroughbred specs vs Barton though. |
|
11-10-2005, 12:27 PM | #6 |
Cooling Neophyte
Join Date: Jan 2003
Location: Spain
Posts: 25
|
Good news. Thank you very much Incoherent
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|