![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Having lived with my water cooled project parallel HEX for a little while now, I thought I would post some analysis to where the idea was originally born; here in the Water Cooling forum. I hope you find it interesting!
Everyone arranges their water cooling blocks in series, what would happen if the blocks were arranged in parallel?
For those that havent read my worklog or dont have the time, heres a brief history:
History
What seems like aeons ago, I set about routing my water blocks in an unorthodox parallel arrangement instead of the usual series methodology. The theory behind this was based on the same principals of calculating the total resistance in an electrical circuit. Quoted from wikipedia:
Hydraulic Analogs
While the terms voltage, current and resistance are fairly intuitive terms, beginning students of electrical engineering might find the analog terms for water flow helpful. Water pressure, measured by international units in pascals (and commonly in units of pounds-force per square inch), is the analog of voltage because establishing a water pressure difference between two points along a (horizontal) pipe causes water to flow. Water flow rate, as in liters (or gallons) of water per minute, is the analog of current, as in coulombs per second. Finally, flow restrictors such as apertures placed in pipes between points where the water pressure is measured are the analog of resistors. We say that the rate of water flow through an aperture restrictor is proportional to the difference in water pressure across the restrictor. Similarly, the rate of flow of electrical charge, i.e. the electrical current, passing through an electrical resistor is proportional to the difference in voltage measured across the resistor.
Still with me? I'll carry on...
The resistance of a series circuit compared to a parallel circuit (electrically speaking):
(Rm1 in the above equation is the resistance of the manifold used to split the circuit in to 6, which is in series with the parallel blocks and must have some resistance.)
Much less resistance right?
The Circuit
Heres a schematic of how I was going to achieve connecting the blocks in parallel.
The Collector and Distributor
To be able to split a single loop into six, I build the Distributor:
To merge the loops back into one and to act as a reservoir and drain point, I built the Collector:
I also had to modify the AC Cuplex XT cpu block to be 3-ports (1 in, 2 out):
Here's some gratuitous, self pimping pics of the finished article:
More pictures are over in my worklog.
Temperatures
All this is pointless without temperatures, so here they are. These are hardly scientific, with all the variance of temp sensors, changing ambients and the relatively uncontrolled enviroment. Still what I would call good though!
Stock Air Temperatures:
Water Cooling - 1 'D5' pump, 1 ThermoChill 'PA120.2' radiator (no case):
Water Cooling - 2 'D5' pump, 1 ThermoChill 'PA120.2' radiator (no case):
Water Cooling - 2 'D5' pump, 1 ThermoChill 'PA120.2' radiator (Final in the case):
The ambients are pretty high in the last chart, I've been meaning to do another set when it's a bit cooler. I included the Dt columns to show the difference in temps from the ambient, to show the change.
I'm pretty happy with the temps. I need to crank all fans up to about 80% to keep the machine happy when playing games. The pump speed only changes temps about 2C from the '1' setting to the '5' setting, I'll admit I expected more of a change. When the pc is idling or performing general low cpu tasks, I can turn get away with 2 fans off and 2 fans on min! (I rigged the fan controller so that one dial controls one radiator fan and one case fan simultaneously, as does the other dial).
The good, the bad and the downright wierd
One effect of setting the pump any higher than about 4, was the turbulence in the Collector. It was pretty bad, so much so the entire circuit was a mix of air and coolant.
When I was benching the system out of the case, this settled down considerably because I filled the Collector as high as I could, plus most of the air was out of the system.
Once I eventually built the case and put everything in situ, the turbulence in the Collector took on a very strange effect. I can only describe it as foaming! It didn't do it before, everything was cleaned thoroughly before assembly. The only thing I changed was the UV additive.....I used AC's BlueMotion this time.
Here's what it's like after 10 minutes with the pumps on max:
A few days later, and my lovely blue loops were completely clear, no colour to them at all and certainly not UV reactive. The blue you see above is the UV leds, which my camera pick up as blue
.
Strangely, the foaming in the Collector had it's benefits. Look closely at the images before. The separation between the 'foam' and the undisturbed water is clearly visible. I've watched it and no bubbles are dragged into the circuit. Quite remarkable.
See for yourself here:
Video 1 of 6 - Night time illumination, shows the bubbling effect in the Collector.
Video 2 of 6 - Day time lighting, starts with the Collector effect, then reducing it.
Video 3 of 6 - Day time lighting, close up of the Collector effect and a general pan.
Video 4 of 6 - Night time illumination, with the laser being activated.
Video 5 of 6 - Night time illumination, with the laser being activated (again but longer).
Video 6 of 6 - Day time lighting, with the collector effect very visible.
Cheers, and thanks for the forum all the help!
Everyone arranges their water cooling blocks in series, what would happen if the blocks were arranged in parallel?
For those that havent read my worklog or dont have the time, heres a brief history:
History
What seems like aeons ago, I set about routing my water blocks in an unorthodox parallel arrangement instead of the usual series methodology. The theory behind this was based on the same principals of calculating the total resistance in an electrical circuit. Quoted from wikipedia:
Hydraulic Analogs
While the terms voltage, current and resistance are fairly intuitive terms, beginning students of electrical engineering might find the analog terms for water flow helpful. Water pressure, measured by international units in pascals (and commonly in units of pounds-force per square inch), is the analog of voltage because establishing a water pressure difference between two points along a (horizontal) pipe causes water to flow. Water flow rate, as in liters (or gallons) of water per minute, is the analog of current, as in coulombs per second. Finally, flow restrictors such as apertures placed in pipes between points where the water pressure is measured are the analog of resistors. We say that the rate of water flow through an aperture restrictor is proportional to the difference in water pressure across the restrictor. Similarly, the rate of flow of electrical charge, i.e. the electrical current, passing through an electrical resistor is proportional to the difference in voltage measured across the resistor.
Still with me? I'll carry on...
The resistance of a series circuit compared to a parallel circuit (electrically speaking):
(Rm1 in the above equation is the resistance of the manifold used to split the circuit in to 6, which is in series with the parallel blocks and must have some resistance.)
Much less resistance right?
The Circuit
Heres a schematic of how I was going to achieve connecting the blocks in parallel.
The Collector and Distributor
To be able to split a single loop into six, I build the Distributor:
To merge the loops back into one and to act as a reservoir and drain point, I built the Collector:
I also had to modify the AC Cuplex XT cpu block to be 3-ports (1 in, 2 out):
Here's some gratuitous, self pimping pics of the finished article:
More pictures are over in my worklog.
Temperatures
All this is pointless without temperatures, so here they are. These are hardly scientific, with all the variance of temp sensors, changing ambients and the relatively uncontrolled enviroment. Still what I would call good though!
Stock Air Temperatures:
Water Cooling - 1 'D5' pump, 1 ThermoChill 'PA120.2' radiator (no case):
Water Cooling - 2 'D5' pump, 1 ThermoChill 'PA120.2' radiator (no case):
Water Cooling - 2 'D5' pump, 1 ThermoChill 'PA120.2' radiator (Final in the case):
The ambients are pretty high in the last chart, I've been meaning to do another set when it's a bit cooler. I included the Dt columns to show the difference in temps from the ambient, to show the change.
I'm pretty happy with the temps. I need to crank all fans up to about 80% to keep the machine happy when playing games. The pump speed only changes temps about 2C from the '1' setting to the '5' setting, I'll admit I expected more of a change. When the pc is idling or performing general low cpu tasks, I can turn get away with 2 fans off and 2 fans on min! (I rigged the fan controller so that one dial controls one radiator fan and one case fan simultaneously, as does the other dial).
The good, the bad and the downright wierd
One effect of setting the pump any higher than about 4, was the turbulence in the Collector. It was pretty bad, so much so the entire circuit was a mix of air and coolant.
When I was benching the system out of the case, this settled down considerably because I filled the Collector as high as I could, plus most of the air was out of the system.
Once I eventually built the case and put everything in situ, the turbulence in the Collector took on a very strange effect. I can only describe it as foaming! It didn't do it before, everything was cleaned thoroughly before assembly. The only thing I changed was the UV additive.....I used AC's BlueMotion this time.
Here's what it's like after 10 minutes with the pumps on max:
A few days later, and my lovely blue loops were completely clear, no colour to them at all and certainly not UV reactive. The blue you see above is the UV leds, which my camera pick up as blue
.Strangely, the foaming in the Collector had it's benefits. Look closely at the images before. The separation between the 'foam' and the undisturbed water is clearly visible. I've watched it and no bubbles are dragged into the circuit. Quite remarkable.
See for yourself here:
Video 1 of 6 - Night time illumination, shows the bubbling effect in the Collector.
Video 2 of 6 - Day time lighting, starts with the Collector effect, then reducing it.
Video 3 of 6 - Day time lighting, close up of the Collector effect and a general pan.
Video 4 of 6 - Night time illumination, with the laser being activated.
Video 5 of 6 - Night time illumination, with the laser being activated (again but longer).
Video 6 of 6 - Day time lighting, with the collector effect very visible.
Cheers, and thanks for the forum all the help!