What don't you like about your current WC setup?

Watercooling grommets are a standard on nearly all mid-range cases and high-end ones. I'm talking about internalized radiator support, where the mounting holes are present and can natively support a 4x120 or 3x140 radiator internally. I know the TJ11 (or one of those cases) can natively support a 4x140 radiator on the bottom, but I don't quite like the design of it (PSU blocks one end of it). Currently the most that cases will support without being a Mountain Mods case is a 3x120.
 
Tsumi said:
What the hell are you talking about? Fluid volume has absolutely no factor on performance unless the levels are so low that air gets into your system. And the only sealed systems are the Corsair H series or similar.

It's not too unreasonable to have to refill every so often, but once a month is a bit excessive. Just how much are you actually losing though?
Click to expand...

I think it'd just like to clear something up about this. Volume of fluid in the system ABSOLUTELY determines the performance of your loop because it dictates how much heat can be carried away from the chips to be dissipated by your rad/heat exchanger. A small volume of fluid circulating through the system will be at a much, much higher equilibrium temperature once the chips start adding heat. My loop has approx. 6L of coolant in it and the coolant temperature never goes higher than 32C with a GTX470 and i7 950 running at peak. It should also be noted that volume of coolant in the rad/exchanger is what factors in most here.

Fluid flow rate also matters because it increases the coolant in contact with the blocks over time and pulls heat away faster.There is a butter zone for any particular system that balances the heat pulled from the chips against the heat dissipated by the rads.

At least, this is what I learned in my heat transfer class at university for mechanical engineering.
 
Colonel_Panic said:
I think it'd just like to clear something up about this. Volume of fluid in the system ABSOLUTELY determines the performance of your loop because it dictates how much heat can be carried away from the chips to be dissipated by your rad/heat exchanger. A small volume of fluid circulating through the system will be at a much, much higher equilibrium temperature once the chips start adding heat. My loop has approx. 6L of coolant in it and the coolant temperature never goes higher than 32C with a GTX470 and i7 950 running at peak. It should also be noted that volume of coolant in the rad/exchanger is what factors in most here.

Fluid flow rate also matters because it increases the coolant in contact with the blocks over time and pulls heat away faster.There is a butter zone for any particular system that balances the heat pulled from the chips against the heat dissipated by the rads.

At least, this is what I learned in my heat transfer class at university for mechanical engineering.
Click to expand...

Once again, it has ABSOLUTELY NO FACTOR in performance. As long as your loop has no air in it, it does not matter how large your reservoir is. Either way, it will reach an equilibrium point, and that equilibrium point is determined by the temperature at which the rate of heat dissipated by the radiator equals the rate of heat added by the heat sources. The rate of heat dissipated by the radiator is determined by the amount of surface area the radiator has to the air and to the water, and the amount of airflow. It is NOT RELATED IN ANY WAY TO RESERVOIR SIZE. Larger radiators of course take more water and have more surface area in which to dissipate heat, but the point here becomes radiator size, not amount of coolant. It is common sense to fill the loop up to capacity, only an idiot runs a loop with air flowing through the blocks and radiator.

It has been proven multiple times that any flow rate over 1 GPM through each block has minimal (less than 1 C) effects on temperature. I have done my research, you clearly have not. I have read actual articles with proven testing and facts, whereas you're just spouting theories.
 
I suppose a particularly good scenario in which to prove such a point is a setup resembling the following:

- Overvolted (1.5v+) and overclocked (4.5Ghz+) i7 Bloomfield (45nm)
- GTX 480's (maybe QUAD-SLI) also overvolted and overclocked

Control WC Loop: CPU and GPU blocks (in parallel), pump (plenty of flow), and a single HW-Labs Black Ice GTX 560 (4x140, HIGH FPI radiator: probably the smallest, thinnest radiator (lowest volume of fluid) capable of handling this setup).

The catch: The high-FPI radiator is running shrouded, high-RPM (4K+) Delta or Sanyo Denki fans (say 220+ CFM) in push-pull. This obviously does not affect the volume of water. These radiators can dissipate ridiculous amounts of heat using high-pressure fans... they dominate

Experimental Loop: Same, but with a rather large (1 gallon perhaps), thermally insulated reservoir.

Alternate Experiment: Use JUST the CPU and a single GTX 140 radiator. Same idea, but lower heatload and water-volume.


I think the differences will be quite predictable. The setup with the larger reservoir will produce slightly better temperatures, but not by much, and it'd require a long duration of testing. Setups of large volume can exploit the high specific-heat of water... but this buffers only the water temperature over time. We must remember that a regular reservoir will slowly dissipate heat on its own = confounding factor... it's a (poorly performing) passive-radiator.

Such a buffering capacity is likely more useful in quiet setups, and... heh.. the reservoir probably shouldn't be thermally insulated.
 
Archmage said:
I suppose a particularly good scenario in which to prove such a point is a setup resembling the following:

- Overvolted (1.5v+) and overclocked (4.5Ghz+) i7 Bloomfield (45nm)
- GTX 480's (maybe QUAD-SLI) also overvolted and overclocked

Control WC Loop: CPU and GPU blocks (in parallel), pump (plenty of flow), and a single HW-Labs Black Ice GTX 560 (4x140, HIGH FPI radiator: probably the smallest, thinnest radiator (lowest volume of fluid) capable of handling this setup).

The catch: The high-FPI radiator is running shrouded, high-RPM (4K+) Delta or Sanyo Denki fans (say 220+ CFM) in push-pull. This obviously does not affect the volume of water. These radiators can dissipate ridiculous amounts of heat using high-pressure fans... they dominate

Experimental Loop: Same, but with a rather large (1 gallon perhaps), thermally insulated reservoir.

Alternate Experiment: Use JUST the CPU and a single GTX 140 radiator. Same idea, but lower heatload and water-volume.


I think the differences will be quite predictable. The setup with the larger reservoir will produce slightly better temperatures, but not by much, and it'd require a long duration of testing. Setups of large volume can exploit the high specific-heat of water... but this buffers only the water temperature over time. We must remember that a regular reservoir will slowly dissipate heat on its own = confounding factor... it's a (poorly performing) passive-radiator.

Such a buffering capacity is likely more useful in quiet setups, and... heh.. the reservoir probably shouldn't be thermally insulated.
Click to expand...

Which I would then estimate the difference after temps have fully stabilized to be maybe 2-3 C at the most, only due to slow heat loss from the reservoir. Nothing like the difference Colonel_Panic is trying to say.
 
Hmm. I guess these are all things to keep in mind when designing a loop. What component/element would you maximize to create the best possible loop?

- High flow rate
- super large radiator
- something I'm probably missing
 
I completely agree with Tsumi, reservoir size has no impact on your equilibrium loop temperatures. The best thing you can do for your temperatures is lots of radiator surface area and a lot of air flow over that surface area. The ability to remove the heat is your loops greatest strength.

The space of having 7 liters of reservoir would better be served with another radiator and fans.
 
Colonel_Panic said:
Hmm. I guess these are all things to keep in mind when designing a loop. What component/element would you maximize to create the best possible loop?

- High flow rate
- super large radiator
- something I'm probably missing
Click to expand...

Super large radiator. The biggest factor in temperature is how much heat the radiator can disperse, and that is mostly determined by the surface area of the radiator. I can get into more technical details with regards to airflow and fans and whatnot, but I'll stop here for now, unless I get a reply from you saying that you do want more technical details. However, the issue here becomes space limitation and cost issues, as large radiators take up a lot of space which some do not have room for, and they cost a whole lot.

Second factor in temperatures is the design of the blocks and how quickly they can transfer heat to the water.

Flow rate does not matter as long as it's kept above a certain level.
 
Tsumi said:
Once again, it has ABSOLUTELY NO FACTOR in performance. As long as your loop has no air in it, it does not matter how large your reservoir is.
Click to expand...

Aren't you being a bit simplistic about this? What if you don't have enough fluid in your loop? Then total fluid volume would be the limiting factor?:confused:


There will always be a limiting factor, unless you can somehow design the system to be in "perfect balance" between fluid volume, pump capacity, tube diameter, blocks, radiators and fans. And I'm sure I left something off that list.:D
 
x509 said:
Aren't you being a bit simplistic about this? What if you don't have enough fluid in your loop? Then total fluid volume would be the limiting factor?:confused:


There will always be a limiting factor, unless you can somehow design the system to be in "perfect balance" between fluid volume, pump capacity, tube diameter, blocks, radiators and fans. And I'm sure I left something off that list.:D
Click to expand...

Or how about you do your research, as I have done? Try reading my posts again and see if you can learn a few things.

If you're not running with enough water in your system, you're a complete idiot, deserve a complete bashing (well, not really), and should be banned from watercooling altogether. And seriously, how hard is it to keep your system full of water? Just go out to your local grocery store and buy a gallon of distilled water for ~$2.

Biggest limiting factors are radiator size, and how well the blocks can transfer heat. Smaller factors are how much water a pump can pump, but as I've said in the post you quoted and multiple times, flow rate does not matter as long as it's above a certain level. Beyond that, any other factors are negligible, unless you decide to be a complete idiot.
 
I have to refill my system about every 6 to 10 month and only a dribble at that.

I want fans that are temp controlled. They are all over, but I guess I am a complete dumb a$$ for not getting them working. The temp controlled fans would be on the radiator & turn on when necessary (under temp load). This would help keep the amount of dirt & lint accumulation down in the radiator.
 
I water cool to overclock the crap out of my components, and still run them cool and nearly silently.
What i dont like about water cooling in general is the lack of auxiliary air flow around the components. and full gpu water blocks are too expensive/only work on one product. Im still using my 8 year old DD maze4 because nothing i like more has come out yet. end up doing crazy-glue/as5 trick to get swiftech ramsinks on gpu memory and end up having to point a fan at the pcb to handle the vreg/ram.

ditto on motherboard vreg area, but asus has fixed that in the last bunch of years.. much heatsinking/surface area, if you bust off the heat pipe that leads to the nb and wc that by itself its all good.(even less of a issue these days with i7/etc)
 
Tsumi said:
I have a Swiftech Microres V2, and I do lose a bit of water, especially when things get hot. I can't seem to get the top cap tight enough. For example, I didn't game for ~2-3 weeks (finals and spring break), and there was barely any water loss. After I got back, I went on a gaming spree for about 4 hours a day, for 3-4 days, with fans running on low (which allowed the loop to get fairly hot). I easily lost ~ half in. of water in just that time period.

Also, the D5 is probably the quietest thing in my system. When my secondary hard drive is seeking (an old Maxtor), or my Velociraptor is defragmenting, I can hear those over the fans, which are louder than the pump.
Click to expand...
I don't think this is the root cause of my problems. After my last fill I used a quarter to tighten the cap beyond my normal finger tight level. When I did todays topup there was a hiss of pressure equalization when I unscrewed the cap; so I don't think it's the source of any loss.
 
The part I hate the most is the amount of pet hair and dust the system collects on the rad.
 
Oh, just to add to the maximizing pump flow rate thing that I just remembered.

The higher flow rate might keep water from heating up much in the waterblock, theoretically providing lower temps. But then the opposite thing happens with the radiators, where the water flows through the radiator too fast for the heat to be removed. So it all balances out above a certain flow rate. Now, if you had a water source that was at a constant temperature, and the water after the blocks are disposed of, then high flow rates would matter.
 
You must log in or register to reply here.
Back
Top