Single Loop Liquid Cooling Help.

windsymphony

Honorable
Jul 9, 2012
52
0
10,630
8xlusj.jpg


This is m y first time on liquid cooling.

The components are .

MCP 355, RES/PUMP COMBO
Swiftech Apogee Drive II Block/Pump Combo
EK-ASUS GTX 680 Direct CU II Waterblock.

1*120mm RAD EK Cool Stream
3*120mm RAD Back Ice GTX Xtreme 360 Radiator

Tubing is : Tygon B-44-4X 1/4" ID (3/8" OD) - "Non-Porous" Smooth Tubing.

...............

Iam conflicted as i came to know that flow through pump puts some heat on the loop , so taking it straight to RAD and than to MY GTX680 .

My CPU is 3770k i use it on turbo , but no overclocking.


Input is greatly appreciated ., thank you.
 
Solution
The pump should NEVER be running dry - it's got ceramic in it that will grind and ruin itself without lubricant, even if it's for just a few seconds. Basically what you want to do is fill the reservoir to the point where the water is sitting at the intake of the pump... that's why your reservoir should always be directly before and physically above your pump.

For two pumps, you want to keep the second one off until the loop is filled, yes.

HugoStiglitz

Distinguished
loop order makes no difference to component temperature.

because water is quite a good conductor of heat (which is why its used in cooling loops, cars & etc) the temperature difference from one point in you loop to another. (e.g. from just after your cpu to just after your radiator) is so tiny it makes no difference to temperatures of your components.

the diff is less than 0.1 of a degree.

the only thing that matters is you must have you res higher than your pump (in the case of split res & pump combo)
 
First, why in the world are you water cooling if you aren't going to overclock? Even if it's for silence, it seems like a ridiculous expense.

Second, stop worrying about the parts making heat flowing to radiators. It's not that the water goes here, gets warmed up, goes here, ect... it's moving so quickly that there's NOTHING like that; all you should worry about is total heat produced and total heat removed from the loop. The order should be whatever restricts flow the least and looks best.
 

windsymphony

Honorable
Jul 9, 2012
52
0
10,630
Will there be a problem between the pump of Apogee Drive II and the MCP 355 (Two Pumps in Loop)....?

and dark sable , i live in Middle East . Air cooling is just too noisy for a "Over the Desk" Case and not effective for 6 months( summer)
 


Absolutely not - the Apogee II works great with a second pump in the equation, and takes a lot of the stress off of longer loops.

That being said, you probably only need one or the other.
 

HugoStiglitz

Distinguished
yes, definatly DO NOT have 2 pumps in a loop

unless they are exactly matched speed (even 2 pumps of the same brand / model will not be exact) it causes major problems, it forces the stronger pump to push the weaker pump which puts additional stress on BOTH pumps.

you can cause a pump to fail and seize, causing your system to rupture and spill liquid everywhere.
unless your going to run split loops or run them parallel instead of series.(complicated double Y pieces) for a redundant pump setup I would get rid of the 2nd pump.

get rid of the Apogee Drive II and get a Raystorm
 

windsymphony

Honorable
Jul 9, 2012
52
0
10,630
In Middle East summer time , Hot Air gets pumped into the case and , i dont want to overclock CPU cuz i believe 3.9 is good for any game, while iam going to overclock my GPU to get good output.

And i think i need a different option than getting rid of apogee , cuz i just got it shipped from USA , IT COST ME A LOT :( .

SO I HAVE TO MAKE DO WITH WHAT I HAVE GOT., There is next to nothing available here in my country on water cooling .


so is there anything else i can do to make a good loop , that wont urt my components causing leak and the pump themselves.
 
Hugo, people have been using pumps in series for a long time; it works perfectly well and provides redundancy.

Read this article: http://martinsliquidlab.org/2011/04/26/pump-setup-series-vs-parallel/2/

That being said, you absolutely don't need to do it. OP, if you haven't already bought the reservoir, then you don't need the built in pump; the apogee has a great one built in already.
 


Nah, there's no need to remove it. If you're worried about the pumps fighting each other, just don't plug one of them in. That being said, there should be no problem with running both pumps at once - it won't really give any benefit, but it shouldn't hurt unless one of the pumps is producing a lot of extra heat.
 
The 335 is pretty noisy, but it produces 15ft of head as opposed to the Apogee's... 14.5ft.

There's really basically no difference between them. If noise matters, use the Apogee. If it doesn't, see if you can take off the pump and just use the waterblock underneath, as it'll look cleaner.
 

windsymphony

Honorable
Jul 9, 2012
52
0
10,630
and if possible , can you brief me on " run dry" , i do know a little about the term and what it means here in liquid cooling , but i do not understand it fully. for example when i first connect a reservoir top of a pump and pour coolant , and hit start on test loop, wont the motor run dry for atleast a second or two ., and lets say that i just let both pump be in equation ., and letting the reservoir on top of the first pump lets a flow on the first pump, but what about the second pump ?. it would run dry for few seconds for sure.


should i first let the "natural flow" for sometime without turning on both the pumps , so first motor gets it den , keep second motor off until a flow is registered and than turn on second motor?
 
The pump should NEVER be running dry - it's got ceramic in it that will grind and ruin itself without lubricant, even if it's for just a few seconds. Basically what you want to do is fill the reservoir to the point where the water is sitting at the intake of the pump... that's why your reservoir should always be directly before and physically above your pump.

For two pumps, you want to keep the second one off until the loop is filled, yes.
 
Solution

HugoStiglitz

Distinguished
DarkSable

I am aware that adding a 2nd pump will give better head but the added power draw is not worth it until you have the largest possible single pump available. then adding a 2nd will add extra head that is not possible with a single pump scenario.

however being that he has only 2 blocks, and neither of which is overly restrictive making dual pumps a mute point. your introducing another point of failure for no gain not to mention more power draw for do difference in cooling ability.

once you get far enough that a single high power pumps is not able to cope I would recommend dual loops rather than pushing the pressure up higher and higher with dual pumps as your more likely to bust an O ring.

either of those 2 pumps on their own will be able to handle the cooling of the system fine.
 
It's not another point of failure, it's a point of redundancy preventing failure; if one pump fails, the other one continues to operate.

That being said, no, there's no point in doing it, and it's not needed, but I disagree that it's dangerous in any way, shape, or form. Just run each pump at a lower speed, or turn one off.
 

HugoStiglitz

Distinguished
but a better option for redundancy is for parallel not series.

I agree that series adds redundancy,.
I agree that series adds extra head to the loop

but series vs parallel, it adds failure for no benefit.

in series if one pump fails and seize up, it will cause a blockage, causing the other pump to pressurise the tubes which will cause a major leak.
in parallel you get the same advantages of double pumps but instead of having the small chance of catastrophic pump failure & seize the flow rate will drop but you will not have a the same chance of leaking.

yes we are talking a very small chance, but putting in dual pumps for redundancy is already working with small chances.
 


Parallel loops, or parallel pumps? The article I linked above pretty much sinks parallel pumps, and using two loops wouldn't really add redundancy. (They'd have to never come in contact, so unless there are waterblocks that have two separate channels within them, then...)

With two pumps in parallel on the same loop, if one fails, the pressure drops horribly and the system is crippled. With two pumps in series, if one fails, it doesn't create a blockage - the only way that would happen is if it got stuck, and there aren't going to be an contaminants large enough for that to happen with. Go read the article I linked; it explains all this.
 


Again, go read the article I linked. If you're running pumps in parallel and one of them seizes, you're in big trouble.

That being said, I'm confused. Unless the motor somehow absolutely seizes on the impeller, how would that cause blockage? Also, you'd still be just as screwed if you were running with a single pump and that happened - at least with two you have one that can at least try to move water still.

 

HugoStiglitz

Distinguished
I have read it.

in parallel you would be fine if 1 pump seized(zero effect to the system, other than reduced flow rate). 2nd pump goes as normal, it would be exactly the same as having a single pump loop.

c4079169_vbattach227565.png


in parallel a blockage in pump A would mean the water would just flow via pump B (flow rate would be reduced to that of a single pump)

yes I know its a very small chance of a pump failure causing it to seize. but redundancy is planning for as many failures as you can.

 
Parallel:
Rather than the DDC pump pushing through the intended loop, it is now bypassing that intended loop via the PMP-450 subloop. That’s bad new for parallel loops, as you lose not only the one pump that quit, but you also loose about 83% of the remaining pump’s power.

However….it does still have “Some” pumping power left over. While it’s not going to be of any good levels, chances are your loop will still get between .3GPM and .8GPM which will function without causing catastrophic failure. In the end…it will still serve a redundancy benefit..but barely.

Series:
Many forum members have suggested that a pump is highly restrictive when stopped and imply that performance in the remaining pump is degraded heavily…is it true? [...] No.

While you do lose the power of the second pump that was turned off, the restriction of the still pump is is very small. On average you will only see about a 2% loss in pumping power vs. a single pump setup. Series setups retain nearly full pumping power of single pump setup and therefore pass the redundancy check with flying colors. Not only do you get adequate redundancy, you pretty much get full single pump setup power. No problem…
 

HugoStiglitz

Distinguished
Yup,

only because you do not have a non return valve

in applications where the extra head is not needed almost all redundant pumps will be configured in parallel.
I work at a water treatment facility and have double checked with the engineers here.
 

HugoStiglitz

Distinguished