Reservoir bypass ???

momo56

momo56

momo56

Soldato
Joined
18 Jul 2009
Posts
3,148
Apparantly a loop with no reservoir will mean a much higher flowrate so i was
thinking to create a by pass with a combination of tlines and ball valves. This way i can switch between a tube or a reservoir in the loop. Is this worth doing? Also will the increased flow rate make the pump louder?

My loops will be:

Pump>CPU>Reservoir>Motherboard,Ram>Reservoir
Pump>GPU1 tline GPU2>reservoir

All my 3 rad are 360 mm and both pumps are D5's
 
Not sure I would agree that res-less loops have higher flow-rate. This is presumably based on the idea that the water slows down in the res? Hmm. The pump creates a pressure difference between its inlet and its outlet causing water to be moved from the inlet to the outlet. The pressure difference is all that matters.

I have tried using pumps is series and although the flowrate is indeed higher than with one, it would be flawed logic to attribute this to the higher flowrate entering the second one. More, it is the combined pressure differential that makes it flow faster.

So, I think you're wrong. In addition, there's no such thing as a closed loop. Even with Tygon, there is water loss through the tube and this will mean that the water will need topping up. how will you do this without a res (or a T-line, my personal preference)?
 
There is a flow rate differance - martin and skinnee have lots of info on it.

it's due to water being a non compressable liquid so in a 'closed' loop water out of the pump = water in.
Just use martins flow rate caculator and add a res (flow rate drops)
Less work the pump needs to do (with a given wattage) = greater flow.
There is a down side but only if you over spec your loop.
(like a D5@5 with a single block highflow loop) - resulting in possable cavitation or pump burn out.

I use an inline res - but fully fill it a day or two after bleeding with a fill line to remove the bubble (compressable) and form a closed loop.
best of both worlds
 
Last edited:
Flow rate makes next to no difference to cooling performance anyway as long as it's not a trickle, just go with what's least hassle - in my opinion a T-line.
 
All it will do is keep temps down a bit for the first few minutes - after that the temperatures will even out over the whole loop and no amount of flow will make any difference.
 
Flow does make a differance - over 1.5 gmp maybe not.
but at .5 to 1 it make a massive differance.
and at 1 to 1.5 it make a small be significant differance.

We don't all run over spec'ed loops - my D5 runs at 2 for near silence.
so larger tubing, no res, ek top all add up to give me 1gmp flow with a multi block loop

edit: use martins flow rate caculator momo56 and get the actual numbers of different loops.
Why rely on forum guesses when a very reliable guy has does LOTS of testing for you.

As a rule any thing over 1 gpm is fine - 1.5+ is even better.
and don't over engineer you loop - K.I.S.S. is allways a good starting point.

also if you have a D5 the typhoon3 res is worth a look for twin loop set ups
 
Last edited:
MikeTimbers said:
Not sure I would agree that res-less loops have higher flow-rate. This is presumably based on the idea that the water slows down in the res? Hmm. The pump creates a pressure difference between its inlet and its outlet causing water to be moved from the inlet to the outlet. The pressure difference is all that matters.

I have tried using pumps is series and although the flowrate is indeed higher than with one, it would be flawed logic to attribute this to the higher flowrate entering the second one. More, it is the combined pressure differential that makes it flow faster.

So, I think you're wrong. In addition, there's no such thing as a closed loop. Even with Tygon, there is water loss through the tube and this will mean that the water will need topping up. how will you do this without a res (or a T-line, my personal preference)?
Click to expand...


i see what you're saying but i'll have to disagree about needing to top up the water, i've been running a closed loop for over a month now and i found that if you get the air bubbles out, once it's filled it's filled.

I had no cooling issues with it at all
 
Mike is right about the water loss - but we are talking above VERY small quantities.
and the tube will easily flex to account for any reduction after a month (possable a year - but thats a guess)
 
The question you have to ask is if water is leaching out of the tube, is it being replaced by air, or is the volume lost forever such that the tubing must suck itself inwards to compensate? Over 12 months the water level in my EK150 reservoirs must have dropped by 2" now I haven't actually took the top blank off to see if theres a vacuum in there or whether air has actually managed to replace the water, would be interesting to find out when I get round to strip down for maintenance. My feeling is it must be dragging air in somehow, as that is quite a change in volume without seeing tubing collapsing anywhere in the loop.

As for flowrates, the flowrate you get in any pumping system is a function of the head the pump can deliver versus flow, i.e. the pump curve, and the system pressure, i.e. the system curve. Where these two curves cross is what flow you are going to get. The system curve is dependant on static and dynamic pressure losses, but the key thing is that any fitting will have an acsciated pressure loss with it. So a res will have a certain pressure loss as a unit, forget about whether water velcity changes as it goes through, its irrelevant, its just what happens to the pressure across the inlet and outlet and how this affects the system curve. If you are inserting a bypass then the two T pieces, and the ball valve if you choose to use a ball valve between them, all have their own pressure losses as well, so what you might gain from taking out the res, you might lose with the three extra fittings, may be better, may be worse.
 
good point bubo - ever barb and bend has an effect too
and fluids being pushed round 90 degree 'bends' in a T can also be modeled in the flow rate caculator.

the thing with a t-line in a loop is the flow travels straight across it, not up it.
so if making your ball valve bypass - remember to have the res on the non-direct path.

.........l
.-----T
.l.......l
res.....l
.l.......l
.-----T
.........l

edit: should have just used paint
 
Last edited:
still being made (available direct)
Production is buy two guys in a workshop - hence it's lack of stock, I think they make about 20 a day (they are both XS forum members).

edit: looks like primochill are pushing they new res, and was some issues with the glue in the T3, and increased noise
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom