How much extra pressure for an Eisbaer 420 ?

R9 Fury

New member
Hi all !

I recently acquired an Eisbaer 420, and for maximum performance at the lowest possible noise level, decided to install the radiator outside of the PC case, a Corsair Carbide 300R, some 25mm above the mesh on the top and with the rearmost third of the radiator sticking out over the back of the case, as shown in the attached picture.

To do so, I had to add 2 blue tubes, 11/8mm and some 40cm long each, to extend the loop enough, and about a month after putting everything together, having decided to give a quick check, tilted the tower and added some extra water in the loop. In the process, the CPU temperatures jumped a few degrees, and since the waterblock is tightened enough, it's most likely some air bubble trapped in some inconvenient place.

Since I intend to add a future GPU to the loop, and having seen that some air bubbles may have a hard time moving around the loop, I'm thinking of adding an extra pump to help push the bubbles to the right spots more quickly, and of course for slightly improving thermals.

But according to the Eisbaer 420's technical data, the radiator is tested to 1.5 bars, and the included pump already has a maximum head of 0.85m, so is there enough margin to add a powerful pump to the loop ? For example, if I add the Eisstation VPP with its VPP755 pump and its 4.25m max head pressure, this would result in up to 5.1 bars if both pumps are directly in series in the loop.

So would it be safe to add such a pump ? Or would you say there's a (lower) maximum value for the pressure of the extra pump not to exceed ?

See you
 

Attachments

War4peace

Member
1. You don't need another pump, but I usually recommend dual-pump setups for redundancy and peace of mind.
2. Pressure is not a problem unless you have a blockage somewhere. Think of it this way: you could unscrew the reservoir cap while the system is running and liquid will not jump out of there, meaning the internal pressure is equal to atmospheric pressure. So if nothing is blocking your liquid path, there is no problem.
3. More pumps mean faster liquid flow but would not necessarily kick out bubbles from where they are. In order to get the bubbles out (and I suspect they form in the uppermost area of the loop, which is the radiator) I suggest the old and tried method of "dancing" with the PC but in your case it's much simpler. While the system is running, take the radiator from where it is and move it below the reservoir (vertically) in such a way that the reservoir is the uppermost area of the loop. Then slowly tilt the radiator from one side to the other on all axis (pitch, yaw, roll) - the bubbles will come out of the radiator and into your reservoir. Afterwards, shut down the system and top up your reservoir, then put the radiator back in its original position.
 

Vanzin

Support
Staff member
HI R9 Fury,

as War4peace already told you try to dance with your case or shake the radiator.
Regarding the VPP755 pump they do not have high pressure. They have <0,5Bar so you could add one of the pump to the system. If you want to expand the system, we would also recommend to add an second DC-LT pump to the system like the Eiswolf GPX Pro or an VPP755 to your system.
 

R9 Fury

New member
OK, so I detached the radiator from the case, in order to shake it like a madman while holding it at a lower height than the pump (didn't want to disconnect the tubes to take the whole AIO out of the case though), and that did push a few air bubbles to the pump, so I could add maybe an extra 20-40ml of fluid to the loop.

The temperatures didn't seem to change, at least not something that is sticking out like a sore thumb.

But here's something that's more striking ; using a meat thermometer pressed against the fittings on the radiator, I measured 33.0°C on the hot fitting, 24.4°C on the cold one. You can use "A Heat Transfer Textbook" to derive what would be the water temperatures inside the tubes themselves, but these temps should be already a pretty accurate proxy of the water temperatures.

Since P = c*delta_T*dm/dt, you get : dm/dt = P/(c*delta_T)

The dissipated power P of the loop is about 70W with the CPU and pump combined, c = 4180 J/(kg*Kelvin) for specific heat of water, so one gets : dm/dt = 70/(4180*8.6) = 0.0019 kg/s = 7.01 liters per hour.

So there's something clearly reducing the flow rate from its theoretical maximum of 70 l/h, and the pump's RPM, at least reported by the motherboard, is between 1232 and 2662 RPM, so nothing special here.

Any guess as to what causes the flow rate to be so low ?
 

War4peace

Member
The maximum flow rate for the VPP755 pump is 350 L/h. 7 L/h is insanely slow, this is 116 ml/minute, it's not possible to have such low flow unless there's a serious blockage somewhere.
For any watercooling loop I always recommend getting a flow meter and at least one temperature probe.
 

War4peace

Member
Oh, so sorry, I misread your OP.
In that case do not under any circumstance keep that pump. It's too weak for a full loop.
 

R9 Fury

New member
But the loop is just the original Eisbaer 420 with 2 extra tubes, each around 40cm long, so in the worst case scenario, that's an extra...

0.0010005 Pa*second for mu the dynamic viscosity of water at 20°C (it decreases with higher temps)

So the Hagen-Poiseuille equation gives : Delta_P = 8*mu*L*Q/(pi*R^4) = 8*0.0010005*0.8*1.94E-5/(pi*0.004^4) = 155 Pa, or 1.55mm of H2O for extra pressure drop with the absolute maximum flow rate of 70 L/h with the pump. And I doubt the fittings to connect these 2 tubes create that much extra of a pressure drop either.
 

War4peace

Member
We're not on the same page, so please allow me to detail a bit.
Equations are great... in theory. They work very well on paper, but looking at them alone will tell you zilch about what is actually happening in reality.
To be honest I didn't even know there was a pump so weak, for example a D5 pump has a maximum theoretical flow rate of 1500 L/h, with the emphasis on "theoretical". In practice, I have measured 650 L/h on a D5 pump using nothing but the pump, a 30cm 16/10 hose sunk in a bucket of water and a flow meter. My PC has two D5 pumps in series with a pretty complex loop and the flow rate is 313 L/h right now, and that's because my waterblocks are linked in parallel, not in series. If I turn off one pump, the flow rate drops to 240 L/h, which means the flow rates don't add up like they would in theory. Therefore adding a pump won't magically boost your flow rate, instead you would be better off with replacing the weak pump with a strong one, because the weak one would likely slow down the strong one by adding flow restriction.

Now, please don't focus on "pressure" because in a closed circuit loop, the "pressure" of a pump is misleading. The correct term is "head pressure" which means how high would a pump raise liquid in a vertical column until the state of equilibrium is reached (liquid no longer rises inside the column with the pump turned on). A pump's head pressure is only useful if you have the highest point of the loop at a certain height, and that height must be lower than the pump's head pressure. For example, if a pump has a head pressure of 1m and the highest point of the loop is at 1.1m height from the pump, the flow will be zero because the pump won't be strong enough to push the liquid over the highest point and establish flow. If the highest point is 90 cm above the pump, the flow will be established and once the liquid starts flowing, all that matters is flow restriction from components (waterblocks, radiator and bends such as 90 degree fittings).

To summarize: restriction decreases flow, not head pressure. Height reduces head pressure but not flow rate. Think of it this way: you have a recipient with a liquid and you put a hose in it, the hose goes up, over the container edge and then down, with the other end of the hose at a height lower than the bottom of the recipient. If you put your mouth on the lower end of the hose and draw liquid from the recipient, the liquid would go up and once it reaches the apex of the hose, it will start to flow downwards by itself, without external help. It's the communicating vessels principle, which is applicable to a closed liquid loop.
 

davido_labido

Moderator
Staff member
The dc-lt should be fine for this, but it will struggle without any help if there is an airlock.

Adding an extra pump would work, but also will manually manipulating the cooler. This isn't always easy to do though as it's like one of the marble mazes on occasion and unless you get it to the exact angle that the unit needs for the bubble to escape, it can be tricky.

Also, war4peace is correct In pretty much everything he has said, the only thing that I would mention is that adding a faster pump another pump to a slower pump won't be that detrimental as the restriction would be minor in your instance. In much larger builds other builds where there are many blocks and points of restriction, a dc-lt added would do as he says and actually cause more of a restriction than it would help.
 

davido_labido

Moderator
Staff member
I know, this isn't always the case though depending on the pump/top and block, especially if it is in a smaller loop. It won't increase the flow at all, but it also won't hinder it any in a small loop.
 
Top