Doing it wrong - AIO Orientation

[Mikkle]

I understand that when an AIO is mounted with the pump at the highest point in the loop, it will likely suffer due to air collecting in the pump, which can be bad for the motor/impeller/etc. This is a good reason not to mount rads below the CPU block in most typical AIOs. Problem is, in some (mostly ITX SFF) cases, there is simply no other option. 
 

For cases that require an “inverted” orientation as mentioned above, I’m wondering whether it’s enough of a solution to select one of those oddball AIOs that locate the pump in the radiator or inline with the tubes? Certainly this would protect the pump, keeping it submerged in liquid. However, any air in the loop is still going to find its way to the CPU block, which will eventually hurt cooling potential, correct? This would hurt thermal performance and could potentially even damage the CPU block as plastics are exposed to heat that can’t be dissipated by the liquid as designed. I doubt it would be a problem from day one (assuming you get a reasonably full AIO) but as permeation occurs and more air finds its way into the system, I can imagine that the relatively weak impeller pump would struggle to push liquid across a dried out CPU block. In fact if this were the case, wouldn’t you still eventually see too much strain on the pump, if it has to work harder and harder with degrading results?

 

I’m kind of wondering... if I already don’t have any other options aside from the inverted orientation (CPU block up top), does it even matter where the pump is?

[8-Bit Ninja]

It's unlikely to be an issue honestly, however if your concerned about it, then I've noticed a few of those AIO's with the pumps located in the rad also have a reservoir with a fill port, which means you would be able to bleed and fill the loop as required. Meaning getting enough air in the loop to reduce the flow by enough to hurt cooling would be near impossible. 

[Mikkle]

12 minutes ago, 8-Bit Ninja said:

It's unlikely to be an issue honestly, however if your concerned about it, then I've noticed a few of those AIO's with the pumps located in the rad also have a reservoir with a fill port, which means you would be able to bleed and fill the loop as required. Meaning getting enough air in the loop to reduce the flow by enough to hurt cooling would be near impossible. 

Yeah, that’s kind of where I’m leaning...  I know there’s a general wisdom advocating against the inverted layout... these oddball AIOs sort of offer a silver bullet to pump failure, but I generally don’t trust silver bullets on principle. While they might solve part of the problem, they could still be vulnerable in other ways - whether or not I’m aware. Inverted may be suboptimal regardless, but I end up wondering how much I care about that, if it still basically works. If it means the thing will go kaput in a few months, or there’s a risk of damage to the CPU/system, I definitely care. Just hard to judge how likely any of that is. 

[emosun]

Just turn the case upsidedown

[Entropy.]

1 hour ago, Mikkle said:

I understand that when an AIO is mounted with the pump at the highest point in the loop, it will likely suffer due to air collecting in the pump, which can be bad for the motor/impeller/etc. This is a good reason not to mount rads below the CPU block in most typical AIOs. Problem is, in some (mostly ITX SFF) cases, there is simply no other option. 
 

For cases that require an “inverted” orientation as mentioned above, I’m wondering whether it’s enough of a solution to select one of those oddball AIOs that locate the pump in the radiator or inline with the tubes? Certainly this would protect the pump, keeping it submerged in liquid. However, any air in the loop is still going to find its way to the CPU block, which will eventually hurt cooling potential, correct? This would hurt thermal performance and could potentially even damage the CPU block as plastics are exposed to heat that can’t be dissipated by the liquid as designed. I doubt it would be a problem from day one (assuming you get a reasonably full AIO) but as permeation occurs and more air finds its way into the system, I can imagine that the relatively weak impeller pump would struggle to push liquid across a dried out CPU block. In fact if this were the case, wouldn’t you still eventually see too much strain on the pump, if it has to work harder and harder with degrading results?

 

I’m kind of wondering... if I already don’t have any other options aside from the inverted orientation (CPU block up top), does it even matter where the pump is?

Well, yes a lot of that is true. An AIO with the pump integrated with another part of the loop could help, but there might be some drawbacks. Tbh I'd just get the ssupd meshlicious. 

[8-Bit Ninja]

28 minutes ago, Mikkle said:

Yeah, that’s kind of where I’m leaning...  I know there’s a general wisdom advocating against the inverted layout... these oddball AIOs sort of offer a silver bullet to pump failure, but I generally don’t trust silver bullets on principle. While they might solve part of the problem, they could still be vulnerable in other ways - whether or not I’m aware. Inverted may be suboptimal regardless, but I end up wondering how much I care about that, if it still basically works. If it means the thing will go kaput in a few months, or there’s a risk of damage to the CPU/system, I definitely care. Just hard to judge how likely any of that is. 

The CPU would shutdown before any damage is caused, and most AIO's carry a multi year warranty. While not ideal the worse that would happen is your system being down for a week or two whilst you wait for an RMA