I have owned and used this "Flow sensor high flow G1/4 for aquaero" for many years. It is hooked into an Aquaero 5. On startup, the unit reads up to 30 l/h however it "decays". By this, I mean the reading slowly reduces until 15 l/h (which I assume is the lower limit) and then reads 0.0 l/h.


This water loop has had this problem before, but it went away somehow. The only reason why it is not working anymore is because I drained the loop and re-filled. It was reading 22 l/h before I removed some constrictions (the purpose of the drain).

An odd note is that the flow rate momentarily increases over 15 l/h when the reservoir is opened and air pressure is released if done quickly after startup.
The loop is actually working. The D5 PWM is moving at 4800 rpm, I can see flow in the reservoir, and the CPUs are getting cooled, the radiator has a sub-cooling margin of 5 K.

Could any of these be the problem?

• The coolant is not pure water, it is mayhems pastel. Possible but unlikely because it worked before on the same coolant.
• The coolant is not fully deaerated. How would this affect the flow sensor?
• The loop has air pockets that prevent full coolant flow. Impossible because it should never increase above 0.0 l/h in this case.
• The flow sensor is old and broken. How come it works sometimes then?
• The flow sensor is positioned too close to the pump outlet. This wasn't a problem before, it wasn't moved during the drain.
• The new style cable is a much smaller AWG than the old one I replaced it with. (I broke the original, both OEM
• The Aquaero is getting a bad signal due to floating ground or something.
• The conductivity is too low on the new coolant. I am really reaching on this one.
  

I am really out of ideas after those. Google searches have turned up nothing.

What could be the problem? Thanks in advance.

Zitat von »Dundys«

That is way below flow rate you should be getting.

I was thinking this. But I have no way to confirm. Only one flow sensor.

Zitat von »Dundys«

If your pump shows that is running at 4800RPM I wouldn't think that there is anything wrong with your wiring, unless if you try to adjust the pumps speed say via-voltage just for the purpose of test and the RPM it's not changing then I would suspect something its up with connections.

Nope, Adjustments haven't been made yet.

Zitat von »Dundys«

I don't know if your blocks are clear and you can see it through, unless there is a blockage of any kind or perhaps there is something abstracting the paddle wheel of the USB flow meter itself from spinning freely I would look into that. Another thing could be your pumps impeller itself but to be honest, that one is a long shot. With your pump running at that speed which is MAX, even with big loop and thick Pastel, you should be getting close to 100l/h at least.....

Yeah, which is why I upgraded to a D5 because the previous pump was showing so low flow.

Zitat von »Dundys«

I would check the wheel paddle first then the blocks

If I am going to check the paddles, I might as well get another flow sensor.

Do the MPS flow sensors work with pastels?

Zitat von »InfoSeeker«

Zitat von »sunborn«

On startup, the unit reads up to 30 l/h

Your best flow rate, 30 l/h, is a very low flow rate. You are running below the flow sensor's specifications. I would not trust those result.
You may need a second pump. And yes, pastels can be pretty, but they generally gunk up over time.

I do have a second pump for the second CPU. It is running at 3200 rpms.

I am going to do one better and take the silly thing apart from the top. I am sure it will void the warranty if it was still in effect.

Zitat von »Dundys«

Simple way to test for the paddle issue is and you will know this right away, when you have your meter out look into inner-out thread and just blow gently air into it, the wheal should spin for few seconds but make sure inside is dry, if it spins its all good, if it stops dead or seems like it has hard time spinning consistently right after you blow the air then there is problem with the wheal.
TBh. you shouldn't need second pump, you can have it if you have huge system and need for bleeding purpose, or for redundancy, but not for the sake of increasing flow rate from 30l/h upward, single pump should push that with eaz

I took it apart but left it connected to the loop. The wheel appears to spin fine. On testing I noticed that the flow goes up to 16 l/h when horizontal, but only horizontal.

Update.

Flow numbers are correct, therefore there is nothing wrong with the original flow sensor. I installed a low mps flow meter. It gets the same results.

Therefore the problem lies in the loop somehow. The loop layout goes:

Main pump -> CPU 1 - > radiator -> CPU 2 -> reservoir -> quick disconnect.

The loop arrangement should work as the whole line is solid (no voids) until the reservoir.

It cannot be the quick disconnect as the main pump would cavitate if the flow problem was there. I removed all other quick connections. The only other connections are two bulk-head connections.

So that leaves the heat sinks or radiator as the problem.
The radiator is: Alphacool NexXxos ST30
CPU 1 has a: Alphacool XP3
CPU 2 has a: Alphacool Eisbaer which includes a pump I use for a booster (which has no effect).

Main pump as previously mentioned is a D5.

Fundamentally that leaves only the heat sinks with a possible clog, or the Eisbear having a really terrible pressure drop.

I just assume it is the eisbaer cooler causing the problems. If I cared to get more than 20 l/h, I would replace it.