How to tell if water is flowing through the whole loop?

[Carsomyr]

First time trying black zmt. I'm using a pump I know is working (was working on pc before redoinf the water loop config). I added a flow sensor but it shows 0 rpm and 24c Celsius on the water. I hear the pump working, it really soundd as though the water isn't leaving the Reservoir because I hear the water moving around the pump. When I feel the zmt tubes it feels as though water is flowing but I cannot tell for sure.

I followed the instructions for the flow sensor so it should work fine.
Freezemods lsj-znbj is the model number for the flow sensor.

My cpu water block is transparent but since I'm using distilled water plus anti growth/corrosion liquid (also transparent) I can't tell whether the water is running thru the whole loop. I suspect its. What can I do to know for sure?

Heres a quick overview of my current loop just in case. In the video I don't run the pump. Just trying to show you the fittings in case I missed something.

 

[Waldorf]

did you fill right after putting it together?
i tend to "flush" the open loop first (4-5L of distilled water) by filling the res and the return dumping it into a bucket (still good for cleaning other things),
just to make sure there is nothing left from producing the parts etc, and that i have proper flow.

while sensors are nice, i learned that sometimes non-powered stuff is the better solution, my MB has 4 connections for external probes (flow/temp etc),
and im using zero, as the manual type ALWAYS works...

flow indicator

 

[Shihab]

Pinch the tube!

Disclaimer: I have absolutely no experience with PC water cooling loops.
But I know a bit of fluid mechanics. Less flow area -> more speed / pressure -> stronger "feel." Ghetto-style venturi gauges.

Then again, said feel may still be too subtle to notice. The only "sure" way to tell, barring getting an ultrasonic flowmeter (not cheap), is breaking the loop and holding the outlet tube over the reservoir's level, then see if it fluid comes out. :|

A more practical solution: infer that fact from stress testing your machine for an extended duration, and see if the temps are comparable to -or better than- your previous loop.

 

[taka]

Since you don't have a reservoir just get a flow indicator for $10-15

 

[Outback Bronze]

Turn on the System, put it under some load and see if it cools/overheats. You will know if it's not cooling.

 

[Waldorf]

@taka
you mean like i had already posted.
way to get the mssg counter up

 

[taka]

@Waldorf didn't see the link, confused with the signature

 

[Waldorf]

lol. all good, happens.

 

[A Computer Guy]

I added a flow sensor but it shows 0 rpm and 24c Celsius on the water.

Make sure your flow sensor isn't installed backwards.

 

[Carsomyr]

Ok well it turns out the water is indeed flowing. I can see the water level inside the cpu water block.

Two things seem a bit weird:

1. The cpu water block isn't getting filled entirely. It's like 50% filled with water.

2. The pump sometimes seems to be taking in air. It's a bit intermittent. Should I open my fill port at the top of my Reservoir to let the air out of the Reservoir or should I leave it closed off (there's a bit of room at the top of the Reservoir for air to go to anyways

 

[DeathtoGnomes]

Do as suggested, bleed the air.

 

[ir_cow]

Need to get the air out. Tilt the case, shake the case, idk..lots of different methods.

 

[Carsomyr]

Ok so I did some further testing. Now I can see the water climbing up the first tube bend towards the flow sensor. But it seems there's some obstruction somewhere because the D5 vario pump struggles yet the water up there part the flow sensor. When I increase the pump speed it doesn't seem to help.

Plus the flow sensor isn't displaying any rpm which makes sense, because the only way I can get the water into the cpu water block is by playing with the soft tubes a bit to help...

The water loop isn't completing the whole loop. It's getting stopped at the flow sensor and the cpu block.

I'm now strongly considering removing the flow sensor, inspecting and maybe get rid of it all together.

 

[Waldorf]

no, never fill a loop completely, you need air to allow for change from temp changes, you want a pressure reliefe valve, usually from things like goretex, able to let air pass in/out, but not water.

this has nothing to do with properly filling the loop and letting air out, part of the reason i would use an external power adapter to run the pump without having to turn on the pc, and able to move/tilt/rotate it, so the res is at the highest point making it easier to travel there.

 

[A Computer Guy]

Ok so I did some further testing. Now I can see the water climbing up the first tube bend towards the flow sensor. But it seems there's some obstruction somewhere because the D5 vario pump struggles yet the water up there part the flow sensor. When I increase the pump speed it doesn't seem to help.

Plus the flow sensor isn't displaying any rpm which makes sense, because the only way I can get the water into the cpu water block is by playing with the soft tubes a bit to help...

The water loop isn't completing the whole loop. It's getting stopped at the flow sensor and the cpu block.

I'm now strongly considering removing the flow sensor, inspecting and maybe get rid of it all together.

A D5 shouldn't struggle with a loop that small unless it's being restricted, broken, or air is trapped in the pump. It would really help if you had a clear reservoir upstream of your pump to gravity fill the pump chamber and bleed off air pockets at the same time to be sure you aren't running it dry. Also that would help you inspect the water quality for growth and low fluid levels.

For example if you had a simple 3 port tube reservoir above the input for your pump when you fill the reservoir gravity will prime the chamber as air is being displaced. You fill the tube full enough not to spill from the initial pump cavitation when you turn the pump on, then when turning on the pump it will begin to empty the reservoir. Before the tube is empty you turn off the pump to avoid running it dry and fill the reservoir again, this helps ensure you don't get air locked in the pump so each pump action is guaranteed to push water. Repeat until return water enters the reservoir return port enough that the reservoir will not fully empty. Then with a flow established you can top off the tube with water but leaving some air gap in the reservoir for expansion.

I had a small loop at one point in my NR200P. A problem with a small pump/res combo like what you have is the D5 is mighty strong and will spit out and suck in air almost instantly, also causing it to run dry. I had to buffer mine using my other pump/tube reservoir combo to get it filled fast without it taking all day with umteen repeat micro fills. QDC's also helped make it happen.

... i would use an external power adapter to run the pump without having to turn on the pc ...

^^^ This is an important point in combination with making sure the PWM header if applicable is not connected while priming the loop.

 

[Carsomyr]

no, never fill a loop completely, you need air to allow for change from temp changes, you want a pressure reliefe valve, usually from things like goretex, able to let air pass in/out, but not water.

this has nothing to do with properly filling the loop and letting air out, part of the reason i would use an external power adapter to run the pump without having to turn on the pc, and able to move/tilt/rotate it, so the res is at the highest point making it easier to travel there.

I've got a motherboard jumper type power switch I'm able to test without the components in. (or most go them aren't there)

I'm doing the same thing I did with my previous 4 water loops and never an issue like this. It's the same pump, so I don't understand why the water is not fully circulating...

 

[A Computer Guy]

I've got a motherboard jumper type power switch I'm able to test without the components in. (or most go them aren't there)

I'm doing the same thing I did with my previous 4 water loops and never an issue like this. It's the same pump, so I don't understand why the water is not fully circulating...

Did you check if your flow sensor is unidirectional? Do you have it attached backwards?

 

[Carsomyr]

A D5 shouldn't struggle with a loop that small unless it's being restricted, broken, or air is trapped in the pump. It would really help if you had a clear reservoir upstream of your pump to gravity fill the pump chamber and bleed off air pockets at the same time to be sure you aren't running it dry. Also that would help you inspect the water quality for growth and low fluid levels.

For example if you had a simple 3 port tube reservoir above the input for your pump when you fill the reservoir gravity will prime the chamber as air is being displaced. You fill the tube full enough not to spill from the initial pump cavitation when you turn the pump on, then when turning on the pump it will begin to empty the reservoir. Before the tube is empty you turn off the pump to avoid running it dry and fill the reservoir again, this helps ensure you don't get air locked in the pump so each pump action is guaranteed to push water. Repeat until return water enters the reservoir return port enough that the reservoir will not fully empty. Then with a flow established you can top off the tube with water but leaving some air gap in the reservoir for expansion.


^^^ This is an important point in combination with making sure the PWM header if applicable is not connected while priming the loop.

Did you see the video I included? I have such a configuration no... I also have an exhaust fitting capable of releasing air at the top of the radiator... Doesn't seem to help. In my previous water loop I didn't have that... Not even a drain port. It was a used pc I had bought from someone else.

I doubt the pump is defective as it was working in the previous loop 48 hours ago

Did you check if your flow sensor is unidirectional? Do you have it attached backwards?

I followed the instructions. It has an in and an out.

I have the pump outlet connected to the inlet of the flow sensor and the outlet of the flow sensor continues the loop to the cpu water block.

Anyhow, I'll do more testing tomorrow morning and I'll figure it out. I'll remove the flow sensor inspect it, see if I did something wrong... I'll also try to run the loop without it to confirm the loop itself is fine...

 

[A Computer Guy]

Did you see the video I included? I have such a configuration no... I also have an exhaust fitting capable of releasing air at the top of the radiator... Doesn't seem to help. In my previous water loop I didn't have that... Not even a drain port. It was a used pc I had bought from someone else.

I doubt the pump is defective as it was working in the previous loop 48 hours ago


I followed the instructions. It has an in and an out.

I have the pump outlet connected to the inlet of the flow sensor and the outlet of the flow sensor continues the loop to the cpu water block.

Anyhow, I'll do more testing tomorrow morning and I'll figure it out. I'll remove the flow sensor inspect it, see if I did something wrong... I'll also try to run the loop without it to confirm the loop itself is fine...

Did you move the loop from somewhere else? Is it possible you left a tube plug (like a paper towel) left in a tube and forgot about it?

 

[Waldorf]

basic test after putting any loop together:
use a 10L bucket as res, let the loop "open" with intake/exhaust going into the bucket, and if it works you will see the water being "returned" to the bucket.

 

[A Computer Guy]

Did you see the video I included? I have such a configuration no...

Yes (minus a decent size res), but maybe you don't have enough displacement from the fill tube, or at least not anymore. That D5 will suck the water out of the fill tube pretty fast with maybe not enough time to push enough water up that hill on the other side before the pump sucks in a pocket of air. Let's say you are careful and the pump doesn't end up sucking air by the time you turn off your pump. Gravity will back fill your fill tube and you will have to microfill your fill tube and pump repeatedly, might take all day to fill that loop that way when the loop is vertical. make the loop temporarily horizontal it might be easier to fill.

I suspect the pump shot a partial load into your CPU block and you now might have an air block to work against without enough displacement from your fill tube to dislodge it vertically. You could try filing both tubes on each side of the pump then reconnect and see if that provides enough water assuming the pump is ok. But you sill have to be able to bleed off the air somehow while being able to fill it. Not sure how you plan on doing that without a res or bucket.

I doubt the pump is defective as it was working in the previous loop 48 hours ago

You might need to take apart the pump to see if the impeller is blocked if you still can't figure it out.

I followed the instructions. It has an in and an out.

Good!

I have the pump outlet connected to the inlet of the flow sensor and the outlet of the flow sensor continues the loop to the cpu water block.

Anyhow, I'll do more testing tomorrow morning and I'll figure it out. I'll remove the flow sensor inspect it, see if I did something wrong... I'll also try to run the loop without it to confirm the loop itself is fine...

Hopefully you will find the issue.