Leak testing

[ChrisZH]

So for the past day I have been trying to figure out where the point of failure in my loop is. I am using ek leak tester and it shows I have a very minor pressure drop. When I test the specific areas of the distro plate there is no pressure drop, once I put the tubes in is when it drops. I'm talking like a mm drop from .5 bar to .45 bar. I have no idea where the point of failure is. Should I add water to the loop and see where it leaks out from?

[ChrisZH]

I have been leak testing in segments as well. Block a certain Part of distro plate off and test. Idk if it could be improper contact of fittings though I have reseated them a couple times.

[Denned]

Does it continue to lose air over time, or is it just that single small drop?

[ChrisZH]

59 minutes ago, Denned said:

Does it continue to lose air over time, or is it just that single small drop?

From what I can tell it's a very very very gradual drop. Like happens after 3-4 hours of testing. It's so small you wouldn't really notice it if I didn't take pictures of it. I'm probably going to have to take it all apart and make sure my tubes are sanded and making good contact again. And make sure my distro plate is sealed completely and my CPU block. But that's why I was thinking of water. To get the exact location of the leak.

[ChrisZH]

All the tubes fit very snugly in the fittings. They are all tightened down to the point it won't go any more. I make sure that they were exact size needed by trial and error. They are all level, so it's just really confusing.

[Denned]

Try to wet your orings before installing both tubes and the fittings themselves.

What happens if you test it with 0,75 or even 1 bar pressure?

[ChrisZH]

2 hours ago, Denned said:

Try to wet your orings before installing both tubes and the fittings themselves.

What happens if you test it with 0,75 or even 1 bar pressure?

The same thing happens with 0,75. It's so minor you wouldn't notice it if you weren't paying attention. I'll try to wet the oring when I get off work and see what happens. Only other thing I can think of is a tube is not long enough or making good enough contact. Wish I knew where the minor leak was coming from... I could fix it then.

[ChrisZH]

2 hours ago, Denned said:

Try to wet your orings before installing both tubes and the fittings themselves.

What happens if you test it with 0,75 or even 1 bar pressure?

What would wetting the o-rings do?

[ChrisZH]

Tubes are supposed to be a tight fit right? I figured a tight fit was better than loose.

[ChrisZH]

Anyone got any advice please? I'm lost with this now..

[For Science!]

1 minute ago, ChrisZH said:

Anyone got any advice please? I'm lost with this now..

You aren't using EK fittings and tubes right? then the EK Leak tester is kind of dumb anyway since you have no idea what your fittings and tubings are rated for. Just do it the classical way of filling it without powering on the system and tissue. This fad of testing with air is pointless imo.

[ChrisZH]

2 minutes ago, For Science! said:

You aren't using EK fittings and tubes right? then the EK Leak tester is kind of dumb anyway since you have no idea what your fittings and tubings are rated for. Just do it the classical way of filling it without powering on the system and tissue. This fad of testing with air is pointless imo.

Well but shouldn't a loop be airtight? So shouldn't this still have a purpose? What kind of pressure for a loop get anyways? But how is it pointless if I'm not using ek products? Don't all do the same thing?

[PowerBaller]

The way how i use that pressure tester is:

1. Test the water block and radiator prior to the installation, in order to rule out faulty parts.

2. Before plugging the reservoir in to your loop, enclose your open loop with stop plugs after leaving one port open for the air pump, test the leak,

to ensure the critical part of you loop isn't leaking, ( the cpu block, vrm block, top rad, gpu block ) ;apply high pressure (.75 bar in the green zone).

3. Isolate the distro plate and test the plate itself with high pressure.

 

The leak test itself: if the air pressure did not drop a bit after 10 min, at the highest pressure in your green zone, it pass the leak test.

 

Now you can connect the reservoir. Even if it leaks. you have 95% confidence to tell that it is the junction between your pump and your distro plate is leaking, and the liquid wont drip on your electronic components, usually.

[ChrisZH]

1 minute ago, PowerBaller said:

The way how i use that pressure tester is:

1. Test the water block and radiator prior to the installation, in order to rule out faulty parts.

2. Before plugging the reservoir in to your loop, enclose your open loop with stop plugs after leaving one port open for the air pump, test the leak,

to ensure the critical part of you loop isn't leaking, ( the cpu block, vrm block, top rad, gpu block ) ;apply high pressure (.75 bar in the green zone).

3. Isolate the distro plate and test the plate itself with high pressure.

 

The leak test itself: if the air pressure did not drop a bit after 10 min, at the highest pressure in your green zone, it pass the leak test.

 

Now you can connect the reservoir. Even if it leaks. you have 95% confidence to tell that it is the junction between your pump and your distro plate is leaking, and the liquid wont drip on your electronic components, usually.

I think I'm going to take my new CPU block and distro plate off and make sure the screws are tight enough. Because I know all my fitting are tight and there is a solid connection between tubes and fittings. Like they won't budge at all.

[For Science!]

12 minutes ago, ChrisZH said:

Well but shouldn't a loop be airtight? So shouldn't this still have a purpose? What kind of pressure for a loop get anyways? But how is it pointless if I'm not using ek products? Don't all do the same thing?

A loop has little to no pressure since water doesn't compress, any pressure produced by the pump is immediately relieved by the fact that the water that was moved is replaced by more water. Think about an aquarium, just because you seal the top of an aquarium tank, a running pump isn't going to cause an explosion, the water moved is replaced by water.

 

Taken from EK themselves:

 

Quote

Keep in mind that the pre-defined safe maximum pressure defined on the gauge is for EK products only. EK cannot guarantee for non-EK products failing under the predefined safe zone marked on the gauge.

 

[ChrisZH]

56 minutes ago, For Science! said:

A loop has little to no pressure since water doesn't compress, any pressure produced by the pump is immediately relieved by the fact that the water that was moved is replaced by more water. Think about an aquarium, just because you seal the top of an aquarium tank, a running pump isn't going to cause an explosion, the water moved is replaced by water.

 

Taken from EK themselves:

 

 

So my distro plates back screws had some more in them to turn. I'm testing each cavity of the plate now for 15min to see how the seal is. I tested CPU block and it's sealed fine. Gpu block is good. I had two tubes that could be a little longer so I went ahead and made a little longer tubes. Hopefully it's figured out now, but only time will tell

 

[Denned]

55 minutes ago, For Science! said:

You aren't using EK fittings and tubes right? then the EK Leak tester is kind of dumb anyway since you have no idea what your fittings and tubings are rated for. Just do it the classical way of filling it without powering on the system and tissue. This fad of testing with air is pointless imo.

That's completely irrelevant. You're missing the point of doing it with air.

We're not testing a loop with air because 'part are rated for x amount of pressure.' We are testing with high pressure to make sure the loop is completely air tight. And higher pressure is just going to make it easier to see if it's leaking.

If you manage to create a loop where air wont escape, then you can be 99,9% certain that water wont escape. If you're just throwing a loop together, then you can't actually be sure that there's no slow leaks. But with air if it isn't leaking, then you can leak test a system in like 15 minutes. Instead of running the loop for x amount of hours while crossing your fingers.

 

4 hours ago, ChrisZH said:

What would wetting the o-rings do?

It's going to make sure the o-ring sets in correctly.

You can also wiggle about the tubing to see if that causes the pressure to drop.

Or use plugs to isolate areas of the loop if your blocks turn out not to be leaking.

[ChrisZH]

9 minutes ago, Denned said:

That's completely irrelevant. You're missing the point of doing it with air.

We're not testing a loop with air because 'part are rated for x amount of pressure.' We are testing with high pressure to make sure the loop is completely air tight. And higher pressure is just going to make it easier to see if it's leaking.

If you manage to create a loop where air wont escape, then you can be 99,9% certain that water wont escape. If you're just throwing a loop together, then you can't actually be sure that there's no slow leaks. But with air if it isn't leaking, then you can leak test a system in like 15 minutes. Instead of running the loop for x amount of hours while crossing your fingers.

 

It's going to make sure the o-ring sets in correctly.

You can also wiggle about the tubing to see if that causes the pressure to drop.

Or use plugs to isolate areas of the loop if your blocks turn out not to be leaking.

Right now I'm testing each cavity of my distroplate individually. The screws on the back had another turn or so in them. So I'm wondering if that was causing bad contact? Also I got two tubes, I made them a tiny bit longer.

[PowerBaller]

1 hour ago, ChrisZH said:

I think I'm going to take my new CPU block and distro plate off and make sure the screws are tight enough. Because I know all my fitting are tight and there is a solid connection between tubes and fittings. Like they won't budge at all.

Be ware of overtightness on acrylic surface; it is notorious of being susceptible against high torque pressure that leads to cracking

[ChrisZH]

1 hour ago, PowerBaller said:

Be ware of overtightness on acrylic surface; it is notorious of being susceptible against high torque pressure that leads to cracking

Yea I just tighten everything until it won't anymore. Im not gonna force any with that because I don't want microfractures 

[Mick Naughty]

Put fluid is. Isn’t gonna cause an issue of it leaks. 

[LOLZR]

Air molecules are smaller than water molecules, even if there is some air loss throughout the system doesn't mean that it is going to leak.

[ChrisZH]

Man it took me a while to find this out. One if my tight tubes was a little too short, wasn't fully making contact. I had to unscrew the distroplate and turn it in order to get the tube in the fitting. Whew, currently leak testing this part. So far everything has been leak free.

[ChrisZH]

This bend was hard to figure out for my first time..

[Tonberry]

Air is less viscous and smaller particles than water, so water gets more difficulties to pass through a micro gap which air can pass through it.

 

I dry tested my loop with it 4x, and I got almost 1 strip drop on the indicator after 2-3 hours on the leak tester when I air pressured my loop. 1 strip is veeeeeeery small delta pressure that water cant pass through it while air itself which less viscous and smaller particles need 3 hr to lose that tiny amount of pressure.

 

You'll be fine.