Guides How To Leak Test a Custom Water-Cooling Loop

Not possible. New W/C loops always make me moist.


All joking aside - great guide. I would like to see you point out the risks around pneumatic testing though. Hydraulic testing is always the preferred method as there is little stored energy, and pneumatic testing is only used as a last resort where hydraulic is impractical, as testing with air creates a lot of stored energy.

I agree that doing it with a Dr Drop is a safer way to do it - but for those who maybe couldn't afford one and grabbed the bike pump out to test their system could end up with parts in more than one place.

 

Hydro-testing is indeed safer, but that's more of a concern for 300-bar cylinders and the like. For a WC loop, if you're hitting even 1 bar ATM then you're already well outside component spec and have probably popped an o-ring, and the volume is so low that a catastrophic blowout would barely register as a light breeze.

 

I vaguely remember a thread on another forum where someone tried using an air compressor.

 

Yeah, sort of. Hydraulic testing is used widely in industry as it’s the safest way to pressure test systems (part of my day job). Pneumatic testing is only used where absolutely necessary, and with strict safety measures.

B-T has written a guide to teach people on how to pressure test, so should inform people properly It’s not a hard, arduous or difficult task to write a couple of lines to make people aware.

I agree that you are out of component spec at those pressures, have you carried out any testing to back up your theory?

That’s not good.

 

Good for the executor, no. Good for the audience, yes.

 

So, we're looking at about 7-10 PSI then for an air pressure test?

 

Aqua Computer state 0.5 Bar (7.2 PSI) as a test pressure, not to exceed 0.6 Bar (8.7 PSI)

 

Cool, cheers

 

Testing first with air is always a lot easier. But it is always smart to follow with an actual leak test because you never know.

 

Presumably a good test pressure would be found by looking up your chosen pump's maximum head pressure, then adding a little extra to avoid passing a marginal setup (e.g. so closing a drain valve doesn't cause a pressure spike that pops out a tube).

 

What does Not .5 mean? I mean I kept hearing (Not Point Five Bars) in the video.

 

Thanks! Haven't heard it said like that before but makes sense.

 

Just a tip I garnered from a lot of experience.. Do not rely on Dr Drop to confirm your loop is sealed. I have had two loops now that held the air pressure for hours, yet once the coolant went in it leaked. Which should not be possible, but some fittings (always rotary) suffer from a flaw. IE if you put any lateral pressure on them they can start leaking.



Air tested perfectly. However I added coolant and this happened.



So, my advice is that you should first pump up your loop. Then, go around and put some gentle pressure on your fittings and hose etc. Give everything a gentle tug and pull it around a bit. Then go back and see if you still have full pressure. As I said, it is usually always caused by rotary fittings. Here is another example, using a 45' fitting. I had air in this thing for 20 hours, then two days later this happened.



Ironically it was the same rig ffs.

 

Finally catching up on my "reading" list - great tutorial thank you

Really interested by the future video reference to making your own version of the Dr. Leak test... Is that still on the horizon or was it dropped due to lack of interest?

 

Just found this:

https://www.overclock.net/forum/61-...-how-correctly-leak-test-your-loop-101-a.html

 

B- also posted it on here

 

Thanks for the guide! A bit less scared now, I’ll hope to try this soon !