Cooling All watercoolers will want to look at this!

During my bordom I was looking at those tests of Gentle Typhoon fans, the site that created them also created this:

Radiator leak test.

Radiator Leak case study.

Watching those was alarming and interesting, so I decided to read the rest of the case study, helped by Google translate (I love the grammar) and the plot thickened.

Basically it appears that several (going by that second video) radiator manufactors have been letting standard drop, and the inlet chamber and outlet chamber are not sealed from each other fully, so the water from the inlet bleeds into the exhause water. The photos on the second page of the case study show this very clearly. So your lovely 360mm radiator maybe only cooling the water by 2 degrees (quoting the figures on the second page of the article).

So, I thought I'd show everyone here, as this is the first I've read about this. And it seems to affect a large array of radiators too. The cheaper ones are the worst offenders, but even TFC and XSPC (which saddens me) appear to be affected.

I'm going to be doing a rebuild of my system next month, so I will test this then and see if my radiator is affected, anyone else wanting to try the test?

All alarming stuff really... Thoughts everyone?

 

(bloody-hell) thats not good there are a few gutted people on them video's i must say hope ur rad's are oky looking out for your water cooling log...

 

I didn't have the patience to watch through the full 13:22 of the second video. Were there any manufacturers whose radiators did not fail the tests?

Nm, saw the text article conclusions.

 

Looks like I'll be buying a Hardware Labs Black Ice rads. I wonder how much it affects Temperatures.

 

only time I have had leaking rads is through being handfisted and when re-threading the holes for larger screws, A bit of plumbers putty . quick re-spray ..jobs a good en

 

This is a different kind of leak...

 

I can't read study stuff as I am in the office its just a bunch of boxes Pete

 

I did explain it above too

 

Lazy bleeders ...guess the QC sticker slapper was just going through the motions huh

 

Well, it's because they leak/pressure test to see if it leaks water externally, but not internally. So to them it has passed QC...

 

+rep for finding this btw

 

never buy a rad made on a Friday... ?

Seriously though, what sort of quality control is that?

 

i used a different testing of quick surface thermometer method while in operation. testing a black ice 120.2. loaded system (core i5 750 @3.8ghz, 1.33v) with occt, average cpu temp via hwmonitor 50c. temp at inlet 49c, temp at opposite endcap 47c, temp between fins 42c at approximate center, temp at outlet 45c. ambient room temp at approximate 21c. center of fins seemed to have the best airflow from my 2 akasa pearl fans, which are quite weak. they have a space of approx 25mm between the fans and radiator as i have it mounted above the fans on ek mounting brackets. i was kind of surprised its effective cooling was only around 5c.
it's possible that the temp rise again on the outlet is due to proximity of inlet rather than internal leaking. i would venture that the few rads that have the inlet and outlet on opposite ends would fair best and avoid the problem all together.

 

Very interesting. I've got mine out ATM so when i flush them out I'm going to be doing some testing.

 

will test tomorrow and make a video on it

 

Cool, see if we can get our own testing going Just looks like it's another test to do as well as leak testing...

The question is, if you get a rad where this happens really badly, is it grounds to RMA it?

 

i would say its grounds for rma under 'defective workmanship' clause. since they failed to properly seal between the inlet and outlet chambers. but if its commonplace as they seem to make it out, how many rma would you be willing to go through?

 

I would say absolutely yes, it is grounds for RMA. To me it's a simple case of, product is not functioning as intended.

It's incredibly unlikely but i would suppose that there is a chance (however small) that this fault could result in damaged hardware. Should that hardware be an SR-2 with a pair of 990x's in which gets damage from excessive heat then i'm pretty sure the rad manufacturer would much rather fork out for a replacement rad than the cost of replacement chips and boards (provided the heat damage could be proven of course).

However from the looks of it some rad manufacturers should be more concerned with this getting some bigger media attention. The way those Phobya's pissed it out immediately is very, very disconcerting

 

How many of you have a temperature sensor for the water? I would think that defects like this would show up with abnormally high water temps?

 

I have one sensor in res 2 and one inline sensor just after the GPU's