Radiator materials used vs claimed

continuum

Ars Legatus Legionis
94,897
Moderator

Or Tom’s Hardware’s summary:

Note: strictly a materials analysis here by Igor’s Labs.
Veteran hardware reviewer Igor Wallossek has published an investigation into AiO CPU cooler radiator materials. Something must have got Wallossek’s spider senses tingling because his investigation into six branded designs uncovered three that were over-spec'd. By that, we mean half of the AiO brands sold radiators using different, cheaper materials than mentioned in their product data sheets, websites, or advertising. The most common switcheroo was advertising copper but instead using a copper/zinc alloy like brass.
 

steelghost

Ars Praefectus
4,975
Subscriptor++
Feeling very slightly smug about this, as my two custom loop systems here have respectively one Alphacool radiator, and two Hardware Labs radiators, both of which come away with a decent verdict. The Bykski one is a bit of a horror show though! As well as the ones that contained "alkali residue" - shows the value of a good flush out before use.

The EKWB one is really disappointing, considering how much they charge for their stuff. At least it appears if you pay the Aquacomputer tax, you really do get top quality gear.
 
  • Like
Reactions: Jeff3F

m0nckywrench

Ars Tribunus Angusticlavius
6,852
Subcontractors have every reason to substitute cheaper metals because pennies per unit is real money in volume. They know many customers really buy for style (comparable to caring how your toilet tank valve looks) not function.

BTW if your project can use a common brass/copper vehicle heater core (the aftermarket still make them for common old cars and trucks, including custom builds) you can have as much radiator surface area as you like for small money. Where appearance is important the automotive aftermarket sell oil and transmission coolers designed for much more demanding use cases.

Pure copper radiators permit smaller radiators but are heinously expensive and that price is not driven by the cost of copper but the usual markups for boutique products. If I do another liquid-cooled desktop I won't worry about the difference between pure copper and brass but instead find the best copper/brass old style aftermarket heater core with connection points where I want them.

If necessary modding inlet and outlet connections by soldering is not difficult on a new uncorroded heater core to include adding hose barbs or threaded bungs of your choice.
 

Paladin

Ars Legatus Legionis
32,552
Subscriptor
Interesting but not super surprising. Though, I would have assumed the companies selling the devices would have their own QA checks on random samples. If that were my company, I certainly would if I were using outsourced manufacturing.

I notice that they are all 120mm units so I wonder if there might be any difference for the more expensive, larger units? Doubtful, but I guess you never know until you cut it open and find out.

The bottom line is probably that the warranty (and actual customer support reputation) is the important part. You won't know what you got until you cut it open so if you encounter any problems, the warranty coverage needs to be good. Otherwise the price better be really good.

Now they need to check the heatsinks on tower coolers and stuff to make sure those are actually as claimed too.
 

steelghost

Ars Praefectus
4,975
Subscriptor++
I notice that they are all 120mm units so I wonder if there might be any difference for the more expensive, larger units? Doubtful, but I guess you never know until you cut it open and find out.
I suspect if you're running a factory making these things, you're using the same basic stock materials for everything, you're just varying the lengths and widths - as such, I don't imagine the different sizes would be any better (assuming both produced at the same time).
I would have assumed the companies selling the devices would have their own QA checks on random samples. If that were my company, I certainly would if I were using outsourced manufacturing.
Yep, this is where Igor dings some of the culprits pretty hard. EKWB don't come off at all well here IMO.
Pure copper radiators permit smaller radiators but are heinously expensive and that price is not driven by the cost of copper but the usual markups for boutique products.
Pricing varies region by region, where I am Alphacool are often at the cheaper end despite being the only 100% copper option.
Now they need to check the heatsinks on tower coolers and stuff to make sure those are actually as claimed too.

View: https://www.youtube.com/watch?v=VmOY1c0Rkno

Slightly difference as it's more about manufacturing process rather than materials, but the same thing applies.
 

Nevarre

Ars Legatus Legionis
24,110

mpat

Ars Praefectus
5,951
Subscriptor
Barrow isn't a brand sold in N. America to my knowledge, but the EU/German distributor Caseking has pulled them from the market entirely and is accepting returns of the affected model no matter how long ago they were purchased.

70% lead in solder? Yeah, those guys better have some good lawyers on retainer and all the documents in order. It is possible to be able to sell part with lead in solder today, but you need to be in one of the areas covered by an exemption.
 

Nevarre

Ars Legatus Legionis
24,110
The EU is strict on ROHS. I have to assume that the Chinese OEM fabricated their compliance and the EU distributor Caseking was operating on the basis that the products were what they were represented to be. It's very possible that the initial product run shown to Caseking was made from compliant solder and then once they had a contract, it's possible Barrow cheaped out at that point. I don't know if Caseking holds some culpability here or not, but taking steps to return all the products they can for refund seems like a good first step for fixing the situation.

Even Barrow's English site is pretty rough: http://www.barrowint.com/index.html

IME, German companies are a little less incentiveized to treat Mainland Chinese suppliers with suspicion but I'm sure this will cost Caseking a lot of money regardless of whether or not they knew or suspected the products might be non-compliant to ROHS.
 

sixstringedthing

Ars Scholae Palatinae
743
Subscriptor
I've got a "fairly expensive in AUD+shipping at the time" Phobya 280 rad that performed brilliantly on my old X58/Xeon rig and is thoughtfully designed in terms of input/output options. So thoughtfully designed, they even thought to solder small lengths of what I assume is narrow-gauge steel or copper wire to the tops of the flat tubes in the area underneath the screw holes, to act as a buffer so you don't puncture your tubes during a Dumbass Moment.

The only problem with this is that they used tin/lead solder to affix steel or copper wire to brass tubes with aluminium fins attached, all in direct contact. Not a chemist so I dunno what kind of funky ionic reactions are going on between that lot, but there's some interestingly-coloured oxidation products in all four corners on both sides of the rad.

It holds lung pressure, but still... I don't trust it.
 
Last edited: