Development : Mayhems Radiators

[mlwood37]

As some of you may know weve been looking into rads for quite some time now.

We sent out what we would call our version 1 Rad to testers a few months ago and sat patently awaiting for the results to come back, just to confirm our own tests. Well now the results are back in from some testers i thought id post some info for you guys.

OFC these results have all been tested on Live real computers and are not in any way done under laboratory conditions how ever the testers whom have had the V1 rads are very happy with them. We how ever have all so had some feed back relating to issue of them being a bit of a nightmare to bleed them so the version 2 will be different and we all so had some other points made that we will be incorporating into the version 2 rads.

I wont be showing any picture of the rads just yet as its quite pointless as version two will look a little different but the same principles over all will be moved across from version 1 to version 2.

Here are what we found with testing and these have been confirmed results with testers as well.

Thermochill 480mm Vs Mayhems 360mm

4 fans on the Themrochill running at 1200 RPM, 3 fans on the Mayhems Rad running at 1200 RPM
A simple ball valve was used to switch between the Thermochill Rad and the Mayhems Rad so nothing was disturbed between switch over.

TC 480mm = 41c After 4 hours of prime
M 360mm = 43c After 4 Hour of Prime

Mayhems 360mm Rad was -2c behind Thermochill's 480mm rad - Test's confirmed from a independent tester who results were 2c behind the same rad.

Thermochill (old style old spacing) 3 x 120mm rads Vs Mayhems 360mm rad

TC 3 x 120mm = 44c
Mayhems 360mm = 41c

Thermochill (new spacing) 360mm Vs Mayhems 360mm

TC 360mm = 46c
Mayhems 360mm = 41c

Black ICE GT Xtreme 360mm Vs Mayhems 360mm rad

Fans running at 1200 RPM. Same again simple ball value in system to switch between rads as not to disturb set up.

Black ICE GT EX 360mm = 45c After 4 hours of prime
Mayhems 360mm rad = 41.5c After 4 hours of prime.

The test difference all takes into account the ambient temps at the time from each tester and are produced from a advantage score.

This was tests done on the Ver 1 rad so really they are irreverent as were working on version 2 to improve the rad some more in the aesthetics department and all so to help bleeding of the system. Some of the changes we are going to incorporate may improve performance or may drop the performance we just don't know yet.

How ever im happy to now move forwards to the version 2 rad.

There will be 4 different types of the rad based around a 240mm version and a 360mm Version. we will not be making a 1 x 120mm version or a 480mm version as there a waste of time. Nor will we be doing a 140mm flavour just yet as again we don't see the point.

I cannot give to much info out just yet except these are made in the UK only and in no way will we have them made out of the UK.

ETA will be round about Spring / Summer 2011. Pricing will be competitive.

[Shteud]

Interesting. wait for more results and fotos

[Avathar77]

Big fan of Mayhems!
But I wish you would do 480s as well!

[JoeM]

ETA will be round about Spring / Summer 2011.

Just guessing you meant 2012 here?

[mlwood37]

Yeh 2012 will be release date.

In our testing there is no pint doing a 480 version as not only are they a waste of space but there is no performance benefit to them. Plus the cost of metal is so high that the cost of production would be high as well. so again pointless.

[Evantaur]

a long way from dyes to rads

fittings next?

[meanmoe]

Very nice... looking forward to more info.

[mlwood37]

Its been some thing ive been wanting to do for a long time. And trust me we dont just do dyes and wont be doing just rads and Fluids as well...

[paulbagz]

Taking pre-orderes? :P

-PB

[pby5cat]

can u put a G1/4 threded fill port on the tank opposite the in/out tanks?

[frank anderson]

Awesome news Mayhem, got your blue and red dye running and they are sweet, I ran it for about 3 weeks 24/7 and drained my system, no staining what so ever (distilled and silver coil only).. will have details on my build log of course

Back on topic.. personal preference really, but these are some things I like to see..

1) Paint not so easily chipped (not sure how HWL does it, but their paints is pretty amazing... Thermochill TA series chip too easily..)
2) inbuild shroud (SR1 thick)
2A) or a thin version (swiftech MCQ)
3) 11 FPI (I think this is the sweet spot for us 800 rpm fan users), 20-30? FPI for those Panaflo/Delta users..
4) G1/4 (standard)
5) G1/4 on oppisite side, for those MM owners that put their rads in a upright position
6) Quality Quality Quality (I would pay extra just for quality)

I dont know when it comes to rads and it's so hard to decide.. they would have to really "shoot themselves in the foot" to make a crappy rad, I have SR1 120, 240 and 360, TA120.3 and a Swiftech 360 MCQ quiet series..
They are all pretty good quality with the exception of a few quirks here and there...

Thermochill came with a lot of gunk inside, salt composite build up on the threads, took me 2 days of cleaning and flushing with vinegar to clean it.. tho this may have just been my luck with the "ship by sea" from my local distributor and not manufacturers fault, the paint chips so easily it's not even funny... Performance is awesome tho.. Unit itself feels like quality and pretty heavy when holding it..

SR1, feels heavy, can feel the quality when holding it in my hand, pretty clean when arrived and hardly any gunk inside..

Swiftech QA is top I think, that res came pristine, I rinsed it and found absolutely nothing inside.. the metal housing that wraps the unit (fins) seems a bit on the light side so the unit feels a bit flimsy, I can feel the rad center buckling a bit when holding it during installation, other then that, a top notch product..

oh yea, I prefer the M4 screw tap !!! :P M3.5 is fine and the self tapping is just annoying when trying to find the right screws for push, push pull, push shroud, etc...

Just sharing some of my experience and thoughts on this topic..

[Church]

frank anderson: i'd advise against built in SR1 shroud, as it 1) limits ways to mount rad being from just one side, 2) is not of most effective depth (1.5cm vs ~3cm), 3) leaves out cases where extra thickness simply will make rad not fit.
+1 for M4 tap.
mlwood37: if you are going to make 140mm rads as well, please don't repeat HWLabs mistake of making their 140mm rads little bit, but too wide to fit in 5.25 bays.

[mlwood37]

were not making 140mm version as i don't see the point.

1) there are no really decent 140mm cases for the mass market or fans for that matter that ive come across.
2) 120mm is established
3) there is very little performance gain

So you all prefer M4 tapped. (ill look into that we were going to do m3 tapping)

the shroud as such will be built in.

We cannot do 4 different fittings if we are to be competitive due to the expense of having them fitted (its not cheap over here in the UK). If we were to have them made in china we could throw all sorts of crap on because it would cost nothing to make but were looking for quality of build and performance not some cheap quick profit, quick turn over rad.

The paint we are using is thin paint that is used on gas heating in fires as it works fine. (how ever it will show though if scratched)

this is our first venture into the rad side of the market and we would like to make a impact from the word go with them so we are tying how ever we have to stay with in reason and costs. We don't want gimmicks or hype just some thing you can hold and go yeh its a good rad and then put in you system and smile. If we can achieve that then we will be happy.

There is very little profit in rads so what we make has to be good else we simply wont do it.

Mick

[Avathar77]

Fittings Next ?

[Vetalar]

waterblocks next???

[Utnorris]

6-32 tap is my preference, but I am in the U.S. and they are easier to get than M4 or M3 screws.

[bds71]

personal opinion: i like the 20 FPI rads - one of the most popular fans is the GT AP15 which runs at max 1800 which is the near perfect spot for 20 FPI rads (assuming 12V operation). and, in P/P the AP15 can be turned down to 5V with still very good operation. less (FPI) than 20 and you lose the high performance crowd. more than that and you lose the "quiet" crowd. just my 2c.

[Annoyingmouse]

were not making 140mm version as i don't see the point.

1) there are no really decent 140mm cases for the mass market or fans for that matter that ive come across.
2) 120mm is established
3) there is very little performance gain

1) The popular Corsair 280 supports 140mm fans, but spacing is needed for the rad because of the hotswap sata bay. Seems like an opportunity for a rad designer to design tanks around that issue.
2) very true
3) Dunno about that. Two extra cm of width is room for two or 3 more tubes, and in the case of a 240 vs 280, the difference in fan-able surface is around 100 square cm. 12cm * 24cm vs 14cm * 28cm.

I think it's at least worthy of consideration. Also, super fine 3m threading is difficult to line of when you're trying to run a screw through a grill, fan, shroud\spacer, and\or case fan mount.

[zeropluszero]

1) pics or it didnt happen
2) skinneelabs review, or the internal and independant testing means approximately what EK's university corrosion testing meant..

[Martinm210]

Was the processor the only heat in the system? What processor was it and what overclock?

FYI, Here is my old test of the original TC 360:


For an approximate 200W processor only heat load I was only getting about a 5-6C delta between air and water at 1200RPM so 3-5C gains is A LOT!... I hope it is, I just haven't seen a gain that large between like sized rads. There must be more heat in there...?

[mlwood37]

@Martin mine was a I7 950 running at 4.2 Ghz on prime runs. Core voltage was 1.42v

@zeropluszero question 1) PSML , Question 2) agreed. We did external testing though via independents because it useful to get feed back .

[Martinm210]

@Martin mine was a I7 950 running at 4.2 Ghz on prime runs. Core voltage was 1.42v

@zeropluszero question 1) PSML , Question 2) agreed. We did external testing though via independents because it useful to get feed back .

Thanks,

FYI, this calculator estimates that processor at about 192 watts at that overclock, so adding another 20 for pump, and you're only looking at about 210 watts.

If a high performing TC120.3 is running at a 5 to 6C water/air delta (at those fan speeds and heat load), you would have to double the radiator performance to see 3 degrees and quadruple it to see 4.5C. Most rads in the 1200RPM range of the same form factor perform within 10-15% of each other. It's not likely to see more than 1-1.5C differences at those fan speeds and heat load unless you are 2X, 3x, 4x radiator size, etc.

You should consider building an environmental chamber if you plan to do a bunch of R&D, that would be the right way to do it. The hobby level testing just isn't controlled enough to split hairs with..

[mlwood37]

We don't have the funding atm to go full blow out testing at our end how ever if you would like to test one of our version 1 rads (it would have to be under NDA) i can do this for you. This is all we can do atm as were raising funding to release these rads and its taken me two years to get this far. The idea has been in place for quite a while (for the rads and other developments i would like to introduce into the community of which i cannot speak about atm).

You can then run the testing your self and see what results you get back to confirm or dissolve our testing which were quite happy to do. How ever as said in the begging statement Ver 1 rads tests are irreverent and are more just to see if we are heading in the right direction. They are a pre production run for testing only and we only had a limited amount of them made (due to costing).

How does this sound to you.

[eXa]

Too bad there wont be any 140mm rad. 140mm is the only option for me, no point in using a smaller 120mm rad when there is plenty of width everywhere to make use of 140mm rads (5 1/4 fits 140mm rads perfectly for instance)

[Martinm210]

How does this sound to you.

Thanks for the offer. I wish I had time to help, but I don't and I'm just a hobby level tester too. I don't have a rad test bench anymore either so I also lack the tools anymore.

For relatively low cost you could build a test bench using a crystalfontz and Dallas one wire probes. This will allow you to log up to 32 sensors with a .2c absolute accuracy. If you can regulate the ambient somewhat and suplly a good contant heat load, it will do pretty good. That's what I was using before I broke it down for parts.