The Phobya CPU waterblock

[MaddHawk]

No No,

That guy is way smarter than me

andyc

Hmm gotcha, I am very much interested in your direct die cooling concept though.

[MaddHawk]

What would be really sweet is AMD and Intel building an integral array of pins or fins and seating into the IHS for building a water block into it. Just buy the top of your choice, bolt it onto the CPU IHS and gtg.

[saxile]

My test runs, please keep in mind this is a mini review and by no means a full out test. I'm no Skinnee, just a normal overclocker trying to get the best temps and look out for my brother's and sister's in arms out there.

Also speaking of Skinnee, I ordered two of these blocks and when the guy they sent swimming from Germany two weeks ago gets here with the other block it's straight in the main to Skinnee for full testing.

Waterblock test runs:

Idle:



Hmm, phobya keeps my idle near water temps.

Intel Burn Test:



Phobya bests my D-Tek Quad by 1c, not much but it is beat. One thing I must note my water temps went way up with this test compaired to the D-Tek, this thing transfers heat into the water a lot faster than the D-Tek.

Core Damage Test:



Again, Phobya comes out 1 degree cooler, and again a massive spike in water temps compaired to D-Tek.


My temps are overall higher with the Phoyba, but the Phoyba is much better at transfering heat due to it's thin base. I used same setup for all tests, and it is as follows:

Rad: Feser 360
Pumps: Dual MCP355 with EK-Dual Turbo Top
Video Block: EK 3870x2 FC
NB Block: Blackops Stock with delrin top.
Coolant: Straight Distilled
Fans: 3x Yate-Loon High Speeds running roughly 1600rpm pulling fresh air thru rad.

I believe if I cranked up the fans or slapped my San Aces on it to keep temps a little lower, Phoyba is a force to reckon with.

[Locke]

Did you do enough mounts so that you can be sure its not a difference in the TIM? 1c isn't a lot so its interesting to see it right in there.

[saxile]

As I said just a quick mini-review the d-tek quad was remounted about 30 times in the past 4 months with pretty much exact same temps (use pro mount too btw), phobya this is the first mount, need more goop if I wanna remove it lol.

[BringerOdeath]

If this block dumps more heat into the water, wouldn't using a better rad and a more powerful pump lower temps even more??

[Eddie3dfx]

If this block dumps more heat into the water, wouldn't using a better rad and a more powerful pump lower temps even more??

Even if the cpu block was not better, a more powerful pump and specifically a better radiator would result in better temps.
I added a 2nd pump to the cpu and the results were the exact same.

[saxile]

If this block dumps more heat into the water, wouldn't using a better rad and a more powerful pump lower temps even more??

pumping wise, no the pump would make little difference. Airflow thru rad, yep, but for fairness I ran fan speeds at the same levels, if I ramped up my fan speeds temps will be lower and be even lower if I slap my san aces on which I may try this weekend. Something else I wan to try is bare die cooling, this block is perfectly flat.

[Hondacity]

45c and 42c...wow thats a huge improvement...are they going to retail in the USA?

[saxile]

45c and 42c...wow thats a huge improvement...are they going to retail in the USA?

I'm guessing after Skinnee runs his tests and depending on how it runs for him Petra, Sidewinder, Jab, etc may start carrying them. But the way highflow talks it's hard for them to keep em in stock currently.

[marcus000]

hehe, just got one to play with!

[Pedalmonkey]

Block looks about as close to Mcoffeys direct Die cooling as you could get. Looks like it would need a powerful pump too. lets see some testing!!!!

[headala]

Cool, we look forward to seeing your results! Could you post a few internal pics? The ones at the beginning of this thread are too small to see the flow route.

[marcus000]

Internal Pictures

[skinnee]

Good pics Marcus!

Did you have any difficulty getting that o-ring back in? I struggled with that thing for a good 20 minutes after disassembling the block for pictures.

[anzial]

highly restrictive, me thinks.

[HESmelaugh]

Great pics, marcus!

[skinnee]

Great stuff guys

One question. With a base that thin, do you see any problem with warping if to much mounting pressure is used? Or does the mid plate allign with the base in such a way as to keep that from happening?

Also, how thinn is that base plate?

Thanks for the post and testing, very very interesting stuff.

andyc

according to my calipers base plate is just shy of 1mm. the block construction is interesting/different from what I have seen from other companies. I would be surprised to see the plate warp with the stock mounting hardware. There is a slight center aligned bow to the base.

[Broken]

You know, that is a sharp looking block. Ok, where's the clear top version?

[headala]

Highly parallel design, just like IBM's internal design. But I worry it would be prone to jets clogging.

[skinnee]

according to my calipers base plate is just shy of 2mm. the block construction is interesting/different from what I have seen from other companies. I would be surprised to see the plate warp with the stock mounting hardware. There is a slight center aligned bow to the base.

Thanks Skinnee,

I agree, very cool design. Can't wait to get my hands on one.

andyc

I completely misspoke, baseplate is just over 1mm thickness. I will change my original post. I knew I should have waited till I got home and checked my notes.

[Vapor]

Do want

No VAT for USA buyers, right?

[headala]

Gosh I wish there was more info about this block.

[saxile]

No VAT Vapor, I should know, I bought mine and Skinnee's lol, I love it and find it no more restrictive than the d-tek v1 with the quad split nozzle, course I'm not measuring flowrate with tools just how it passes through my res.

[ILikeCosmosS]

this block is really different i dont think it might succeed it migh just restrict flow