FuZion GFX: huge pressure drop?

[MetalZone]

Has anyone noticed from the flow vs pressure drop graphs at D-tek's website that the FuZion GFX induces a huge pressure drop?
the pressure drop is 6.8 PSI @ 2 GPM! and the Storm G4 Rev2 is only 6 PSI @ 2GM. Are the ratings for real or a mistake?
Looking at the design of the block, does the mid nozzle thing sticking out so far (probably close to the top transparent cover) cause a significant restriction?

[serialk11r]

This is odd...that's some pretty bad restriction there.
But I find this hard to believe...we'll need to see some more tests.

[Petra]

I noticed... but those numbers seem off. Anyway, it's something that I intend to look into once the fittings that I need for my new manometer arrive.

[MetalZone]

Precisely. I was wondering if that was a mistake.
the FuZion CPU's pressure drop is only 2.2 PSI @ 2 GPM. that makes the GFX version 3 times more restrictive.

[septim]

typo error? seen a lot of these lately...

[Pete]

Will await Petra or Nick test it out.

What came of them nozles for the cpu block...think they will do some for the gfx as well?

[Petra]

Sorry to dredge up this thread...but...I did some quick & dirty pressure drop tests this evening at 1GPM and 2GPM and I'm showing about 6.4psig @ 2GPM and about 2psig @ 1GPM. Again, this was just a very quick/dirty/rough test to see whether or not the D-Tek numbers were way off or not...and they're not.

This basically means that the FuZion GFX incurs a pressure drop that's more than four times greater than an MCW60 and about 1.7-2 times greater than the MP-1 GPU. :eek:

At the very least, it's really cool to watch bubbles go through the FuZion GFX... :zombie:

[SparkyJJO]

Weird, it doesn't look all that restrictive. And considering what they started with, a CPU block that is a flow king it seems strange that they'd go the completely opposite direction with the GPU block.

Guess I'm glad I went with the MCW60 instead of the fuzion GFX.

[Hard H20]

Alex

Thanks for the results, you saved me from throwing $100.00 away.

[Jedda]

Hmmm
I'd like to understand this a little deeper.
Reserving judgment for now. Although it isn't looking great as a Sli cooling solution.

[SNiiPE_DoGG]

see i knew i saw this somewhere, mcw 60 FTW, i am considering cooling my gx2 and now i know the truth

[Skauneboy]

Damn. After looking around on the forums I was convinced that the MCW60 and D-Tek are basically identical performance wise. :shrug: Good thing I only went for a CPU+GPU loop.

[serialk11r]

Damn then the D-tek must have some pretty phenomenal performance/flow ratio...I'm still wondering why ranker did not see huge CPU temp increases.

[MetalZone]

Goodness gracious. I almost wanted to buy the FuZion GFX coz it looks better than the MCW60. looks like i'm sticking to the MCW60 purely for lower pressure drop sake. I'm gonna have a Storm in the same loop. the FuZion GFX+Storm is gonna be a real flow killer... 6.8+6 = 12.8 PSI @ 2GPM pressure drop... darnnn....

[septim]

must be a design problem on their part....

anyone up for custom modding said fuzion gfx block?

[MetalZone]

I'm guessing one of the factors causing it is that tube sticking out the centre.

[ranker]

Damn then the D-tek must have some pretty phenomenal performance/flow ratio...I'm still wondering why ranker did not see huge CPU temp increases.

I didn't see a huge temp increase at all, but I am running those on a separate loop with their own DDC-2 w/ Petra's.

I'm moving my GTX FuZion's to the future planned MM Horizon R600 build I've been eying. I'm putting back on the EK FC's in and see if there's any difference in temps when the blocks come in. I wish I could strip my MCW60's to retest the temps but they're currently in my work machine on my 7900's and I have no desire to take apart that loop.

[hecktic]

What about in a dedicated single 8800GTX loop ?

Tygon 2075 1/2" ID --> Swiftech Mcres Res. --> Petra's Tech DDCT-01s Custom Delrin DDC Pump Top with MCP355 (pre assembled be petras) --> Thermochill PA 120.3 rad -->Fuzion GFX

(Note that is not the actual order but I know its in a push to pull config)

Should I be worried and not assemble this loop with the Fuzion GFX and use a different block?

I have not bought the block yet but was so close to until I read this thread...

[septim]

you shouldn't see a problem if only single fuzion gfx....

[SparkyJJO]

I'm wondering if it is because of the inlet. The water goes in but then kind of meanders around in that top section before finally getting shoved to the center hole. I wonder if that center nub was trimmed down some if it would reduce the restriction somewhat.

[septim]

well then time to bring out the dremel i guess...

[hecktic]

but it should be safe to buy the block anyways right because a recall may just happen with this block ?

[Petra]

but it should be safe to buy the block anyways right because a recall may just happen with this block ?

A recall? Surely you jest :rolleyes: . There's no 'defect' here to warrant a recall, it just has a rather high pressure drop. A revision may be released sometime in the future but that's about it.

[hecktic]

A recall? Surely you jest :rolleyes: . There's no 'defect' here to warrant a recall, it just has a rather high pressure drop. A revision may be released sometime in the future but that's about it.

oh dam

[serialk11r]

This is very very wierd...
To think of it, D-tek just lost a :banana::banana::banana::banana:load of customers because of this...

[quiet]

I agree serial, I was about a week away from buying this block until I saw this thread. Bummer.

[ranker]

I agree serial, I was about a week away from buying this block until I saw this thread. Bummer.

For people planning to use this without a separate GPU loop, I can see the concern. But with its own loop, I'm not sure what the fuss would be all about. I'm switching out these blocks for some EK FC's, but not due to performance reasons.

[2k7zero]

Basically for a cpu+gpu loop it is recommended to get a MCW60 instead of the Fuzion GFX?

[ranker]

Basically for a cpu+gpu loop it is recommended to get a MCW60 instead of the Fuzion GFX?

In my opinion yes, especially with the pressure drop of the GFX.

[hecktic]

In my opinion yes, especially with the pressure drop of the GFX.

But even in a dedicated GPU loop like lets say the one Im going for, even in that case if the pump were to fail then the card would die very quick should the pressure drop take its toll before the gpu core gets heated up by a failing water cooling loop right?

in other words a dedicated gpu loop is just as risky?

[serialk11r]

But even in a dedicated GPU loop like lets say the one Im going for, even in that case if the pump were to fail then the card would die very quick should the pressure drop take its toll before the gpu core gets heated up by a failing water cooling loop right?

in other words a dedicated gpu loop is just as risky?

Huh? Pressure drop is what we call restriction. Once your pump is gone flow pretty much stops, within seconds. With any block if the pump goes then the water in the block and eventually around it (convection and conduction too) will heat up pretty quick because there is no flow to carry away the heat, and likewise, your card will overheat and perhaps die.

[hecktic]

dam.. hmm anyone ever try modifying a Swiftech MCW60-T with the G80 adapter and then reapplying the neoprene on an 8800GTS/GTX ?

[IanY]

Its no concern for systems with powerful pumps.

[serialk11r]

Its no concern for systems with powerful pumps.

Lol I mean your flow is limited to *just* 2.5 gpm or so with 4 DDC+s :D Actually 2 of these in a loop means big big trouble, if you're not IanY. Of course that's if the pressure drop really is that terrible.

[hecktic]

In my opinion yes, especially with the pressure drop of the GFX.

Its no concern for systems with powerful pumps.

MCP655 in the powerful catagory ? :shrug:

[MaxxxRacer]

haven't been on the forums in a week (yes that is a record since i joined). Its pretty shocking how high the pressure drop is on this block.

Looks like MCW60 is the way to go for those large core GPU's (yes, i know they are almost all huge)

[quiet]

Just for the sake of this thread, what's the pressure drop on say an EK FC 8800 block? To compare

[Petra]

Just for the sake of this thread, what's the pressure drop on say an EK FC 8800 block? To compare

Hasn't been tested, that I know of... However, I plan on doing a bunch of pressure drop testing when I get a chance.

[Jedda]

This is very very wierd...
To think of it, D-tek just lost a :banana::banana::banana::banana:load of customers because of this...

More cudos to Alex for posting that while holding stock, knowing full well that people will misconstrue the implications and stampede away from the block.
So far IanY has the only post to focus on the pump block combination and their relevance.