Storm/G7 Prototype + Piccies

Cathar · 09-13-2005, 07:51 PM

In final testing phase now of my "ultimate" waterblock. Seriously considering stopping making CPU waterblocks after this as its getting kinda "out there" in terms of cost vs performance gains.

Anyway here's the pictures:

http://www.employees.org/~slf/g7/

Got some work to do now, will explain the pictures as the thread develops.

What's changed you ask?

Just to recap what the G-numbers mean 'cos there's still some confusion over this:

G1 = 1 jet
G2 = 7 jets
G3 = 19 jets
G4 = 35-37 jets (implemented G4 has 35 jets)
G5 = 55-61 jets (implemented G5 has 59 jets)
G6 = 85-91 jets
G7 = 117-127 jets (implemented G7 has 117 jets)
G8 = 155-169 jets (could only machine on aerospace quality equipment)

As for the design improvements over the original base Storm design specification:

Increased jet inlet plenum area to lower pressure drop (thicker top-plate enables a more voluminous plenum so water can reach all jet inlets with minimal restriction - visible to eye))
Ported jet intakes to lower pressure drop (jet intakes are now all ellipse radiused using a custom manufactured tool - visible to eye)
Tweaked jet lengths to lower pressure drop (jet lengths are shorter - internal + external mod, middle plate is significantly thinner too)
Tweaked jet outlets to lower pressure drop (each jet outlet is chamfered around the outside of the jet tube to offer less restriction for flow exiting the cup structure - barely visible to eye)
Multiple flow path channels to outlet to both decrease pressure drop, and more evenly distribute jet flow and total base-plate cooling (self-explanatory and very visible to eye)
Base-plate bracing to boost structural strength for high-pressure CPU mounting (the 12 little posts that surround the jet area brace the base-plate and also serve as additional water turbulation for water outwash cooling effect over base-plate outside of main jetted area)
Improved top-plate strength for higher pressure mounting capability (just a thicker top-plate, which is offset by thinner middle plate)
Middle-plate bracing to improve jet/cup alignment when used with high-pressure pumps (additional consequence of the middle plate posts - reduces middle plate flexing meaning jets stay better focused with respect to their cup counterparts under very high pressure operation - meaning >>10psi)
Improved O-ring specifications to minimise assembly based warping (smaller O-rings and slightly wider O-ring groove dimensions, thereby meaning less clamping pressure required to seal minimising the amount of pressure that gets applied to the base-plate effectively eliminating base-plate flex on assembly)
Additional O-ring seal between inlet to middle plate plenum and outlet (this one's for the armchair waterblock critics - I never deemed it necessary but people often griped about the design needing one even though it doesn't, so I put it there to keep them happy anyway - may as well)
Improved cup/wall ratios for higher performance (minor ratio adjustments - added base-plate strength too)
Improved jet turbulators for higher performance (two prototypes being tested)
Improved base-plate characteristics for higher performance (adjusted base-plate thickness suitable for G7+silver)
Improved jet/cup ratios for higher performance (tweaked ratio adjustment)
Improved low-flow performance without sacrificing high-flow performance (micro-jetted efficiencies are much higher even with lower flow rates - also a consequence of turbulator modifications and jet/cup ratio adjustments)
Improved small bare-die performance (down to 7x7mm CPU dies) without sacrificing IHS performance. (simulated and tested adjustments to attempt to cover the widest possible variety of CPU types)

There's about 12 month's worth of lessons in this block and I've spared nothing to develop this block that physically embodies everything I've learned.

jaydee · 09-13-2005, 08:49 PM

Sounds and looks pretty good!

DryFire · 09-13-2005, 09:43 PM

If this was made would a copper version be considered? (seeing as teh G5 was only silver)

Cathar · 09-13-2005, 10:55 PM

Quote:

Originally Posted by DryFire

If this was made would a copper version be considered? (seeing as teh G5 was only silver)

If people want a copper block version of the Storm, go buy the Swiftech version.

At this level of machining, the silver is merely a minor part of the overall expense of the block.

Sin22 · 09-13-2005, 11:07 PM

Very nice stuff mate.

Interested about the shortening of the jet tubes and the ellipse of the jet intakes. Does the elipse go through the whole tube or is it kept at only the intake area.

Cathar · 09-13-2005, 11:18 PM

Quote:

Originally Posted by Sin22

Interested about the shortening of the jet tubes and the ellipse of the jet intakes. Does the elipse go through the whole tube or is it kept at only the intake area.

Extends part-way down into the tube proper.

ricecrispi · 09-13-2005, 11:48 PM

Do you think about having larger porting at the centermost inlets to focus more flow into the center cups would improve performance or is it negligible?

CAN we also get more pics of the mid plate?

starbuck3733t · 09-14-2005, 09:09 AM

I MUST HAVE ONE!!!

No. Seriously. You've got PM.

Nugit · 09-14-2005, 10:51 AM

Magnificent, the details on that block. Wow. What a ripper.. I'm speechless.

What's machining time for one of those beautities?

Breach · 09-14-2005, 11:28 AM

looks like a sweet block Cathar, you have obviously done all of your homework and thensome for this one.

Starbuck how did I know you would be first to raise your hand

starbuck3733t · 09-14-2005, 11:56 AM

Quote:

Originally Posted by Breach

looks like a sweet block Cathar, you have obviously done all of your homework and thensome for this one.

Starbuck how did I know you would be first to raise your hand

Cuz I'm typical

bigben2k · 09-14-2005, 02:12 PM

Nice work.

I like the new outlet scheme (inner channel).

ricecrispi · 09-14-2005, 02:57 PM

Quote:

Originally Posted by Cathar

In final testing phase now of my "ultimate" waterblock. Seriously considering stopping making CPU waterblocks after this.

"Ultimate" is a complete understatement. More of GODLIKE would appropriate.

Do you have any prelime test numbers for us?

iakovl · 09-14-2005, 04:00 PM

silver

so much

Cathar, i knew u r a WC ace
now i know, i was wrong

u r a WC god!!!

tong · 09-14-2005, 04:53 PM

cathar when are you making these available?

murray13 · 09-15-2005, 11:50 AM

Hi Stew, the 'Perfect' Storm. Do you know what the projected or actual pressure drop is like?

Holst · 09-15-2005, 01:50 PM

I dont suppose you have pics of the other blocks for comparison.

Especially G1, 2 and 3

killernoodle · 09-15-2005, 02:02 PM

That machining is incredible. What is the ID of those jets?

mwolfman · 09-15-2005, 03:52 PM

Well done Cathar!

Some questions...
What’s the total production time (Silver and Delrin). I guess the Delrin manufacturing time is about three times the silver time?

What’s the pressure drop? (IE that pump is required to run a G7 CPU AND G7 GPU block inline)

Do you have to machine (mill) the silver block (silver = soft and the pattern is relatively easy to press), what’s the depth on the hols, and diameter (in mm please) any angel (like a upside cone)?

What type of filter is needed so there aren’t any hold-ups in the jets? (I clean my system ones a week with air bubbles...)

dacooltech · 09-16-2005, 04:49 PM

excellent job Stew

I especially loved the Wye connection integrated to the top plate... there are 2 holes in the mid plate too if i'm not mistaking... positioned above the center of the channels on the BP sucking coolant from the channels (some kind of Venturi effect perhaps? ) ... very nice

I think that integrated Wye idea can be adapted to the WW as well, don't you think?

dacooltech · 09-16-2005, 05:26 PM

Quote:

Originally Posted by mwolfman

What type of filter is needed so there aren’t any hold-ups in the jets? (I clean my system ones a week with air bubbles...)

well looks like the block itself is the filter j/k

Cathar · 09-16-2005, 10:40 PM

Quote:

Originally Posted by killernoodle

That machining is incredible. What is the ID of those jets?

0.45mm

Quote:

Originally Posted by mwolfman

What’s the total production time (Silver and Delrin). I guess the Delrin manufacturing time is about three times the silver time?

Total production time is a shade under 3hrs all up (about 2.5hrs CNC work).

Silver CNC time is about bang on 1hr.

Quote:

Originally Posted by mwolfman

What’s the pressure drop? (IE that pump is required to run a G7 CPU AND G7 GPU block inline)

Measured to have a slightly lower pressure drop than my original Storm/G4's.

Quote:

Originally Posted by mwolfman

Do you have to machine (mill) the silver block (silver = soft and the pattern is relatively easy to press), what’s the depth on the hols, and diameter (in mm please) any angel (like a upside cone)?

I would say that the pattern in this block would be extremely difficult to press reliably. As for the other info, no comment.

Quote:

Originally Posted by mwolfman

What type of filter is needed so there aren’t any hold-ups in the jets? (I clean my system ones a week with air bubbles...)

Either ensure that your water & system is clean when you fill it, or alternately use a cheap micro-irrigation filter if you're really worried.

Quote:

Originally Posted by dacooltech

I especially loved the Wye connection integrated to the top plate... there are 2 holes in the mid plate too if i'm not mistaking... positioned above the center of the channels on the BP sucking coolant from the channels (some kind of Venturi effect perhaps? )

Correct on all accounts. I did my best to ensure that the block allowed as free flowing of an exit path for the water as possible, as well as eliminating any and all calls for a possible third barb.

mwolfman · 09-17-2005, 03:44 PM

Quote:

Originally Posted by Cathar

0.45mm.

cool, I’m very happy that I’m not drilling that...

Quote:

Originally Posted by Cathar

Total production time is a shade under 3hrs all up (about 2.5hrs CNC work).

With the exultation for the 0.45mm holes (I might figure that out as well), I can’t see any reason for using the mill when (IF) there series production of the block (just melt press the block and then make the holes with water jets).

Quote:

Originally Posted by Cathar

Silver CNC time is about bang on 1hr.

Ouch, is silver worse than copper to mill? (I made 77*5mm*2mm holes and the rest of the copper block in 20 min, in copper). Cant silver be casted easily?

Quote:

Originally Posted by Cathar

Measured to have a slightly lower pressure drop than my original Storm/G4's.

Nice!

Quote:

Originally Posted by Cathar

I would say that the pattern in this block would be extremely difficult to press reliably.

Depends on the temperature of the silver.

Quote:

Originally Posted by Cathar

As for the other info, no comment.

I guessed that…

Bundles · 09-17-2005, 06:19 PM

They say there is a very, very, VERY fine line between being a genius and being insane, you Mr Cathar good sir, are perched very shakily on that line in my opinion.

However, if it wasn't for the likes of yourself, watercooling would not be where it is today, have the reputation it has today, and just generaly be so enjoyable and interesting as it is.

I just can't see you hanging up your hose though, there would always be those "Hmm, i wonder if....?" and them "Now I think if I...." moments.

AS for the G7, if i could afford one, then i would take one for certain, can waterblocks get much better than this? I think not.

Cathar · 09-17-2005, 07:52 PM

Quote:

Originally Posted by Bundles

can waterblocks get much better than this? I think not.

Have gathered some results with the G7 vs the G5, and am busy constructing such a post. The G7 is NOT a quantum leap forwards at the higher pumping powered end, nor was I ever expecting it to be such. The G7 represents my version of a near completely "flow agnostic" block (within reasonable limits of course).

Can it get better, yes, very very slightly, but only by specifically tuning for very high-end pumping powers, but for a while now I've realised that water-cooling's back is against the wall with respect to seeing further improvements and with the G7 I think I'm about as close as one can get excluding optimising for specific pumping powers.

I'll explain more when I finish my write-up.

Oh, and I agree - this is pretty close to sheer madness.

09-13-2005, 07:51 PM	#1
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Storm/G7 Prototype + Piccies In final testing phase now of my "ultimate" waterblock. Seriously considering stopping making CPU waterblocks after this as its getting kinda "out there" in terms of cost vs performance gains. Anyway here's the pictures: http://www.employees.org/~slf/g7/ Got some work to do now, will explain the pictures as the thread develops. What's changed you ask? Just to recap what the G-numbers mean 'cos there's still some confusion over this: G1 = 1 jet G2 = 7 jets G3 = 19 jets G4 = 35-37 jets (implemented G4 has 35 jets) G5 = 55-61 jets (implemented G5 has 59 jets) G6 = 85-91 jets G7 = 117-127 jets (implemented G7 has 117 jets) G8 = 155-169 jets (could only machine on aerospace quality equipment) As for the design improvements over the original base Storm design specification: Increased jet inlet plenum area to lower pressure drop (thicker top-plate enables a more voluminous plenum so water can reach all jet inlets with minimal restriction - visible to eye)) Ported jet intakes to lower pressure drop (jet intakes are now all ellipse radiused using a custom manufactured tool - visible to eye) Tweaked jet lengths to lower pressure drop (jet lengths are shorter - internal + external mod, middle plate is significantly thinner too) Tweaked jet outlets to lower pressure drop (each jet outlet is chamfered around the outside of the jet tube to offer less restriction for flow exiting the cup structure - barely visible to eye) Multiple flow path channels to outlet to both decrease pressure drop, and more evenly distribute jet flow and total base-plate cooling (self-explanatory and very visible to eye) Base-plate bracing to boost structural strength for high-pressure CPU mounting (the 12 little posts that surround the jet area brace the base-plate and also serve as additional water turbulation for water outwash cooling effect over base-plate outside of main jetted area) Improved top-plate strength for higher pressure mounting capability (just a thicker top-plate, which is offset by thinner middle plate) Middle-plate bracing to improve jet/cup alignment when used with high-pressure pumps (additional consequence of the middle plate posts - reduces middle plate flexing meaning jets stay better focused with respect to their cup counterparts under very high pressure operation - meaning >>10psi) Improved O-ring specifications to minimise assembly based warping (smaller O-rings and slightly wider O-ring groove dimensions, thereby meaning less clamping pressure required to seal minimising the amount of pressure that gets applied to the base-plate effectively eliminating base-plate flex on assembly) Additional O-ring seal between inlet to middle plate plenum and outlet (this one's for the armchair waterblock critics - I never deemed it necessary but people often griped about the design needing one even though it doesn't, so I put it there to keep them happy anyway - may as well) Improved cup/wall ratios for higher performance (minor ratio adjustments - added base-plate strength too) Improved jet turbulators for higher performance (two prototypes being tested) Improved base-plate characteristics for higher performance (adjusted base-plate thickness suitable for G7+silver) Improved jet/cup ratios for higher performance (tweaked ratio adjustment) Improved low-flow performance without sacrificing high-flow performance (micro-jetted efficiencies are much higher even with lower flow rates - also a consequence of turbulator modifications and jet/cup ratio adjustments) Improved small bare-die performance (down to 7x7mm CPU dies) without sacrificing IHS performance. (simulated and tested adjustments to attempt to cover the widest possible variety of CPU types) There's about 12 month's worth of lessons in this block and I've spared nothing to develop this block that physically embodies everything I've learned.

09-14-2005, 09:09 AM	#8
starbuck3733t Cooling Savant Join Date: Nov 2002 Location: State College, PA Posts: 338	I MUST HAVE ONE!!! No. Seriously. You've got PM. __________________ Goliath: 3.4E@3.91/Abit IC7, Maze4 (temporarily) + custom splitter to crazy 4-way watercooling parallel loop: X800XT @ 520/1280 + AC Twinplex, AC Twinplex Northbridge, Silenstar Dual HDD Cooler, Eheim1250, '85 econoline van HC + 2x120, 1x120 exhaust - polished aluminum frame panaflo L1As, 2x18GB 10K RPM U160 SCSI, 4GB PC4000. I wanna be BladeRunner when I grow up! Project Goliath - nearing completion.

09-14-2005, 02:12 PM	#12
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Nice work. I like the new outlet scheme (inner channel). __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

09-14-2005, 04:53 PM	#15
tong Cooling Neophyte Join Date: Mar 2005 Location: Ludlow, MA Posts: 89	cathar when are you making these available? __________________ My Heatware My Ebay

09-15-2005, 02:02 PM	#18
killernoodle Thermophile Join Date: Sep 2002 Location: Charlotte, NC Posts: 1,014	That machining is incredible. What is the ID of those jets? __________________ I have a nice computer.

09-13-2005, 08:49 PM	#2
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	Sounds and looks pretty good!

09-13-2005, 09:43 PM	#3
DryFire Cooling Savant Join Date: Sep 2004 Location: USA Posts: 148	If this was made would a copper version be considered? (seeing as teh G5 was only silver)

09-13-2005, 11:07 PM	#5
Sin22 Cooling Neophyte Join Date: May 2003 Location: Singapore Posts: 66	Very nice stuff mate. Interested about the shortening of the jet tubes and the ellipse of the jet intakes. Does the elipse go through the whole tube or is it kept at only the intake area.

09-13-2005, 11:48 PM	#7
ricecrispi Cooling Savant Join Date: Jul 2005 Location: california Posts: 429	Do you think about having larger porting at the centermost inlets to focus more flow into the center cups would improve performance or is it negligible? CAN we also get more pics of the mid plate?

09-14-2005, 10:51 AM	#9
Nugit Cooling Neophyte Join Date: Jun 2003 Location: Aalborg, Denmark Posts: 85	Magnificent, the details on that block. Wow. What a ripper.. I'm speechless. What's machining time for one of those beautities?

09-14-2005, 11:28 AM	#10
Breach Cooling Neophyte Join Date: Mar 2005 Location: Washington Posts: 93	looks like a sweet block Cathar, you have obviously done all of your homework and thensome for this one. Starbuck how did I know you would be first to raise your hand

09-14-2005, 04:00 PM	#14
iakovl Cooling Neophyte Join Date: Jan 2004 Location: israel Posts: 57	silver so much Cathar, i knew u r a WC ace now i know, i was wrong u r a WC god!!!

09-15-2005, 11:50 AM	#16
murray13 Cooling Savant Join Date: Sep 2002 Location: Omaha, NE USA Posts: 216	Hi Stew, the 'Perfect' Storm. Do you know what the projected or actual pressure drop is like?

09-15-2005, 01:50 PM	#17
Holst Cooling Neophyte Join Date: Oct 2002 Location: Leics UK Posts: 9	I dont suppose you have pics of the other blocks for comparison. Especially G1, 2 and 3

09-15-2005, 03:52 PM	#19
mwolfman Cooling Savant Join Date: Jun 2004 Location: Sweden Posts: 152	Well done Cathar! Some questions... What’s the total production time (Silver and Delrin). I guess the Delrin manufacturing time is about three times the silver time? What’s the pressure drop? (IE that pump is required to run a G7 CPU AND G7 GPU block inline) Do you have to machine (mill) the silver block (silver = soft and the pattern is relatively easy to press), what’s the depth on the hols, and diameter (in mm please) any angel (like a upside cone)? What type of filter is needed so there aren’t any hold-ups in the jets? (I clean my system ones a week with air bubbles...) __________________ Very old specs below, but I keep it for sentimental values :-/ Gigabyte GA-7N400 Pro2, Barton 2500@3200 -11201MHz, X800 XL , 1 GB Cosair 3200 (6-2-2-2) @400MHz in dual channel TurboBlock Rev. 1.0~0,15452756±0,004125251°C/W , HL120.3* with 3 Papst 4412@7volt , Hydor L30II

09-16-2005, 04:49 PM	#20
dacooltech Pro/Vendor Join Date: Feb 2002 Location: NYC Posts: 267	excellent job Stew I especially loved the Wye connection integrated to the top plate... there are 2 holes in the mid plate too if i'm not mistaking... positioned above the center of the channels on the BP sucking coolant from the channels (some kind of Venturi effect perhaps? ) ... very nice I think that integrated Wye idea can be adapted to the WW as well, don't you think? __________________ www.cooltechnica.com

09-17-2005, 06:19 PM	#24
Bundles Cooling Neophyte Join Date: Mar 2003 Location: Scotland Posts: 32	They say there is a very, very, VERY fine line between being a genius and being insane, you Mr Cathar good sir, are perched very shakily on that line in my opinion. However, if it wasn't for the likes of yourself, watercooling would not be where it is today, have the reputation it has today, and just generaly be so enjoyable and interesting as it is. I just can't see you hanging up your hose though, there would always be those "Hmm, i wonder if....?" and them "Now I think if I...." moments. AS for the G7, if i could afford one, then i would take one for certain, can waterblocks get much better than this? I think not. __________________ "Okay, okay! I take it back. Unf*ck you!!!"

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)