first time i can post some results here

g.l.amour · 11-23-2002, 12:36 PM

too many ppl to thank.

puzzdre,#rotor,maddogme,mfpmax,bigben2k, fragn'stein alot i'm forgetting.

this is a pure ripoff of 3rotor's design. credit where credits due.

top half is brass(70Cu/30Zn)
lower half is electrolitic copper

LiquidRulez · 11-23-2002, 02:20 PM

Good job!

What's it look like on the inside? And what are your temps??

g.l.amour · 11-23-2002, 02:59 PM

well, as i said, this is a ripoff of the purest form of rotor's design. lots of little drillholes connected with a dremel.

at the moment, temps seem to be +/- 2°C better than with my 1/2" maze2. but as i just swapped blox, there is still prolly some air to be purged out. and maybe some settling of thermal grease.

ofcourse, if i follow the billA line of thought. me making comments on temps is a bit stupid. i mounted the block once, so results could be at their peak, or at their low for this specific block. whereas my maze has been mounted that much times that i can positively know what the best mount could get me tempwise.

as it is, i have 2 rads in parallel, so i am already at the limits of what watercooling can offer. so any tempdrop is a hardfought battle.

i do wonder however, if the block would have had an advantage if the top half was also in electrolitic copper, instead of brass. i chose for the brass because it doesn't oxidise so easily, and i had already completed the tophalf after noticing that the metals merchant hadn't sold me copper but brass. so being sortof lazy, i just kept the work, and made only the lower half out of pure Cu.

Puzzdre · 11-23-2002, 03:07 PM

Smoooooooth finish!!!

Did you polish the barbs too?

It surely looks very nice!

g.l.amour · 11-23-2002, 03:35 PM

i polished, every-thing

this was the most fun part... spending an afternoon polishing and coveting something that one has spent blood , sweat and tears on over the last couple of weeks.

g.l.amour · 11-23-2002, 03:44 PM

OMG

temps with NO fans running on rads

43°C load!!!!!

koslov · 11-23-2002, 03:49 PM

One thing I don't understand about #rotor's and your blocks: Why is there a channel pattern in the top as well as the base? Seems like you would want all the water to come into contact with the base, where the heat source is. Did you do this for turbulence?

g.l.amour · 11-23-2002, 03:57 PM

if there is 1 mistake i made in this design, then u have pointed right at it koslov...

rotor ofcourse isn't a n00b as i am. he prolly doesn't drill half as deep on his top half as on his bottom half. on pix , something like that is really hard to see.

so , what i would change in an upcoming block.

lets say, top half gets 2mm drill; lower half gets a 6mm drill. instead off both being about 5mm deep. next time i would also use a narrow dremel wheel for the center part. and larger diameter wheel for outer parts. i feel that now my block has indeed a bit too low flow resistance. and maybe too much of the water is indeed flowing via the top half where it is doing basically much less than it could be doing flowing in the bottom half.

but but but, cuts and drills can't be uncut and undrilled so, aiming even a bit higher for the next block

BillA · 11-23-2002, 04:45 PM

your first variant should simply be a flat top, copper - brass - whatever

don't be concerned with the flow rate, you're after a boost in the convection rate in the bottom

let us know

g.l.amour · 11-23-2002, 05:28 PM

i most certainly will.

atm i think much can be improved indeed.

a baseplate thickness of 3 - 2 mm should be possible. the pin that is centered on the intake nozzle could be sharpened with a dremel, so that the incoming water doesn't bounce off at the first plane of the concave square (this prolly only sounds understandable to me, lol)

anyway, i will try my best to improve the design, then post some pix and results so it becomes a bit more clear.

BillA · 11-23-2002, 05:36 PM

scrounge up a view of the Atlantis
look at its pin center
rough up the bottom in the central area

g.l.amour · 11-23-2002, 05:45 PM

there is one thing on the atlantis block, i can hardly make out.

u have the spirally, starshaped form, with above its center the inlet. is the center part of the star hollow or filled (as with TT's block) on the available pics i seem to be seeing that the central colum is hollow with a canal at the furtest from outlet spiral. damn it is hard to say it in words.

anyway it was made, it must be possible to rework it in a rudimentary way to a #rotor compatible design.

FRAGN'STIEN · 11-23-2002, 05:46 PM

AWESOME JOB!!!

Glad to see you made one g.l.amour. I love the contrast between the copper and the brass. Very nice indeed. I agree with what bill is saying, but in my case I used 1/4" plate for both top and bottom and I think resistance would have been terrible if I would have only drilled the bottom half.

g.l.amour · 11-23-2002, 06:05 PM

indeed, at 1/4" it seems a bit risky letting only the lower half do the work.

to take this to a more theoretic level. one can easily calculate the surface where the water enters the block at the nozzle (circular). one can also easily calculate the total surfaces where the water can evacuate to the exits points (small rectangles)

if i follow bill's line of thought, it would be better to design it that the evacuating water has a narrower total surface, than the total surface of entry.

what would be the best tradeoff. a jet effect in the hosebarb. or a larger hosebarb with multiple small jets inside the waterblock. (i feel again, that i will prolly be the only one understanding this mumbo jumbo)

FRAGN'STIEN · 11-23-2002, 06:19 PM

I agree with that theory g.l.amour. That was my reasoning for trading the 1/2" outlets on mine and replacing them 3/8" barbs. I figured this would speed the velocity of the water comming out of the block.

#Rotor · 11-23-2002, 06:29 PM

the ideal would be to have the two halves of the block in different thickness. say the bottom half to be 1/2" (13mm) and the top half to be 1/4" (6mm) then you can do the drilling and dremeling in only the bottom. One problem will arise from doing this, those Dremel cutting disks will be able to reach the bottom of the channel, only while they are brand new.... by the time you get to the 2nd row, the wheel will be to small to go in all the way.... from there my method of using two sides that are the same as far as grid design. The depth of the channels in my blocks are more or less the same, for both sides.... I do however use VDFC extensively, and for that to work properly, I need to do both the top and bottom halves. The block gets bolted together so tight, that the formed pins, as far as the heat is concerned, would appear to be one continues long pin, all the way to the top.

the other very big benifit of having both sides the same, is the fact that you don't need to buy stock in 2 different sizes. In my case that means I can get away with one 6feet lenth per order... saves me having to shell out $350+ every time I need copper.

g.l.amour · 11-23-2002, 06:36 PM

lol, minimum length i can buy copper in is 18ft

@ rotor, ever considered using some plastic top (plexi). then u could dremel only the 3/8 - 1/2 copper piece. maybe a bit deeper so one could still have sufficient flow.

BillA · 11-23-2002, 06:45 PM

FRAGN'STIEN

no, minimize head losses every place in the system (including the wb exit)
except in that specific area where you wish to maximize the convection rate
~ a 1" diameter

g.l.amour

for your 2ed variant reverse your connection sizes
3/8" inlet, 1/2" outlet

#Rotor

"The block gets bolted together so tight, that the formed pins, as far as the heat is concerned, would appear to be one continues long pin, all the way to the top."

nope, not unless you vacuum weld the pieces
you have a very conventional thermal joint using water as the thermal gap filler
better than air, but not so good
the reason you cannot 'see' any difference (if you've tested this) is that no heat is going up those pins anyway

#Rotor · 11-23-2002, 06:55 PM

Quote:

Originally posted by g.l.amour
if i follow bill's line of thought, it would be better to design it that the evacuating water has a narrower total surface, than the total surface of entry.

what would be the best tradeoff. a jet effect in the hosebarb. or a larger hosebarb with multiple small jets inside the waterblock. (i feel again, that i will prolly be the only one understanding this mumbo jumbo)

remember that water-in always equal water-out.....

so no matter what you do, do not restrict the entry or the exit.

this design relies heavily on surface area and turbulence, and turbulence is directly proportional to flow-velocity, not flow-rate... there is a big difference.

I found that by having the inlet right on top of the core (hotspot), I get the "eye of the storm" syndrome. Meaning that right in the middle, there is barely any relative movement. the most turbulent region in such a configuration would be in circular patterns out from the middle, with more verocity as you approach midway between the inlet and outlets.

Virtually the same problem high density heatsinks have, when the fan is right on top, centered to the core. That's why a side mounted drafting fan config, usually gives better core temps, in heatsink world of coarse.

#Rotor · 11-23-2002, 07:11 PM

Quote:

Originally posted by unregistered

#Rotor

"The block gets bolted together so tight, that the formed pins, as far as the heat is concerned, would appear to be one continues long pin, all the way to the top."

nope, not unless you vacuum weld the pieces
you have a very conventional thermal joint using water as the thermal gap filler
better than air, but not so good
the reason you cannot 'see' any difference (if you've tested this) is that no heat is going up those pins anyway

I agree, definitely not as good as a solid copper pin, but considering all other factors, be it manufacturing, or cost-effectiveness. I see it to be as effective as it can be...

true that the heat might not even get that far up and that would in fact be the ideal situation. The faster the heat can be suspended in the liquid, the faster is can get removed by the liquid.

g.l.amour

I have tried it yes...

.

let's just say reliability is not where I want it to be... never had trouble with it, but I'm not willing to risk potential catastrophe, just for some eye-candy, besides, it looks good for the first week......

g.l.amour · 11-23-2002, 07:13 PM

it will become mighty interesting to start testing some different designs out, now that i have the tools...

thx for all the input.

Bill, u got some daring ideas. they will be considered; thx

g.l.amour · 11-23-2002, 07:17 PM

well, it wasn't as much for the pretty fx i asked. but more towards u saying cost effectiveness, not wanting to buy 2 diff thicknesses of copper; if u were to use some PE top, then u can stick to one copper bar dimension

FRAGN'STIEN · 11-23-2002, 07:40 PM

Quote:

Originally posted by unregistered
FRAGN'STIEN

no, minimize head losses every place in the system (including the wb exit)
except in that specific area where you wish to maximize the convection rate
~ a 1" diameter

Hmmm. So reverse the inlet and outlets(3/8" inlet and 1/2" outlets in this case)? Also would it make sense to create a spray(nozzle) effect over the core, similar to what cathar did? I've been working on an idea using small tapered holes at the inlet.

g.l.amour · 11-23-2002, 07:48 PM

i made a 3/8 nozzle very low down in the barb. but i think the jet is being deflected on the first standing concave pin. if that pin could be more pointy...

BillA · 11-23-2002, 08:29 PM

Quote:

Originally posted by g.l.amour
. . . . if that pin could be more pointy...

yup

FRAGN'STIEN
Cathar's wb has no spray, directed flow would be a better description

Violenti has the jets
but you, nor he, have enough pressure to make it work

11-23-2002, 12:36 PM	#1
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	first time i can post some results here too many ppl to thank. puzzdre,#rotor,maddogme,mfpmax,bigben2k, fragn'stein alot i'm forgetting. this is a pure ripoff of 3rotor's design. credit where credits due. top half is brass(70Cu/30Zn) lower half is electrolitic copper __________________ yo soy un tiburón

11-23-2002, 02:59 PM	#3
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	well, as i said, this is a ripoff of the purest form of rotor's design. lots of little drillholes connected with a dremel. at the moment, temps seem to be +/- 2°C better than with my 1/2" maze2. but as i just swapped blox, there is still prolly some air to be purged out. and maybe some settling of thermal grease. ofcourse, if i follow the billA line of thought. me making comments on temps is a bit stupid. i mounted the block once, so results could be at their peak, or at their low for this specific block. whereas my maze has been mounted that much times that i can positively know what the best mount could get me tempwise. as it is, i have 2 rads in parallel, so i am already at the limits of what watercooling can offer. so any tempdrop is a hardfought battle. i do wonder however, if the block would have had an advantage if the top half was also in electrolitic copper, instead of brass. i chose for the brass because it doesn't oxidise so easily, and i had already completed the tophalf after noticing that the metals merchant hadn't sold me copper but brass. so being sortof lazy, i just kept the work, and made only the lower half out of pure Cu. __________________ yo soy un tiburón

11-23-2002, 03:07 PM	#4
Puzzdre Cooling Savant Join Date: Aug 2002 Location: Croatia Posts: 969	Smoooooooth finish!!! Did you polish the barbs too? It surely looks very nice! __________________ 'Out of cheese error... ...please reboot the universe (press the GBL to continue)'

11-23-2002, 03:35 PM	#5
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	i polished, every-thing this was the most fun part... spending an afternoon polishing and coveting something that one has spent blood , sweat and tears on over the last couple of weeks. __________________ yo soy un tiburón

11-23-2002, 03:44 PM	#6
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	OMG temps with NO fans running on rads 43°C load!!!!! __________________ yo soy un tiburón

11-23-2002, 02:20 PM	#2
LiquidRulez Cooling Savant Join Date: Sep 2002 Location: In Hell Posts: 322	Good job! What's it look like on the inside? And what are your temps??

11-23-2002, 03:49 PM	#7
koslov Cooling Savant Join Date: Jul 2002 Location: Posts: 264	One thing I don't understand about #rotor's and your blocks: Why is there a channel pattern in the top as well as the base? Seems like you would want all the water to come into contact with the base, where the heat source is. Did you do this for turbulence?

11-23-2002, 03:57 PM	#8
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	if there is 1 mistake i made in this design, then u have pointed right at it koslov... rotor ofcourse isn't a n00b as i am. he prolly doesn't drill half as deep on his top half as on his bottom half. on pix , something like that is really hard to see. so , what i would change in an upcoming block. lets say, top half gets 2mm drill; lower half gets a 6mm drill. instead off both being about 5mm deep. next time i would also use a narrow dremel wheel for the center part. and larger diameter wheel for outer parts. i feel that now my block has indeed a bit too low flow resistance. and maybe too much of the water is indeed flowing via the top half where it is doing basically much less than it could be doing flowing in the bottom half. but but but, cuts and drills can't be uncut and undrilled so, aiming even a bit higher for the next block __________________ yo soy un tiburón

11-23-2002, 04:45 PM	#9
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	your first variant should simply be a flat top, copper - brass - whatever don't be concerned with the flow rate, you're after a boost in the convection rate in the bottom let us know

11-23-2002, 05:28 PM	#10
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	i most certainly will. atm i think much can be improved indeed. a baseplate thickness of 3 - 2 mm should be possible. the pin that is centered on the intake nozzle could be sharpened with a dremel, so that the incoming water doesn't bounce off at the first plane of the concave square (this prolly only sounds understandable to me, lol) anyway, i will try my best to improve the design, then post some pix and results so it becomes a bit more clear. __________________ yo soy un tiburón

11-23-2002, 05:36 PM	#11
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	scrounge up a view of the Atlantis look at its pin center rough up the bottom in the central area

11-23-2002, 05:45 PM	#12
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	there is one thing on the atlantis block, i can hardly make out. u have the spirally, starshaped form, with above its center the inlet. is the center part of the star hollow or filled (as with TT's block) on the available pics i seem to be seeing that the central colum is hollow with a canal at the furtest from outlet spiral. damn it is hard to say it in words. anyway it was made, it must be possible to rework it in a rudimentary way to a #rotor compatible design. __________________ yo soy un tiburón

11-23-2002, 05:46 PM	#13
FRAGN'STIEN Cooling Savant Join Date: May 2002 Location: Rathdrum, ID Posts: 380	AWESOME JOB!!! Glad to see you made one g.l.amour. I love the contrast between the copper and the brass. Very nice indeed. I agree with what bill is saying, but in my case I used 1/4" plate for both top and bottom and I think resistance would have been terrible if I would have only drilled the bottom half. __________________ This space for rent

11-23-2002, 06:05 PM	#14
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	indeed, at 1/4" it seems a bit risky letting only the lower half do the work. to take this to a more theoretic level. one can easily calculate the surface where the water enters the block at the nozzle (circular). one can also easily calculate the total surfaces where the water can evacuate to the exits points (small rectangles) if i follow bill's line of thought, it would be better to design it that the evacuating water has a narrower total surface, than the total surface of entry. what would be the best tradeoff. a jet effect in the hosebarb. or a larger hosebarb with multiple small jets inside the waterblock. (i feel again, that i will prolly be the only one understanding this mumbo jumbo) __________________ yo soy un tiburón

11-23-2002, 06:19 PM	#15
FRAGN'STIEN Cooling Savant Join Date: May 2002 Location: Rathdrum, ID Posts: 380	I agree with that theory g.l.amour. That was my reasoning for trading the 1/2" outlets on mine and replacing them 3/8" barbs. I figured this would speed the velocity of the water comming out of the block. __________________ This space for rent

11-23-2002, 06:29 PM	#16
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	the ideal would be to have the two halves of the block in different thickness. say the bottom half to be 1/2" (13mm) and the top half to be 1/4" (6mm) then you can do the drilling and dremeling in only the bottom. One problem will arise from doing this, those Dremel cutting disks will be able to reach the bottom of the channel, only while they are brand new.... by the time you get to the 2nd row, the wheel will be to small to go in all the way.... from there my method of using two sides that are the same as far as grid design. The depth of the channels in my blocks are more or less the same, for both sides.... I do however use VDFC extensively, and for that to work properly, I need to do both the top and bottom halves. The block gets bolted together so tight, that the formed pins, as far as the heat is concerned, would appear to be one continues long pin, all the way to the top. the other very big benifit of having both sides the same, is the fact that you don't need to buy stock in 2 different sizes. In my case that means I can get away with one 6feet lenth per order... saves me having to shell out $350+ every time I need copper. __________________ There is no Spoon....

11-23-2002, 06:36 PM	#17
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	lol, minimum length i can buy copper in is 18ft @ rotor, ever considered using some plastic top (plexi). then u could dremel only the 3/8 - 1/2 copper piece. maybe a bit deeper so one could still have sufficient flow. __________________ yo soy un tiburón

11-23-2002, 06:45 PM	#18
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	FRAGN'STIEN no, minimize head losses every place in the system (including the wb exit) except in that specific area where you wish to maximize the convection rate ~ a 1" diameter g.l.amour for your 2ed variant reverse your connection sizes 3/8" inlet, 1/2" outlet #Rotor "The block gets bolted together so tight, that the formed pins, as far as the heat is concerned, would appear to be one continues long pin, all the way to the top." nope, not unless you vacuum weld the pieces you have a very conventional thermal joint using water as the thermal gap filler better than air, but not so good the reason you cannot 'see' any difference (if you've tested this) is that no heat is going up those pins anyway

11-23-2002, 07:13 PM	#21
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	it will become mighty interesting to start testing some different designs out, now that i have the tools... thx for all the input. Bill, u got some daring ideas. they will be considered; thx __________________ yo soy un tiburón

11-23-2002, 07:17 PM	#22
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	well, it wasn't as much for the pretty fx i asked. but more towards u saying cost effectiveness, not wanting to buy 2 diff thicknesses of copper; if u were to use some PE top, then u can stick to one copper bar dimension __________________ yo soy un tiburón

11-23-2002, 07:48 PM	#24
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	i made a 3/8 nozzle very low down in the barb. but i think the jet is being deflected on the first standing concave pin. if that pin could be more pointy... __________________ yo soy un tiburón

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