What do you want to see in a Water Block??? - Page 3

Brad · 01-02-2002, 06:09 PM

yeah, I was talking about how busy fixittt is, but you should look at dual pelt ideas, and other ways of doing it

jaydee · 01-02-2002, 06:13 PM

Quote:

Originally posted by Brad
yeah, I was talking about how busy fixittt is, but you should look at dual pelt ideas, and other ways of doing it

I will try and come up with something. it will be a different design then what i have now but that is the only way to get over a 1/4 NPT thread.

Brad · 01-02-2002, 06:34 PM

yeah, I definately feel you need to get over the 1/4" ID fittings.

have you thought of doing a whole pile of thin fins arranged in a sprial pattern, with a central inlet?

jaydee · 01-02-2002, 06:46 PM

Quote:

Originally posted by Brad
yeah, I definately feel you need to get over the 1/4" ID fittings.

have you thought of doing a whole pile of thin fins arranged in a sprial pattern, with a central inlet?

Yes, but it isn't practical, or cost effective. And I am not sure it would help anything.

To do that would cost the buyer about $90-$100 a unit if i made it. It would take a good 2 hrs on my mill to carve all that out. Probably take one bit a block providing it dosn't break. Thats $8 alone. Maybe someone with a bigger mill with a faster spindal speed could do it in less than an hour. To much for my machine.

Brad · 01-02-2002, 06:52 PM

if it performed a lot better, then I know I'd pay that sort of price. if it only performed a bit better, I wouldn't bother with it.

still, I think you should make one, just to see how it works

jaydee · 01-02-2002, 08:10 PM

Quote:

Originally posted by Brad
if it performed a lot better, then I know I'd pay that sort of price. if it only performed a bit better, I wouldn't bother with it.

still, I think you should make one, just to see how it works

If I get some extra time when my Copper stock gets here I will try and see if I can do one and see what it takes. The hardest part is going to be making the code for the CNC.

Brad · 01-02-2002, 08:52 PM

also, you should make it at first so the base is 10mm thick or something.

test it, then take 1mm off it, then test again, etc, see what happens with your temps

jaydee · 01-02-2002, 09:02 PM

Quote:

Originally posted by Brad
also, you should make it at first so the base is 10mm thick or something.

test it, then take 1mm off it, then test again, etc, see what happens with your temps

Yeah good idea. It is easy to take material off with the mill.

Brad · 01-03-2002, 05:50 AM

you should get a parabola type effect, it goes down slowly, levels out, then rapidly goes up

jaydee · 01-03-2002, 11:04 AM

Quote:

Originally posted by Brad
you should get a parabola type effect, it goes down slowly, levels out, then rapidly goes up

Not sure how effective that would be but it sounds cool. I giess the outlet would be narrower than the inlet? causing a speed up as it tries to puch through the narrower outlet?

resago · 01-03-2002, 11:05 AM

I think he was refering to the results you'd have from shaving the base.

jaydee · 01-03-2002, 11:37 AM

Quote:

Originally posted by resago
I think he was refering to the results you'd have from shaving the base.

If that is the case I have no idea what that means then?

resago · 01-03-2002, 01:29 PM

means temps would look like this: 27C, 26C, 25.5C,25.6C,27C,30C,35C, etc.

each temp would correspond to a base thickness.

decodeddiesel · 01-03-2002, 01:30 PM

Basically he was stating the resulting temp pattern would be like a "bell curve" or parabolic in nature. In other words you would see the temps getting better and better as the base got thinner and thinner, once you reached optimal base thickness the temps would level out, then as you continued to get thinner the temps would begin to rise again, though at a much steeper curve (well in theory anyway). Regardless, that would be a very interesting experiment and the results would be a huge contribution to the watercooling community.

decodeddiesel · 01-03-2002, 01:31 PM

Quote:

Originally posted by resago
means temps would look like this: 27C, 26C, 25.5C,25.6C,27C,30C,35C, etc.

each temp would correspond to a base thickness.

thank you resago

jaydee · 01-03-2002, 03:30 PM

I see!!!

Well I just got a block of Copper to work with and it was sheer torchure on my Mill but I got a Copper block Milled out now. I am going to have to rework the CNC code to make it easier on the mill.

The rest of my stock should be here Friday. I will post pics of the copper block on my website and i will be testing it that same way I did with the Aluminum one to see what different temps I can get between the two blocks.

I think I may have found a flaw in my last testing aswell. There was to much airflow around the back of the motherboard which may have caused lower temps. I will retest each block better this time.

Butcher · 01-03-2002, 04:24 PM

Using motherbard sensors is a bad way to test, you should use a proper probe either in the block baseplate or use a chip with an internal diode.

decodeddiesel · 01-03-2002, 04:26 PM

Quote:

Originally posted by Butcher
Using motherbard sensors is a bad way to test, you should use a proper probe either in the block baseplate or use a chip with an internal diode.

I agree, see what pHaestus has done on http://www.voidyourwarranty.net or Joe's article "Anatomy of a review" for the best way to measure die temps.

Brad · 01-03-2002, 05:20 PM

resago got it.

the only thing is that each block will be differant, and Al vs Cu will give differant results. IE, I'd expect the Al to need a bit thicker base than Cu.

Also, with pelts I think you could get away with something a bit thinner because the heat is already spread out a bit more, over 40mm x 40mm or 48mm x 48mm, whereas a normal cpu die is like 11mm x 12mm

jaydee · 01-03-2002, 05:21 PM

I have an external probe that I will try to use for the next test. It is real tight in between the CPU and the water block, even more so with the XP's. I will look at the links above and try to figure out a better way.

I tested the on board prpobes agains the external probe the best I could and the on board was very close. it is hard to get a good temp reading off the CPU though. Even more so with the XP's with the little chips on the bottom and the CPU itself is narrower.

I burned up my T-Bird which I would rather do the tests with.

Always make sure you unplug the power supply before removing the HS/water block. I got in a hurry and forgot to and the system booted up (bumped the on button)and smoked the T-bird.

decodeddiesel · 01-03-2002, 05:24 PM

Quote:

Originally posted by jaydee116
...Always make sure you unplug the power supply before removing the HS/water block. I got in a hurry and forgot to and the system booted up (bumped the on button)and smoked the T-bird.

Ouch, that sucks

The smell of burnt silicon is about as bad as the smell of burnt flesh IMO.

Brad · 01-03-2002, 05:28 PM

maybe you could try mounting the prob on the underside of the cpu, as close to the core as possible without touching any caps

Fixittt · 01-03-2002, 09:20 PM

Hey Jaydee,

here is what I have done to allow more clearance for a temp probe.

Mill a small channel on the under side of the block, that leads uo to the edge of the cpu core.
Take some heatsink coumpond and apply it to a processor core, then slip a waterblock on the moutning bolts. press down firmly to contact the CPU. Pull the block off, and you have an exact location where the CPU touches the block. Take something and scribe a line where one side of the CPU stops, then just mill you a small channel to that line maybe .003 past the line, so U know U have a good reading. and POOF!

Hope it might help

check you email.

jaydee · 01-03-2002, 09:35 PM

Yeah I was thinking about something like that. I will give it a try.

My e-mail has been screwed up today. I am not getting half my e-mails. My FTP is going on and off aswell.

Quote:

Originally posted by Fixittt
Hey Jaydee,

here is what I have done to allow more clearance for a temp probe.

Mill a small channel on the under side of the block, that leads uo to the edge of the cpu core.
Take some heatsink coumpond and apply it to a processor core, then slip a waterblock on the moutning bolts. press down firmly to contact the CPU. Pull the block off, and you have an exact location where the CPU touches the block. Take something and scribe a line where one side of the CPU stops, then just mill you a small channel to that line maybe .003 past the line, so U know U have a good reading. and POOF!

Hope it might help

check you email.

jaydee · 01-03-2002, 09:54 PM

My personal e-mail address is jaydee116@attbi.com . Feel free to use that one aswell. I hope this server issue with my host is taken care of real soon. i have had a lot of problems in the last few days.

01-03-2002, 04:24 PM	#67
Butcher Thermophile Join Date: May 2001 Location: UK Posts: 1,064	Using motherbard sensors is a bad way to test, you should use a proper probe either in the block baseplate or use a chip with an internal diode. __________________ Once upon a time, in a land far far away...

01-03-2002, 09:20 PM	#73
Fixittt CNC Beyatch Join Date: Aug 2001 Location: Tulsa Spell it backwards Posts: 721	Hey Jaydee, here is what I have done to allow more clearance for a temp probe. Mill a small channel on the under side of the block, that leads uo to the edge of the cpu core. Take some heatsink coumpond and apply it to a processor core, then slip a waterblock on the moutning bolts. press down firmly to contact the CPU. Pull the block off, and you have an exact location where the CPU touches the block. Take something and scribe a line where one side of the CPU stops, then just mill you a small channel to that line maybe .003 past the line, so U know U have a good reading. and POOF! Hope it might help check you email. __________________ Creator of the Spir@l Block Longest post ever http://forums.procooling.com/vbb/showthread.php?s=&postid=43808#post43808

01-02-2002, 06:09 PM	#51
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	yeah, I was talking about how busy fixittt is, but you should look at dual pelt ideas, and other ways of doing it

01-02-2002, 06:34 PM	#53
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	yeah, I definately feel you need to get over the 1/4" ID fittings. have you thought of doing a whole pile of thin fins arranged in a sprial pattern, with a central inlet?

01-02-2002, 06:52 PM	#55
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	if it performed a lot better, then I know I'd pay that sort of price. if it only performed a bit better, I wouldn't bother with it. still, I think you should make one, just to see how it works

01-02-2002, 08:52 PM	#57
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	also, you should make it at first so the base is 10mm thick or something. test it, then take 1mm off it, then test again, etc, see what happens with your temps

01-03-2002, 05:50 AM	#59
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	you should get a parabola type effect, it goes down slowly, levels out, then rapidly goes up

01-03-2002, 11:05 AM	#61
resago Cooling Savant Join Date: Mar 2001 Location: state of denial Posts: 488	I think he was refering to the results you'd have from shaving the base.

01-03-2002, 01:29 PM	#63
resago Cooling Savant Join Date: Mar 2001 Location: state of denial Posts: 488	means temps would look like this: 27C, 26C, 25.5C,25.6C,27C,30C,35C, etc. each temp would correspond to a base thickness.

01-03-2002, 01:30 PM	#64
decodeddiesel Cooling Savant Join Date: Oct 2001 Location: classified Posts: 534	Basically he was stating the resulting temp pattern would be like a "bell curve" or parabolic in nature. In other words you would see the temps getting better and better as the base got thinner and thinner, once you reached optimal base thickness the temps would level out, then as you continued to get thinner the temps would begin to rise again, though at a much steeper curve (well in theory anyway). Regardless, that would be a very interesting experiment and the results would be a huge contribution to the watercooling community.

01-03-2002, 03:30 PM	#66
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	I see!!! Well I just got a block of Copper to work with and it was sheer torchure on my Mill but I got a Copper block Milled out now. I am going to have to rework the CNC code to make it easier on the mill. The rest of my stock should be here Friday. I will post pics of the copper block on my website and i will be testing it that same way I did with the Aluminum one to see what different temps I can get between the two blocks. I think I may have found a flaw in my last testing aswell. There was to much airflow around the back of the motherboard which may have caused lower temps. I will retest each block better this time.

01-03-2002, 05:20 PM	#69
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	resago got it. the only thing is that each block will be differant, and Al vs Cu will give differant results. IE, I'd expect the Al to need a bit thicker base than Cu. Also, with pelts I think you could get away with something a bit thinner because the heat is already spread out a bit more, over 40mm x 40mm or 48mm x 48mm, whereas a normal cpu die is like 11mm x 12mm

01-03-2002, 05:21 PM	#70
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	I have an external probe that I will try to use for the next test. It is real tight in between the CPU and the water block, even more so with the XP's. I will look at the links above and try to figure out a better way. I tested the on board prpobes agains the external probe the best I could and the on board was very close. it is hard to get a good temp reading off the CPU though. Even more so with the XP's with the little chips on the bottom and the CPU itself is narrower. I burned up my T-Bird which I would rather do the tests with. Always make sure you unplug the power supply before removing the HS/water block. I got in a hurry and forgot to and the system booted up (bumped the on button)and smoked the T-bird.

01-03-2002, 05:28 PM	#72
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	maybe you could try mounting the prob on the underside of the cpu, as close to the core as possible without touching any caps

01-03-2002, 09:54 PM	#75
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	My personal e-mail address is jaydee116@attbi.com . Feel free to use that one aswell. I hope this server issue with my host is taken care of real soon. i have had a lot of problems in the last few days.

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