5days of block testing finished, My best WB is...

morphling1 · 10-07-2002, 04:12 PM

Now that was same work... I just finished testing my different waterblock designs, and now installed top block back inside my case again, along with my new northbridge waterblock. And now I'm running my new agoia 1600+ at 2GHz 2.1V rock stable, and look at my max oc. at 2.2V 15°C water and unstable

Now I need to get all the data and conclusions in presentable form, I'll post all the details later and for now just show you my current top block:

g.l.amour · 10-07-2002, 04:31 PM

JEEEESUS, awestruck. congrats m8

Fixittt · 10-07-2002, 04:33 PM

ever think about going thrue the chipset first, that way you are not heating it up. That might get you a more stable clock. And it wont add enought heat to the coolant to matter to the Processor, Just an observation.

Looks really good.

Also......... what you have something against the stock ATX plug?

gone_fishin · 10-07-2002, 04:44 PM

Looking good, nice overclock.
Makes me wish I had a good digital cam also

I can see the flowpath good in them pics, the water directed to the bottom needs to travel to the top of the block then back down again in the clear top to exit, am I seeing it correctly?

Puzzdre · 10-07-2002, 05:38 PM

AAAARGHHHH!!!
I think that price for that block just went up !!!

Anyway, congratulations for looks and performace!

I'll be checking my pm's more frequently now

BRAVO!

morphling1 · 10-08-2002, 08:00 AM

Quick reply, then back to writing.
Fixittt, it doesn't realy matter which way first because I have flow rate of 10 l/min and at that flow water going in cpu block and water going out is practicaly the same temp.
Yes I have something agains atx, it was complet crap and it burned, check it out.

And this is how I fixed it

Now it works great I have good voltages on all rails and it doesn't heat up any more.

g_f , I don't think you got it right (hehe, the same way Cathar had problems with your new block), water goes in in the center then the midle block have four outles where the two distant ones from the final outlet are a little wider so the flow rate is pretty much the same, then all four outlest combined into two flow paths in top block and then exit through one outlet. Simple

Oh yeah just to rub it in here is my top overclock (stable just to get screen shoot) with 2.3V and 13°C water on this block YESS!

Note, this is not t-bred, this is old palomono core axp 1600+ agoia -y week 13, just love this cheap AMD chips.

Demoman · 10-08-2002, 05:19 PM

test looks nice.

have to say it again that blocks looks like

that burned atx connector did that happen during test or a other time?

Can O' Beans · 10-09-2002, 12:14 AM

Grats on comming up with a design that kicks ass

I agree that they have to come up with a better design for the PSU connector... I've seen the above problem quite a few times, even on cheap normal user PCs.

gone_fishin · 10-09-2002, 12:33 AM

Ok, I did not see the bottom holes because you kept the top so nice and clear

Are you chilling your water to get 13C water temps or are you testing with a hooded parka on and airconditioning?

morphling1 · 10-09-2002, 01:42 AM

Funny thing, I'm in the attics where I have temporary wireless link with my friend (I made a mistake buying usb adapter and not second access point that could be connected to network cable), and that room isn't heated so that is the room temp and also my water temp, and yes I am wearing my thermo velur

Better pic of the block and layers:

Brad · 10-09-2002, 04:06 AM

never mind the atx header or the block, look at the radiator, it's freaking awesome!

morphling1 · 10-09-2002, 11:17 AM

Ok ,here are result of my test.
First let me describe how I did the test. First the room was unheated one in the attics with temp of around 15°C, motherboard was Epox 8k3a, placed in the table with no fans anywhere to influence air temp. around socket. The pumps were Maxy- Jet 1000 and I've done two test. First with one pump then I added the second one in series. The hosing was 1/2" silicon ones total lenght 1m, the pumps were in 10l reservoar and from and then they were conected to copper heatercore 25*18*6cm with two Sunon high speed 120mm fans on top, then to the block and then back to reservoar. For thermal greese I used silicon goop that was very low viscosity, I choose this instead of as2 because I could aply it a lot more constant then as2, plus because of low viscosity and even when aplied very thin the redundant goop would squeeze out from contact between core and block. Next the clamping force, I made special nuts that hold the spring in position, and I turned those nuts the same so all the spring were aplying the same force, and that force was equal for all the blocks and was also biger then amd recomended clamping force.

Then the test, cpu frequency was 1950MHz. core volatge 2,1V the loading program was Toast which from all the programs give me highest and least fluctuating temps. Block was heat loaded for 1h until water temp reach equilibrium, then I run another 10 min where I took the measurment each second, so after 600 measures, the test was stopped and I calculate the average max temp. for one block. All this was done with a help of mbm great util and exporting results to excell. Then I re run the test with two pumps in series.
First here are the tested blocks.

MWBC mk1

MWBC mk2

MWBC mk3

morphling1 · 10-09-2002, 11:19 AM

And the last three.
M&B al spirall

M&B cu spirall

And the block I got from local store
Magnum 390

Results next, I still need to make graphs in excell

morphling1 · 10-09-2002, 02:23 PM

Ok, here are the results,

And this are the actual flow rates of the tested blocks.

I'm sure I forgot something to mention, but that's your job to remind me

Demoman · 10-09-2002, 03:01 PM

you didnt test the magnum on temp?

and things i noticed is that your mwbc blocks almost have the the same flowrate and same difference as with 2 pumps

M&B little less flow

and magnum bit more less.

can that be that these blocks use other fittings that the mwbc blocks? or just because of the design?

and that there is much difference in temp of the M&B blocks 7-9 degrees. i thought that would be less.

and ofcourse nice test

it can help people who make there own blocks to see what some things effect. and ofcourse nice for you too see that your block works pretty nice

keep up the work

and hope see more things like this. because all what matters are pics and data

(sorry for some bad english

)

bigben2k · 10-09-2002, 03:32 PM

It'd be nice to see a flow test at higher flow rates... What your graphs show, is that by increasing the flow rate about 25%, your temps dropped 1 degC, so it might not be important.

Those are nice temps, congratulations!

morphling1 · 10-09-2002, 03:44 PM

Silly me I completly forget water temps. those average max. temp are normalized to 20°C water, so the differences are only from design and flow rate.

Well no temps for Magnum, why ? Because it's mounting mechanism sucks, and I manage to burn my old 1700+ ,because the springs wasn't strong enough for good contact between block and core, so when I start up at 1950@2,1V cpu just died.
With new cpu I was a little more carefull, and started at default setting 1600+ 1.75V , and then slowly raised the clock, but again cpu just become unstable even at 1700Mhz, and the temperature was pretty normal for that clock, so like I said one part of the block was in contact and some part wasn't so I just gave up, because I'm not made out of money to just burn those cpu's.
As for fittings all block were mounted the same 1/2 od 8mm ID brass fitting, except for Magnum which was 6mm ID. Oh yeah the Magnum is one spiral chanell milled just for info

bigben2k · 10-09-2002, 03:47 PM

Let me get this straight...

Your water temp is 20 degC, and your best CPU temp is 38.4?

morphling1 · 10-09-2002, 03:49 PM

Well yes, why do you think that sucks

Puzzdre · 10-09-2002, 05:09 PM

Shheesh, sometimes I just refuse to believe that you are made of flesh and blood!

And to think that you bought that mill maybe month ago...

... can't wait to see what happens when you get your hands on CNC...

MAN !!!!

bigben2k · 10-09-2002, 05:11 PM

If you run the C/W numbers... let's see...

Delta T = 38.4 - 15.0 = 23.4
Heatload of CPU = ??

18.4 / x = your C/W rating.

According to Dan at Dansdata, The Maze 2 will do 0,35 C/W. His methodology is outlined here:
http://www.dansdata.com/coolercomp.htm

His Delta T is made up of the temp difference between air (ambient) and the thermal probe inside his simulated die.

So your heatload must be at least 52 Watts to beat the Maze 2.

gone_fishin · 10-09-2002, 08:29 PM

XP1600 @ 1950, 2.1vcore

110 watts according to benchtest.com

Benchtest is looked at to have high numbers for wattage, but they are certainly not 2X the wattage, so I think you beat the maze 2 easily

morphling1 · 10-10-2002, 02:57 AM

bb2k, first of all delta T would be 38,4-20 =18,4.
Then to heatload,according to amd specs on palomino core cpu, 1600+ (1400Mhz 1,75V) maximum thermal power is 62.8W and 56.3W is typical thermal power, if I take this numbers in calculator, typical thermal power for 1950@2,1V would be 112W similar benctest number.
You do the math if thats better then dd2

Puzzdre , when I started cnc machining... ufff, only imagination is the limit

10-07-2002, 04:12 PM	#1
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	5days of block testing finished, My best WB is... Now that was same work... I just finished testing my different waterblock designs, and now installed top block back inside my case again, along with my new northbridge waterblock. And now I'm running my new agoia 1600+ at 2GHz 2.1V rock stable, and look at my max oc. at 2.2V 15°C water and unstable Now I need to get all the data and conclusions in presentable form, I'll post all the details later and for now just show you my current top block: __________________ [My ftp, with lots of pics, hope to be home page someday

10-07-2002, 04:33 PM	#3
Fixittt CNC Beyatch Join Date: Aug 2001 Location: Tulsa Spell it backwards Posts: 721	ever think about going thrue the chipset first, that way you are not heating it up. That might get you a more stable clock. And it wont add enought heat to the coolant to matter to the Processor, Just an observation. Looks really good. Also......... what you have something against the stock ATX plug? __________________ Creator of the Spir@l Block Longest post ever http://forums.procooling.com/vbb/showthread.php?s=&postid=43808#post43808

10-07-2002, 05:38 PM	#5
Puzzdre Cooling Savant Join Date: Aug 2002 Location: Croatia Posts: 969	AAAARGHHHH!!! I think that price for that block just went up !!! Anyway, congratulations for looks and performace! I'll be checking my pm's more frequently now BRAVO! __________________ 'Out of cheese error... ...please reboot the universe (press the GBL to continue)'

10-08-2002, 08:00 AM	#6
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Quick reply, then back to writing. Fixittt, it doesn't realy matter which way first because I have flow rate of 10 l/min and at that flow water going in cpu block and water going out is practicaly the same temp. Yes I have something agains atx, it was complet crap and it burned, check it out. And this is how I fixed it Now it works great I have good voltages on all rails and it doesn't heat up any more. g_f , I don't think you got it right (hehe, the same way Cathar had problems with your new block), water goes in in the center then the midle block have four outles where the two distant ones from the final outlet are a little wider so the flow rate is pretty much the same, then all four outlest combined into two flow paths in top block and then exit through one outlet. Simple Oh yeah just to rub it in here is my top overclock (stable just to get screen shoot) with 2.3V and 13°C water on this block YESS! Note, this is not t-bred, this is old palomono core axp 1600+ agoia -y week 13, just love this cheap AMD chips. __________________ [My ftp, with lots of pics, hope to be home page someday

10-08-2002, 05:19 PM	#7
Demoman Cooling Neophyte Join Date: Jul 2002 Location: Netherlandz Posts: 8	test looks nice. have to say it again that blocks looks like that burned atx connector did that happen during test or a other time? __________________ If it gets hot! you can watercool it.

10-07-2002, 04:31 PM	#2
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	JEEEESUS, awestruck. congrats m8

10-07-2002, 04:44 PM	#4
gone_fishin Cooling Savant Join Date: May 2002 Location: Da UP Posts: 517	Looking good, nice overclock. Makes me wish I had a good digital cam also I can see the flowpath good in them pics, the water directed to the bottom needs to travel to the top of the block then back down again in the clear top to exit, am I seeing it correctly?

10-09-2002, 12:14 AM	#8
Can O' Beans Cooling Savant Join Date: Aug 2002 Location: Pa - USA Posts: 264	Grats on comming up with a design that kicks ass I agree that they have to come up with a better design for the PSU connector... I've seen the above problem quite a few times, even on cheap normal user PCs.

10-09-2002, 12:33 AM	#9
gone_fishin Cooling Savant Join Date: May 2002 Location: Da UP Posts: 517	Ok, I did not see the bottom holes because you kept the top so nice and clear Are you chilling your water to get 13C water temps or are you testing with a hooded parka on and airconditioning?

10-09-2002, 01:42 AM	#10
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Funny thing, I'm in the attics where I have temporary wireless link with my friend (I made a mistake buying usb adapter and not second access point that could be connected to network cable), and that room isn't heated so that is the room temp and also my water temp, and yes I am wearing my thermo velur Better pic of the block and layers: __________________ [My ftp, with lots of pics, hope to be home page someday

10-09-2002, 04:06 AM	#11
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	never mind the atx header or the block, look at the radiator, it's freaking awesome! __________________ 2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water

10-09-2002, 11:17 AM	#12
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Ok ,here are result of my test. First let me describe how I did the test. First the room was unheated one in the attics with temp of around 15°C, motherboard was Epox 8k3a, placed in the table with no fans anywhere to influence air temp. around socket. The pumps were Maxy- Jet 1000 and I've done two test. First with one pump then I added the second one in series. The hosing was 1/2" silicon ones total lenght 1m, the pumps were in 10l reservoar and from and then they were conected to copper heatercore 25186cm with two Sunon high speed 120mm fans on top, then to the block and then back to reservoar. For thermal greese I used silicon goop that was very low viscosity, I choose this instead of as2 because I could aply it a lot more constant then as2, plus because of low viscosity and even when aplied very thin the redundant goop would squeeze out from contact between core and block. Next the clamping force, I made special nuts that hold the spring in position, and I turned those nuts the same so all the spring were aplying the same force, and that force was equal for all the blocks and was also biger then amd recomended clamping force. Then the test, cpu frequency was 1950MHz. core volatge 2,1V the loading program was Toast which from all the programs give me highest and least fluctuating temps. Block was heat loaded for 1h until water temp reach equilibrium, then I run another 10 min where I took the measurment each second, so after 600 measures, the test was stopped and I calculate the average max temp. for one block. All this was done with a help of mbm great util and exporting results to excell. Then I re run the test with two pumps in series. First here are the tested blocks. MWBC mk1 MWBC mk2 MWBC mk3 __________________ [My ftp, with lots of pics, hope to be home page someday Last edited by morphling1; 10-15-2002 at 02:49 AM.

10-09-2002, 11:19 AM	#13
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	And the last three. M&B al spirall M&B cu spirall And the block I got from local store Magnum 390 Results next, I still need to make graphs in excell __________________ [My ftp, with lots of pics, hope to be home page someday

10-09-2002, 02:23 PM	#14
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Ok, here are the results, And this are the actual flow rates of the tested blocks. I'm sure I forgot something to mention, but that's your job to remind me __________________ [My ftp, with lots of pics, hope to be home page someday

10-09-2002, 03:01 PM	#15
Demoman Cooling Neophyte Join Date: Jul 2002 Location: Netherlandz Posts: 8	you didnt test the magnum on temp? and things i noticed is that your mwbc blocks almost have the the same flowrate and same difference as with 2 pumps M&B little less flow and magnum bit more less. can that be that these blocks use other fittings that the mwbc blocks? or just because of the design? and that there is much difference in temp of the M&B blocks 7-9 degrees. i thought that would be less. and ofcourse nice test it can help people who make there own blocks to see what some things effect. and ofcourse nice for you too see that your block works pretty nice keep up the work and hope see more things like this. because all what matters are pics and data (sorry for some bad english ) __________________ If it gets hot! you can watercool it.

10-09-2002, 03:32 PM	#16
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	It'd be nice to see a flow test at higher flow rates... What your graphs show, is that by increasing the flow rate about 25%, your temps dropped 1 degC, so it might not be important. Those are nice temps, congratulations! __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

10-09-2002, 03:44 PM	#17
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Silly me I completly forget water temps. those average max. temp are normalized to 20°C water, so the differences are only from design and flow rate. Well no temps for Magnum, why ? Because it's mounting mechanism sucks, and I manage to burn my old 1700+ ,because the springs wasn't strong enough for good contact between block and core, so when I start up at 1950@2,1V cpu just died. With new cpu I was a little more carefull, and started at default setting 1600+ 1.75V , and then slowly raised the clock, but again cpu just become unstable even at 1700Mhz, and the temperature was pretty normal for that clock, so like I said one part of the block was in contact and some part wasn't so I just gave up, because I'm not made out of money to just burn those cpu's. As for fittings all block were mounted the same 1/2 od 8mm ID brass fitting, except for Magnum which was 6mm ID. Oh yeah the Magnum is one spiral chanell milled just for info __________________ [My ftp, with lots of pics, hope to be home page someday

10-09-2002, 03:47 PM	#18
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Let me get this straight... Your water temp is 20 degC, and your best CPU temp is 38.4? __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

10-09-2002, 03:49 PM	#19
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	Well yes, why do you think that sucks __________________ [My ftp, with lots of pics, hope to be home page someday

10-09-2002, 05:09 PM	#20
Puzzdre Cooling Savant Join Date: Aug 2002 Location: Croatia Posts: 969	Shheesh, sometimes I just refuse to believe that you are made of flesh and blood! And to think that you bought that mill maybe month ago... ... can't wait to see what happens when you get your hands on CNC... MAN !!!! __________________ 'Out of cheese error... ...please reboot the universe (press the GBL to continue)'

10-09-2002, 05:11 PM	#21
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	If you run the C/W numbers... let's see... Delta T = 38.4 - 15.0 = 23.4 Heatload of CPU = ?? 18.4 / x = your C/W rating. According to Dan at Dansdata, The Maze 2 will do 0,35 C/W. His methodology is outlined here: http://www.dansdata.com/coolercomp.htm His Delta T is made up of the temp difference between air (ambient) and the thermal probe inside his simulated die. So your heatload must be at least 52 Watts to beat the Maze 2. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

10-09-2002, 08:29 PM	#22
gone_fishin Cooling Savant Join Date: May 2002 Location: Da UP Posts: 517	XP1600 @ 1950, 2.1vcore 110 watts according to benchtest.com Benchtest is looked at to have high numbers for wattage, but they are certainly not 2X the wattage, so I think you beat the maze 2 easily

10-10-2002, 02:57 AM	#23
morphling1 Cooling Savant Join Date: Sep 2001 Location: Slovenia Posts: 468	bb2k, first of all delta T would be 38,4-20 =18,4. Then to heatload,according to amd specs on palomino core cpu, 1600+ (1400Mhz 1,75V) maximum thermal power is 62.8W and 56.3W is typical thermal power, if I take this numbers in calculator, typical thermal power for 1950@2,1V would be 112W similar benctest number. You do the math if thats better then dd2 Puzzdre , when I started cnc machining... ufff, only imagination is the limit __________________ [My ftp, with lots of pics, hope to be home page someday

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