Want your water blocks tested?

Hey all

Pretty soon i am reciving a bunch of non commercial blocks for testing and i think i have 4 so far... i will be doing a round up and posting reults on my site..

Anyways this isnt Billa quality testing. Im still trying to aqquire the funds to get stuff to test c/w but the way i test is i let each computer run with the same load of

TOAST
CPU BURN
SISOFT SANDRA
F@H
SETI

For three days and i use the temp recording software from the external touching the die sensor and i use all the numbers generated (disregaurding the first hour because the CPU may still be heating up then) and i average them. Before this i leave the compute runtouched for 6 hours and use the software to check idle temperature (again by averaging.)

If you have a DIY block you want to submit please PM me or AIM me. I am not charging anything for the tests because they arent nearly exensive enought o charge for. Anyways this should give a pretty accurate representation of which block is the best.

I am counting the turns of the screws for clamping pressure and i am weighing the arctic silver i use on a gram scale and applying it in the same thickness with a credit card.

I am not trying to promote my site but i want to know if more people want to have thier blocks tested.. PM me if you are interested. Thanks.

what qualifies you as a tester?

xX~EXCELSIOR~Xx,

I really don't mind if you are a tester. The community clearly needs several. But what qualifies you to be a tester?

What is your experience? On a scale of 1 to 100 lets make BillA a 90, nobody is 100. Just as in many sporting events we have to have the ability to allow someone to beat Bill so he only gets a 90 because he was first. I believe he is likely to have no equals any time soon. By the way, being interested earns one point.

What is your test procedure?

What equipment do you have and its accuracy? What measurements will you provide? C/W and delta-p are minimal.

This will do as a starting point.

Re: what qualifies you as a tester?

deeppow said:

xX~EXCELSIOR~Xx,

I really don't mind if you are a tester. The community clearly needs several. But what qualifies you to be a tester?

What is your experience? On a scale of 1 to 100 lets make BillA a 90, nobody is 100. Just as in many sporting events we have to have the ability to allow someone to beat Bill so he only gets a 90 because he was first. I believe he is likely to have no equals any time soon. By the way, being interested earns one point.

What is your test procedure?

What equipment do you have and its accuracy? What measurements will you provide? C/W and delta-p are minimal.

This will do as a starting point.


Why am I allowed to ask such question? Because I was an editor of the Journal of Heat Transfer (the best heat transfer journal in the world) for 5 years and am a past chairman of the American Soc. of Mechanical Eng.'s Heat Transfer Division. That qualifies me. Nothing personnal here but you want to do hard science then this is what we're allowed to ask.

Click to expand...

Read my post

My experience is only being into computers for the past 13 or so years and watercooling for the past 2-3

As i said right now i am still trying to aqquire more equipment so i can be more accurate but right now im pretty sure this is as accurate
as ic an get withou expensive equipment to calculate c/w (which i AM working on getting just still need to allocate funds )

First of all what i do is run the computer with no load on. use the temperature monitor software which is taken froma probe PERMANENTLY mounted in one place RIGHT near the CPU die. I let the computer sit there for 6 hours disable any hibernation whatsoever and make sure there are no extra apps for any of the waterbklocks . The temp monitor records the temperatures to a .1 of a degree every 10 minutes. I take the readings from every 10 minutes and average them to get Idle Tempearture. then I use all of the following apps:
TOAST
CPU BURN
SISOFT SANDRA
F@H
to give the processor as much load as possible. Even two of thos eapps would give the processor all the load but i have the extra ones running to make sure EVERY clock cycle is used. I let the monitor sofwtare (which is hoked up to a DIFFERENT computer via the temp diode therefore not using the comptuers CPU) record the temperature at 10 minute intervals for Three days. I disregaurd teh first hour because the cpu may stillh ave been warming up and it may skew results, and then i average all the left over LOAD temperatures to get LOAD. The arctic silver i use is from the same tube measured out on my VERY accurate gram scale using the same amount for each noe. I spread it the same thickness (verified using caliper) with a credit card and i coun the number of turns i use to screw down the block to make sure that the pressure is the same throughout.

This is as accurate as i can get without Bill type of equipment.

Back to the stone age we go, but I think pHaestus at procooling is getting together a simulater and other equiptment, along with some knowledgable advice from the artist formerly known as BillA.
If you are serious then you don't want your first crack at it to be mediocre, research the articles and postings of BillA to get a feel of what you are in for.
So far what you describe is pretty close to what [H] did a while back.

Here we go again.

We always seem to have a disconnect here. The point is that
without the proper test equipment AND procedures AND experience
your results wont be worth spit. There is no shortcut.

Please don't take this personally. Mother Nature has
rigged the rules. If you don't play by Her rules you are
in the penalty box.

Note that BillA has spent $thousands and more time than
some of you have been alive on this and there is no
way to do it for less.

I am NOT talking about COMMERCIAL blocks here... I'd understand if we wer speaking of commercial blocks but im noooot.

There isnt much room for error and whatever block does best i can pretty much garuntee you would be the best.
I have read bill's articles and posts and i do understand the concept.

Sad thing is i just wish i was a magzine Maxpc got around 7 pro type kits from vapochill prometia and other companies and all they did was just use the same mobo and same cpu

I expected all these types of replies and i understand. I am not charging. I actually pay for the shipping. Bill charged a well deserved 90 dollars for his awesome DIY block testing and i am trying to offer a free service that pretty muchw ould proclaim which block would be best.

I AM allocating a few thousand dollars to try to find some nice equip. to do my testing on but that is still a ways away.

Everyone is soo critical of the results but i am wodnering WHY. This is pretty much as controlled as it can get without a couple thousand dollar bill type system and it would show the best block.

The logic cooling person would probably ALSO charge for this... if you have a block thats DIY and you want it to be tested do you want to pay for it? Ahh screw it no one is listening . Delete this thread mods if you think im just causing trouble. sorry

Hey, you aren't causing trouble. There is no problem here. We
can save you a lot of time and pain by showing you what you are
up against. Free or not your results would be anecdotal at best.

SEriously..

I understand that it wont be exactly accurate. But any inaccuracies would be consistant throughout the blocks.

I have eliminated almost ALL possible inacuracies. Except for the way i sporead the arctic silver i cant find anything else that wouldnt remain consistent

look at it from a scientific point of view

If i use ALL the clock cycles and use the SAME software for load.. i mount it the SAME way and ushc as explaine and i use a DIFFERNT computer for the temp readings what varriables are there. The only one i can think of is the waterblock and thats all that that is the factor being tested.

Anecdotal? With the described methods i think it would work all right for proving if one block was better than the other. Did your ead through the entire testing prociedure or did yous e eiwasnt using a couple K worth of testing equip and immediately say it is flawed and wouldtn provide good results.

Btw hopefully WHEN i get the mutli thousand testign equip i would only be charging $20 bucks for the block testing.

Bonne chance, mon ami.

Well i just got 500 bucks closer to my goal for testing equip

Called dad up... willing to lend me out 500 bucks. an di have around 500 i can put up for it.. only around 700-1000 dollars more needed

And thanks i actually understood that (dad taught me french )

Again, I nor do I think anyone else is trying to trash, flame or whatever you. Your interest and desires are admirable.

But after many years of writing simulations of single and two phase flow and doing experiments, I've decided that bad data is worse than no data. It leads you to a completely wrong perspective of the real world.

With C/W or pressure drop you are dealing with defined quantities that represent the physical world . Really doesn't matter if it is a commerical block or someone's home made. The measurement is a quantity that must be comparable with those Bill made, ain't no way around it. It is the definition. And the uncertainity in those measurement are required by any semi-serious publication that might present your results. Realize of course that many web sites don't know the meaning of the terms. But to illustrate, a measurement of a C/W = 0.24 with an uncertainity of 0.1 (due to poor measurements) isn't of much use. Add to that some reasonable confidence level to correct for the finite sample you're going to have and you got meaningless data.

But then, what do I know.


PS. I had read your post.

Al right

Fine.

I geuss i wont even try i talthough i think it would be fairly accurate no one will believe teh results.

WHere is teh link where billa explained ALL his equip. I saw it earlier and read through it all i just need to scavenge to see parts and such. Dad said if i make abusiness proposal he can bump the fundage up to 700 possibly if he accepts

Thx

i think you should go for it, get 10 blocks, that arent mass produced, just peoples blocks they have designed/made etc. and test the suckers. my 2 cents.

Yah taht would be great but i onyl have 4 and probably from this thread no one else will be wiling to let me test thier blocks thioll i get this equip.

man, I have sent so many out to be tested that never did and never got returned. I dont know if I can afford anymore

Just trying to get you to do your homework ahead of time. I've said my piece. I'm outa here, i.e. unsubscribed.

Be

Daddy's 700 won't get you far, add a few zeros to it and see what he says.

gone_fishin said:

Daddy's 700 won't get you far, add a few zeros to it and see what he says.

Click to expand...

ACtually it is a sart.

700+500=1,200. If i find used parts and such i am betting i could build it all for 2-2 1/2 grand max.

I think you're getting flamed because you don't have the gear yet, so it's possible that you might not understand its limits. I have yet to see your proposed testing method, but if it involves using a PC, I think you ought it to us to let us know how you propose to have more accurate readings than any of us can do at home.

I have no doubt that you'll do well, I'd just like to know how!

As for the thermal paste, a thermal pad might give you some more consistency. Have you explored that?

bigben2k said:

I think you're getting flamed because you don't have the gear yet, so it's possible that you might not understand its limits. I have yet to see your proposed testing method, but if it involves using a PC, I think you ought it to us to let us know how you propose to have more accurate readings than any of us can do at home.

I have no doubt that you'll do well, I'd just like to know how!

As for the thermal paste, a thermal pad might give you some more consistency. Have you explored that?

Click to expand...

OOo thermal pad!

Thank you! that actually was some helpful advice