define the relevant topics

[BillA]

a suggestion, define the sub-catagories as separate topics/threads
else confusion will reign

the heat 'die'
the power source and measurement
temperature measurement
flow measurement
pressure measurement
mounting considerations
and overall, accuracy and repeatability

go to it

[Phant0m51]

BillA, did you mean to reply to a thread, instead of make a new one?

If you didn't mean to make a new one, I'm confused as hell...

[BillA]

all this in relation to testing only
my suggestion was merely to start threads for each of the above listed topics (or others ?) so as to hold a bit of focus
- each of those topics is rather distinct

worth noting is that the testing of hsfs is not covered
- some VERY interesting 'things' lie there

[Joe]

I am pondering makint this its own category

Covering all aspects of cooling not just water. HSF's and such would be very good to see that worked on.

[bigben2k]

If ya'll want to get into some preliminary discussions, that's fine, but I still have some administrative "stuff" to take care of, to get the WTA (Waterblock Testing Alliance) going.

I didn't intend on covering HSF: it's a waterblock testing alliance... Yes, there's something there, so let's take a look.

Bill's right on: we need to address these topics, individually:
-the heat 'die'
-the power source and measurement
-temperature measurement
-flow measurement
-pressure measurement
-mounting considerations
-and overall, accuracy and repeatability

Let's hear from everyone, as to what other topics should be included, then I'll open up the threads.


We also have to keep in mind that this is for the purpose of setting a testing standard, not for dictating how a rig should be constructed. What I mean is that I don't want this to end up with: "you must use Fluke meter #xxxx", but rather, "you must use a meter with X features and capabilities". (Am I making sense?)

[BillA]

Originally posted by bigben2k
If ya'll want to get into some preliminary discussions, that's fine, but I still have some administrative "stuff" to take care of, to get the WTA (Waterblock Testing Alliance) going.
. . . .
yea, the decoder rings
Ben, why do you grasp so for control ?

pHaestus is the preeminently qualified individual in your club
let HIM decide the what and how, to whom and by whom

[bigben2k]

I'm not grasping for control at all! I said that ya'll can go ahead and talk about anything you want, didn't I?

I was just trying to say that the WTA isn't ready to vote on anything, because I haven't got the administrative part down yet, plus, I'm still recruiting.

I also stated that I'd open up the threads. This is a free Forum, if you want to open up all those threads, go right ahead: I'm not stopping you!:rolleyes:

[bigben2k]

Originally posted by unregistered
a suggestion, define the sub-catagories as separate topics/threads
else confusion will reign

the heat 'die'
the power source and measurement
temperature measurement
flow measurement
pressure measurement
mounting considerations
and overall, accuracy and repeatability

go to it
AS you stated, you made a "suggestion". So maybe I wasn't clear: I suggest that we go over the topics, to see if there's anything else that anyone wants to discuss.

Personally, I'd like to start a discussion on how all of the cross-testing is going to work, because I'm really fuzzy on it.

So the question here is more: "Why do you want to dictate which topics that are going to be discussed?"

[gone_fishin]

Well, that other thread has seemed to have roll over effect to the new forum here. I like the heading for this new topic btw. Testing can encompass a whole spectrum of individual parts such as pumps for example. There should be a thread the clearly describes how to translate test data or specs over to a real world application. The info is already scattered about on this site but all in one place would be nice for people.
Benchmarking also covers a wide scope and overclockers just love them benchies.:)

Suggestion, Ben the best moderation is done in moderation.

[Since87]

Originally posted by bigben2k
So the question here is more: "Why do you want to dictate which topics that are going to be discussed?"

Bill gave you a succinct list of the fundamental technical issues that MUST be addressed if this testing is to amount to anything.

Adding to that list does nothing to further understanding of the fundamental issues.

Personally, I'd like to start a discussion on how all of the cross-testing is going to work, because I'm really fuzzy on it.

What point is there to discussing the administrative details of a cross-checking scheme for mythical test benches. Until baseline criteria are decided for the test bench itself, such discussion is just noise.

[bigben2k]

Originally posted by gone_fishin
I like the heading for this new topic btw. Testing can encompass a whole spectrum of individual parts such as pumps for example. There should be a thread the clearly describes how to translate test data or specs over to a real world application. The info is already scattered about on this site but all in one place would be nice for people.
Benchmarking also covers a wide scope and overclockers just love them benchies.:)

Suggestion, Ben the best moderation is done in moderation.

I definitely want to get some work done in that area. It might fall outside of the Alliance, I don't know yet.

For all of Bill's test results on OC, the reality is that there's always been a gap between that testing, and how the end user is going to apply it, to build a water-cooled PC.

So all that data about the pressure drop of various heatercores, is really going to come in handy.


Originally posted by Since87
Bill gave you a succinct list of the fundamental technical issues that MUST be addressed if this testing is to amount to anything.

Adding to that list does nothing to further understanding of the fundamental issues.


What point is there to discussing the administrative details of a cross-checking scheme for mythical test benches. Until baseline criteria are decided for the test bench itself, such discussion is just noise.
I agree, but the cross-testing is a very important part too.

The "cross-testing" as a topic isn't meant to be administrative, I meant it to be about how all the data is going to tie together. Similar to the example I gave, if tester A tests a block and gets results Xa with error margin Ea, and tester B runs the same tests, and gets result Xb within error Eb, and so on with testers C, D, E... How do the results tie in together, and how can we use the results to make an adjustment to the results of each of our test benches, if that's possible?

So there, that's what I meant.:p

Otherwise, although Bill's list is fundamental, does it really emcompass everything?

This is one of the problems that I've seen lately: in what I do, I don't consider anything as fact, unless it's been verified by two other independant sources.

So as much as I appreciate Bill's input here, I don't know if he wrote that up quickly with his right hand, while eating a sandwich with his left, or what. I take nothing for granted here, no matter from who it comes.

[arcsylver]

Originally posted by Since87
Bill gave you a succinct list of the fundamental technical issues that MUST be addressed if this testing is to amount to anything.

Adding to that list does nothing to further understanding of the fundamental issues.



What point is there to discussing the administrative details of a cross-checking scheme for mythical test benches. Until baseline criteria are decided for the test bench itself, such discussion is just noise.

I have to agree with Since87 on this one.

Until a design of some sort is ironed out for the benches themselves there is not really any point in discussing the how part of the testing.

First thing we need to iron out is the basic requirements for a test bench for this group of testers so that we are all on the same page.

IMO I think the order that things should be focused on is thus.

The testbench itself.

The tests to be run/Accuracy level of such

The format which data is to be presented.

[bigben2k]

and I have to agree there, that the first step is going to be to define the test bench "bar", but it doesn't prevent us from including another topic.

The reaon that I bring it up, is because I see it as something important, and it worries me that I don't have the details of it, because it's a really important part of what the Alliance is going to be doing.

Heck, maybe I'm just jumping the gun, or being paranoid, I don't know... I'm not twisting anyone's arm here, am I?:shrug:

[arcsylver]

Well BB2K, I hate to say it but lately the pissing contest between you and BillA has been rather annoying to say the least.

The idea is a sound one though the logistics are a nightmare.

My suggestion is to take things one step at a time and iron them at as you go.

It isnt like we are going to be producing any test results next week after all. Take your time (all of us that is not just you) and do it right the first time.

As for BillA. Lay of the cutthroat remarks will ya.

This was started as a project to develop a replacement for your testing methods to be honest. And I have to agree in some aspects that it certainly seems you are dead set on trying to prevent being replaced in that role.

Since the onset of this discussion you have berated, bad mouthed, and from my point of view tryed to kill this idea.

If you are so worried about someone's qualifications for "leading" a group of independent testers why dont you get down off your high horse and do it yourself. Otherwise stick to staying something helpful instead of berating every post you read.

(dons asbestos flame suit and waits for BillA's response)

[bigben2k]

Ok, I'll drop it, for now.


To try to get back on topic...

So far we have:
-the heat 'die'
-the power source and measurement
-temperature measurement
-flow measurement
-pressure measurement
-mounting considerations
-and overall, accuracy and repeatability

(and many thanks to Bill for it!)

Does anyone else think that anything should be added / changed / removed?

If not, I'll go ahead and start a thread for each one.

[jaydee]

Originally posted by arcsylver


IMO I think the order that things should be focused on is thus.

The testbench itself.

The tests to be run/Accuracy level of such

The format which data is to be presented.
nah this doesn't make sence. You cannot define a test bench untill you know what the test bench is going to be used for. You need to:

1) Figure out what measurements are going to be "expected".

2) What kind of form those measurments need to be "formatted" in.

3) A proceedure for taking those measurments.

4) Find the equipment that is capable of doing 1-3.


If you pick your equipment first then your limited to what that equipment can do.

[arcsylver]

Originally posted by jaydee116
nah this doesn't make sence. You cannot define a test bench untill you know what the test bench is going to be used for. You need to:

1) Figure out what measurements are going to be "expected".

2) What kind of form those measurments need to be "formatted" in.

3) A proceedure for taking those measurments.

4) Find the equipment that is capable of doing 1-3.


If you pick your equipment first then your limited to what that equipment can do.

This is true. So I need to reverse the top two on my list I guess to be.

The tests to be run/Accuracy level of such

The testbench itself.

The format which data is to be presented.

[bigben2k]

Originally posted by bigben2k
-the heat 'die'
-the power source and measurement
-temperature measurement
-flow measurement
-pressure measurement
-mounting considerations
-and overall, accuracy and repeatability

I'm going to try to write up a little intro, and open up those topics as seperate threads.

By day's end.

[bigben2k]

Done.

The intro is meant as some kind of guide, and perhaps an intro of the topic. Please feel free to discuss anything you want, any way you see fit.

If need be, we may add/remove or split a thread (as soon as I get mod rights on this particular Forum).


Here's a sample of a soon-to-be-proposed "logo" I drafted up quickly: (It's been a slow day :( )

[MadDogMe]

I really don't see what there is to discuss? It's been done and all you have to do is copy it as close as you can financialy afford. The only question you really need to ask is 'how good is good enough?' (if it's any lower than has already been achieved you need to ask yourself if it's going to do H2o cooling any good). How many of these 'setups' are you planning on making BB? If only one, then why not just donate to Ph like was previously suggested? (if Ph is even interested?). If lot's, then it's going to be like asking 20 people to build a clock each, and then expect them all to keep the same time!...

I don't consider myself a neigh sayer but I don't think this is ever going to happen. An alliance suggests 'lots' of different testers, and I just don't see people pumping money(the kind that's needed!!) into testing waterblocks for the fun of it. This smacks of 'Flogging a dead horse' :shrug: (that's whipping, not trying to sell :p )...

What BillA started at OC'ers makes sense too me, but it has to be paid for because of the huge amount of man~hours involved (which is why I wonder/doubt that Ph would be interested)...

Trying to develope the best sytem would be a better,achievable idea to me. an 'X block with Xrad with Xfan/s with Xamount of tubing with Xpump is the best setup possible today' kindof idea. Something that every new user could benifit from, rather than forming a 'pointless' elite club, it's not as though it's going to improve H2o cooling is it?...

Maybe it's time people expanded into other hobbies as well as watercooling? I just don't see that it warrants this much attention without financial gain involved. If people have so much free time why not start up a 'Pro Users against Kidy porn/Exploits' web policing force, or something?...

[bigben2k]

MadDogMe: I'm going to take a minute here, and try to reply to your concerns, which are valid!

First off, on the money aspect. Yes, I did have a choice to either try to round up people around pHaestus, but if you keep in mind that Bill's test rig costs ~$20'000, even if I was able to get 20 people, there is no way that I could reasonably expect each member to fork out $1'000, for testing gear that's going to end up in Canada, and it's not clear that pHaestus may want to do what everyone else might expect him to do, with that test bench.

So the only viable alternative, from my prespective, is to get a bunch of people together, that would be willing to spend $1'000 on their own rigs, and work off of each other to try to get to some level of accuracy, as cross-testing may allow.

The first step of course, is recognizing that yes, there has been some efforts made by various individuals towards building a test bench, and yes, there have been many discussions. But it's not my position to dictate what is acceptable or not: I'm neither interested nor qualified, single-handedly, to do that. It would be much much more efficient and productive to use those past bits of information, and get a few good heads together, to set a bar.

Each member may (or may not) have an interest in building a test rig of their own. Some members may simply want to voice their opinion, from their perspective (aka manufacturer). So they're welcome to join, even though they may not be able to vote on the more technical issues. There's room for anyone that's interested, that can see the benefit, and that can make a contribution.

Just PM or e-mail me your e-mail address. Try to include your motives for joining.

I hope that I've made that clear, but feel free to ask me any other questions!

[Joe]

WBTA hehehe first thing I thought of was "Why Bother Testing Anything" when you can just give it a 9/10 Editors choice and be leet! :)

hehe

sorry yet another non productive post by yours truly :)

[RoboTech]

Hey Ben,

As I suggested in my e-mail to you yesterday, I would like to see some clarification of goals and objectives - maybe that desrves to be a topic?

Just exactly what are you (we) trying to accomplish here?

Thanks... :)

[jaydee]

Correct me if I am wrong but:

1) Round up people interested in quality WB testing.

2) Putting together a list of procedures and equipment needed to do this.

3) Calibrating all the test benches with each other. (probably the most impossible part)

4) Making this the standard for WB testing so readers can see the WBTA logo on a review and know it was done decently. Kind of like seeing ASE on oil and what not.

[bigben2k]

What Robotech is referring to, is wether the testing that we're trying to discuss is "analytical" or "comparative".

The first is about measuring with as much precision as possible, the actual level of performance of a block. The second, is more simply about being able to compare two or more blocks, and determining, with certainty, which one perform better, and to some limited extent, by how much.

Obviously, the primary purpose of the WBTA is about analytical testing, but I believe that we will be covering comparative testing, one way or the other.

[jaydee]

I would think it would have to be pretty analytical to be accuratly comparitive. Other wise whats the point in comparing if you were not analylitical in the measurments..... Which seems to be the problem today.

[Blackeagle]

I would suggest that a fan area is also needed here.

Not only would fan testing aid those who are making use of air cooling thier CPU, but it would aid in choosing the best fan(s) for use on rads, to get them to perform thier best.

[bigben2k]

Originally posted by jaydee116
I would think it would have to be pretty analytical to be accuratly comparitive. Other wise whats the point in comparing if you were not analylitical in the measurments..... Which seems to be the problem today.
I see what you mean, but if you think about it for a second, analytical would measure the actual performance, and would put an actual "number" on it, where comparative would only give you a rough (but valid) measure of "which is better".

If I use a car analogy here for a second... Picture yourself at a stop sign, in your fat/beefed up Mustang, and some kid pulls up next to you, in some fancy decorated Honda, and he's rev'ing his engine.

From my perspective, I'm going to find out pretty quickly which one of you is the fastest. On the other hand, if I had the proper instruments, I could measure your quarter mile time, and take it anywhere for comparison.

[nikhsub1]

Originally posted by Blackeagle
I would suggest that a fan area is also needed here.

Not only would fan testing aid those who are making use of air cooling thier CPU, but it would aid in choosing the best fan(s) for use on rads, to get them to perform thier best.
Think chiller, not rad for WB testing.

[jaydee]

Originally posted by bigben2k
I see what you mean, but if you think about it for a second, analytical would measure the actual performance, and would put an actual "number" on it, where comparative would only give you a rough (but valid) measure of "which is better".

If I use a car analogy here for a second... Picture yourself at a stop sign, in your fat/beefed up Mustang, and some kid pulls up next to you, in some fancy decorated Honda, and he's rev'ing his engine.

From my perspective, I'm going to find out pretty quickly which one of you is the fastest. On the other hand, if I had the proper instruments, I could measure your quarter mile time, and take it anywhere for comparison.
Then we MUST do anyalitical testing. What if that Honda had a $20,000 engine with NOS and twin turbos and was capable of 0-60 in 8 seconds. Hard to tell from just walking by.

[bigben2k]

Originally posted by Blackeagle
I would suggest that a fan area is also needed here.

Not only would fan testing aid those who are making use of air cooling thier CPU, but it would aid in choosing the best fan(s) for use on rads, to get them to perform thier best.
Well, the WBTA isn't going to get into fan testing, but if you want to open it up, go ahead.

WE will need a "cooling solution" thread though, wether it's a chiller or some rad/fan combo.

[bigben2k]

Originally posted by jaydee116
Then we MUST do anyalitical testing. What if that Honda had a $20,000 engine with NOS and twin turbos and was capable of 0-60 in 8 seconds. Hard to tell from just walking by.
Exactly.

But it still doesn't stop you from comparing blocks. If you smoked the poor fool, let it be known. ;)

[gone_fishin]

Originally posted by bigben2k
Exactly.

But it still doesn't stop you from comparing blocks. If you smoked the poor fool, let it be known. ;)

Is this the goal of the clubhouse?

Your mission statement could read...

Does your block smoke, we do.:D

On a further thought, are members in your club who actually make blocks going to compare theirs to interested folk who make submissions? I can hear the cries of unfair now in the distance should their performance not come out favorably. They may even want their donations back if any were made.
Have you thought this through yet?

Impartiality is a high priority in comparative testing of any kind.

[RoboTech]

Originally posted by bigben2k
What Robotech is referring to, is wether the testing that we're trying to discuss is "analytical" or "comparative". Yes, I assumed the WBTA is all about analytical testing - higher resolution, more accuracy, and a LOT more skill, money and time than even good comparative testing requires.

By clarifying objectives and goals I mean to better define (put into words?) what we expect to accomplish. Is the goal just to be able to accurately measure a block's C/W? - how accurately? Heck, I'm no expert so just thinking through the goals I thought might help us get on (or confirm we are all on) the same path... :)

At the end of next year (?) how will we measure our success?

[jaydee]

Originally posted by bigben2k
Exactly.

But it still doesn't stop you from comparing blocks. If you smoked the poor fool, let it be known. ;)
But what if you smoked him because the USER doesn't know how to drive? Proves nothing on the cars performance. ;) Which leads us back to the problem we have today with a bunch of ignorant fools doing half assed reviews.


gone_fishin has a point. It would be pretty easy for people to start shouting bias if the WBTA members blocks are showing to be better than others. :shrug:

[Groth]

Originally posted by jaydee116
gone_fishin has a point. It would be pretty easy for people to start shouting bias if the WBTA members blocks are showing to be better than others. Credibility. This is the big reason the standards and procedures need to be open and the development process transparent. If someone doesn't believe my results, they can reproduce them themselves.

[jaydee]

Originally posted by Groth
Credibility. This is the big reason the standards and procedures need to be open and the development process transparent. If someone doesn't believe my results, they can reproduce them themselves.
Sure if you have the equipment and knowledge to try and reproduce the results, but I think we were talking about the 99% of the rest of the people out there that don;t know jack about anything. They see WBTA blocks better than other big name companies and they will think the WBTA is BS and just a marketing scam to promote our own blocks.

[Groth]

Well, I don't think anything we do can help the people that don't want to educate themselves. They will always be prey for marketing types and will alway be prone to weird conspiracy beliefs. Screw 'em. The whole idea is accurate information for those who care enough to learn.

[gone_fishin]

Originally posted by jaydee116
They see WBTA blocks better than other big name companies and they will think the WBTA is BS and just a marketing scam to promote our own blocks.

Bingo;)

[jaydee]

Originally posted by Groth
Well, I don't think anything we do can help the people that don't want to educate themselves. They will always be prey for marketing types and will alway be prone to weird conspiracy beliefs. Screw 'em. The whole idea is accurate information for those who care enough to learn.
That pretty much rules out any reason whatsoever to make a WBTA. I thought this was for showing the untrained people that a review has been done under a respected standard. Those who care enough already know better and do not need this WBTA!

[Groth]

Originally posted by jaydee166
That pretty much rules out any reason whatsoever to make a WBTA. I thought this was for showing the untrained people that a review has been done under a respected standard. Those who care enough already know better and do not need this WBTA! Jeez, man, we don't need to obtain god-like status, where everyone bows before the might and power of a certified test. As it stands now, block reviews range from half-assed to fully-assed. I buy stuff and I want accurate reviews.

[jaydee]

Originally posted by Groth
Jeez, man, we don't need to obtain god-like status, where everyone bows before the might and power of a certified test. As it stands now, block reviews range from half-assed to fully-assed. I buy stuff and I want accurate reviews.
Well if we are all going to spend all this time and money each on a test bed then it damn well better be for more than a few people! I don't understand your thinking here?

[bigben2k]

Wow, so much talk, in just 24 hrs!

Gone_Fishin: The WBTA is not a "law enforcement" agency, nor is it a judge of characters. The honesty and integrity of each member is assumed.

However, to earn a WBTA label, a test bench must be presented to all members, and that's going to include the whole testing procedure. This is no small effort. It's a lot of very detailed work.

So if a tester really, really wants to do all of those things, then cheat and make false claims, that's not our problem, at least until it becomes apparent that it's going on.

As with anything else, you should always keep in mind that even if you get one person's opinion, it doesn't necessarily make it a fact. If you want to make sure that a particular block performs as well as it has been claimed, then you can submit it for comparative testing, or submit it for another round of analytical testing by another member, or someone else entirely. That's just common sense, really. Of course, having a test bench with a label, should mean that you won't have to ask a second source to verify the results. No testing should happen without you knowing the testing procedure.

To clarify: the WBTA is going to set standards, and approve testbenches. What a member ends up doing with it, is up to that member. We don't differentiate between testers and blockmakers.

So if you're going to make claims of inappropriateness, I'd suggest that you get your information together, and send it to the WBTA, so that we can review it. (We'll have a process for that).

Thank you.



Robotech: I'm polishing up the mission statement, as well as objectives.


Jaydee116: Yes, driving errors are possible. That's why comparative testing also includes making sure that such variations are accounted for. Did you really think that you'd only get to run a single test, and call it a day? ;)

[jaydee]

Originally posted by bigben2k


Did you really think that you'd only get to run a single test, and call it a day? ;)
Happens everyday....;)

[gone_fishin]

Originally posted by bigben2k
Wow, so much talk, in just 24 hrs!

Gone_Fishin: The WBTA is not a "law enforcement" agency, nor is it a judge of characters. The honesty and integrity of each member is assumed.

However, to earn a WBTA label, a test bench must be presented to all members, and that's going to include the whole testing procedure. This is no small effort. It's a lot of very detailed work.

So if a tester really, really wants to do all of those things, then cheat and make false claims, that's not our problem, at least until it becomes apparent that it's going on.

As with anything else, you should always keep in mind that even if you get one person's opinion, it doesn't necessarily make it a fact. If you want to make sure that a particular block performs as well as it has been claimed, then you can submit it for comparative testing, or submit it for another round of analytical testing by another member, or someone else entirely. That's just common sense, really. Of course, having a test bench with a label, should mean that you won't have to ask a second source to verify the results. No testing should happen without you knowing the testing procedure.

To clarify: the WBTA is going to set standards, and approve testbenches. What a member ends up doing with it, is up to that member. We don't differentiate between testers and blockmakers.

So if you're going to make claims of inappropriateness, I'd suggest that you get your information together, and send it to the WBTA, so that we can review it. (We'll have a process for that).

Thank you.



Robotech: I'm polishing up the mission statement, as well as objectives.


Jaydee116: Yes, driving errors are possible. That's why comparative testing also includes making sure that such variations are accounted for. Did you really think that you'd only get to run a single test, and call it a day? ;)

If I were to spend my hard earned money and time on a test bench, why would it need a label from some club. The equiptment needed is already certified. All that would be needed for anyone to verify its potential accuracy would be a full disclosure of each componant and the procedures for its use along with results to be scrutinized. Which could easily be posted here on this forum.

Is it your hope that blockmakers will use your group as their R&D department? How about anyone trusting you guys with nondisclosure agreements?

[bigben2k]

Originally posted by gone_fishin
If I were to spend my hard earned money and time on a test bench, why would it need a label from some club. The equiptment needed is already certified. All that would be needed for anyone to verify its potential accuracy would be a full disclosure of each componant and the procedures for its use along with results to be scrutinized. Which could easily be posted here on this forum.

Is it your hope that blockmakers will use your group as their R&D department? How about anyone trusting you guys with nondisclosure agreements?
As I've stated before, one of the WBTA's purpose is to set a minimum standard. The benefit to you, if you build a test bench, is that you'll know ahead of time, what kind of accuracy you'll need to achieve. You're free to exceed it.

You may use "certified equipment" (do you mean "calibrated", or spec'ed?), but it doesn't mean that the application is correct. So you submit your test bench details, along with the procedure, to the WBTA, and members will vote on wether or not it can hold the WBTA label. The review of the test bench is by good people, with the purpose of making sure that it meets WBTA standards.

Once you have the WBTA label, what you do with your test bench is up to you. Wether you run testing for a website review, or sell the service to blockmakers, you've got the WBTA backing that your testing is sufficiently accurate, and properly done.

Of course no one is twisting your arm to join the WBTA. If you'd rather have your test bench reviewed in a Forum, you're free to do so. There are more than one means to an end.


As for blockmakers, the WBTA doesn't offer testing services, but individual members may offer it. NDAs are common for new and upcoming products, but that's between the tester and the blockmaker.

[bigben2k]

Just for kicks...

I was just going over Bill Adam's article, titled "Waterblock Bench Testing" (here (http://www.overclockers.com/articles654/)), and I thought I'd quote this paragraph, from page 1:

It is hoped that the test method outlined here might be a 'talking point' for some future collaboration between interested testing parties. After a general agreement on test methods, the real issue of cross-calibration could be addressed to enable the comparison of test results from different sources. And farther yet down the road, standardized test procedures could be drafted and validated with conventional round-robin testing.

[gone_fishin]

Originally posted by bigben2k
Just for kicks...

I was just going over Bill Adam's article, titled "Waterblock Bench Testing" (here (http://www.overclockers.com/articles654/)), and I thought I'd quote this paragraph, from page 1:

Don't ignore the rest of the article.

He has answered all of the relevent questions in that one article. The limitations of his setup. Why cross calibration is required for comparative data. How to cross calibrate with multiple test setups, even a suggested set of standard test conditions for determining a cross calibration offset curve for each unique test die package. That should have been the very first place to start with this whole new forum instead of rehashing and trying to cut corners.

The best thing a so called WBTA could offer is to make and loan out a single standard test die with insulation, probes and heat elements for other testers to cross calibrate with. That is the way I see the model BillA laid out. If this WBTA is to lead the way it must do so by example (do it right).

Also needed is to win the lotto, mortgage the house or wait for that rich uncle to kick off.

[pHaestus]

i do collaborative research for a living, and we don't usually do it by cobbling together a bunch of half assed equipment for every lab involved in the project. Instead, we decide upon the project's needs and grant a single top notch solution. Does a single website or forum need 10 testers all cross calibrating with one another at all? I don't see why. Would the hardware community as a whole benefit from this kind of collaboration? Sure. This exercise is basically preaching to the choir at least in its current form.

What I see in these threads for the most part is the following thought process:

Hmm I wanna test some waterblocks and not get flamed by Jaydee and/or BillA on the internet when I'm done. How do I test them? Hey let's look at Bill's bench testing article; I bet it explains it! Ok this seems reasonable; how much will it cost? Holy mother of god HOW MUCH? Ok I better e-mail Bill and find out if there's a cheaper way. Shit there's NOT? I better make an alliance and a bunch of posts on a forum then. Maybe with enough people involved we can talk this whole cost thing out of existance. Or at least get rid of the error bars.

I want to make a waterblock testing setup just like Bill's except with hookers and blackjack. On second thought never mind the blackjack.....and the testing setup.

[jaydee]

Originally posted by pHaestus


What I see in these threads for the most part is the following thought process:

Hmm I wanna test some waterblocks and not get flamed by Jaydee and/or BillA on the internet when I'm done.
Might be easier just to Ban me and Bill eh? :D I will tell you now though, my flaming (I never really seen it as flaming but whatever) days are going to be winding down. As will my posts here and everywhere. Newbs can relax a little now! Got more important things going on now and coming up soon....

[Joe]

Originally posted by pHaestus
What I see in these threads for the most part is the following thought process:

Hmm I wanna test some waterblocks and not get flamed by Jaydee and/or BillA on the internet when I'm done. How do I test them? Hey let's look at Bill's bench testing article; I bet it explains it! Ok this seems reasonable; how much will it cost? Holy mother of god HOW MUCH? Ok I better e-mail Bill and find out if there's a cheaper way. Shit there's NOT? I better make an alliance and a bunch of posts on a forum then. Maybe with enough people involved we can talk this whole cost thing out of existance. Or at least get rid of the error bars.


Hmmm seems the nail on the head has been hit :D

[Groth]

<begin rant>

Good god, no wonder so few results and useful articles get posted! Is the point of these forums to mock people and revel in your supposed superior knowledge?

Yes, Bill had a wonderful test bench, but do we need to deify him? Is worshipful replication of his testing temple the 'one true path'? Is a test suite with a margin of error larger that of The Perfect One completely useless, or would you prefer to read more of the currently available reviews based on socket thermistor readings?

I read all the frickin' posts with no useful content, and it pisses me off. If something is wrong or inadequate, then explain why! Constructive criticism, analysis, numbers! If you ain't got 'em, then STFU!

And stop throw your damned preconceived notions on everything. Who the hell claimed a test bench could be talked down to free? Why do you think error margins will be glossed over? There'll always be trade-offs between money and accuracy. If you don't like the direction it's headed, say something specific, like "if you spend an extra hundreds buck on XX, you can reduced the error by YY percent."

Either post your own testing results, or help out here, or go away.

<end rant>

I need a drink....

[gone_fishin]

Originally posted by Groth
<begin rant>


Either post your own testing results, or help out here, or go away.

<end rant>

I need a drink....

Everyone is trying to help out. What's wrong with do it right or don't do it at all?

A work group never sets out to establish standards that are poorer than what came before. That's not progress it's called regression and a waste of resources.

[satanicoo]

Are the normal thermometers that we use to measure our own temperature (to check fever and stuff) reliable?

I know they have a low resolution, but if they measure 37.2 degrees, is it 37.2 degrees or they have an error?

[Zhentar]

since one degree fahrenheit is considerably important for our bodies, I would assume they are pretty accurate; but I don't think they'd be suitable here, because of they're designed range; 38C plus or minus 2C is all they need.

[bigben2k]

Originally posted by pHaestus
...

What I see in these threads for the most part is the following thought process:

Hmm I wanna test some waterblocks and not get flamed by Jaydee and/or BillA on the internet when I'm done. How do I test them? Hey let's look at Bill's bench testing article; I bet it explains it! Ok this seems reasonable; how much will it cost? Holy mother of god HOW MUCH? Ok I better e-mail Bill and find out if there's a cheaper way. Shit there's NOT? I better make an alliance and a bunch of posts on a forum then. Maybe with enough people involved we can talk this whole cost thing out of existance. Or at least get rid of the error bars.

...
I definitely understand that you may be used to doing things a certain way, but it doesn't exclude the possibility of doing it another way now does it?

As I stated earlier, even if we got 20 members, there is no way in h*ll that we could reasonably expect each member to fork out $1'000, so that one member can have a test bench of Bill's caliber.

As I also stated, the idea of going with multiple test benches is one of many possible directions, and the one that I envisioned for the WBTA. If you believe that you could start gathering $20'000 in contributions/donations, etc, then go right ahead, but we both know the odds of that happening.

I certainly never made any claims that we would reach Bill's level of testing, but all of us will certainly give it our best shot. It's one of those "ideal goals", a target to shoot for, sort of speak.

As for the cost issue, that's very simply an every day thing. Accuracy still remains the top priority, but cost is definitely close behind. If there is any way that I can trade "putting some time and effort", instead of buying a ready made solution, you can bet that I'm going to consider it: that's just common sense.

As I also stated (in another thread here), there may be a need for some higher accuracy, strictly for building. From a cost perspective, that leaves me with a possible option of borrowing or renting a tool.


Satanicoo: what your real question is, is "Is the resolution related to the accuracy?" and the answer is "no". A meter could have a resolution of +/- 0.1 (i.e. display), but an accuracy of +/- 0.5. That's where it gets interesting, because you have to go over the specifications of the device. Then it extends into "repeatability" and then into "calibration".

[gone_fishin]

This is a slippery fish indeed.

[pHaestus]

Accuracy is basically the product of resolution and calibration.

Stated accuracy on a product spec sheet is what should be possible sans any calibration (out of the box).

Some medical thermometers are presumably pretty accurate: keeping track of a woman's temperature as part of a way to monitor fertility would require a decent accuracy.

Groth: I was quite honestly in the mood to make smart ass comments the other day. Is Bill's data the holy grail? Nope. Does his approach make it harder for testers? Yep. He has assuredly stopped me in my tracks for publishing any test data over the last year or so as I improve my data collection and measurement at the house. Is this a bad thing? To me, no. I would feel bad if myths were perpetuated or if waterblocks were bought and sold based on erroneous data from me. Is this a bad thing to the WBTA? Not so sure. I get the vibe that this is a "good enough" type group.

Don't think that cross calibration will get you where you want to go with cheap gear. Think about this logically. If the error bars are large, and we include them as we correlate the blocks, then what have we gained? Statistics is the one true friend that tells you that your fiance is a whore.