Opinion: Water Block Testing Alliance.

[bigben2k]

I just PM'd pHaestus, JayDee116, and Robotech, about starting a water block testing alliance, since we all seem to have an interest in getting into it.

So I'm starting this thread to see if there's an interest out there, to have your block tested. (this thread is not for technical/administrative discussions, that's on private e-mail.)



In the mean time, if there's anyone else out there that would be interested in setting up a test bed and join us, PM or e-mail me. I'll work on putting up a webpage with more details. This is just starting up, and we're some ways from being able to offer this service, so be patient!

[pHaestus]

I am not sure I got the same impression from your PMs that I do from this thread. I thought the idea was some sort of cross correlation and comparison among testers. Didn't realize it was immediately scaling up to public service via block testing.

Not so sure I would be asking for blocks before you can get them tested. Learned that one the hard way...

[bigben2k]

I don't know about testing commercial products yet, but I think we could start offer user block tests. Like I said, we're a long way from there. We'd need a fair amount of valid data before being able to claim any kind of accuracy.

I was thinking about breaking up North America into districts: I've got Texas (Southwest), you've got Canada, JayDee's got the East coast, RoboTech's got the North. (Darn, we need one of those West Coast person! LOL!). Territories are non-commital, it's just a guide.

[pHaestus]

I am having a hard time seeing how this could possibly ever work. If you want to cross calibrate or get assistance from me then that is fine. However, I don't think it is in my best interest to limit the waterblocks that I test (for procooling) to be only from a certain region or of a certain number. I am sure you can understand that.

[bigben2k]

Hum, I see your point. I didn't mean to imply that we'd be limited.

Let's take it to e-mail and sort it out.

[BillA]

how about each specializing in a particular facet ?
lets see now . . . .
Ben in equivocation
jd in contention
pHaestus in delay
(Robotech I do not know)

jk jk - or not ?

[jaydee]

Originally posted by unregistered
how about each specializing in a particular facet ?
lets see now . . . .
Ben in equivocation
jd in contention
pHaestus in delay
(Robotech I do not know)

jk jk - or not ?

Joke or not I would have a hard time arguing that.

[GTA]

Originally posted by unregistered
how about each specializing in a particular facet ?
lets see now . . . .
Ben in equivocation
jd in contention
pHaestus in delay
(Robotech I do not know)

jk jk - or not ?

How is that helping?

You constantly bemoan the need for accurate waterblock testing, and yet when someone starts working towards that testing, you mock them.

Yes, you used to do waterblock testing, and it was the best around, no-one disputes that.

Then you stopped, and started working for Swiftech, so your services are no longer available.

So I'd suggest that instead of writing faintly mocking posts, with semi-veiled insults, you help someone else produce a reliable and objective testing setup.

And we all know that you consider everyone else beneth you in terms of scientific accuracy, but we're all still trying to get along with watercooling over here, and you're simply not helping one bit.

Either offer to help, or don't post, as you suggested you wouldn't after you started working for Swiftech.

Anyway.


BB2k + Co, I would like to build a die simulator and help with this project. I have money set aside to do this. If someone could tell me what I would need to get and build to help with a waterblock testing service, then I would love to help.

[pHaestus]

Bill: LOL

GTA: Relax

[BillA]

GTA
you are wide of the mark

I have a demonstrated history of HELPING all who have an interest in semi-accurate testing spanning 2+ years
- do you have any idea at all as to how many threads I have initiated and contributed to on this subject ?
-> understand that it IS necessary also to say that a scheme will NOT work

I know Ben, I know jd, and I know pHaestus; and I'm sure they understand my comment

was there a 'cutting edge' to the comment ? sure, such is an element to all humor
was there also truth being expressed ? you bet your butt

jd has already concurred
why don't you poll Ben and pHaestus ?

correction rejected, sorry

[arcsylver]

In all seriousness,

This does sound like a good idea considering the limited resources availible for such testing.

That and the level of accuracy often demanded by the users of this forum would most assuredly end up pushing the quality of those tests up a notch or two.

As for me The rig I am planning currently would be an excellent Source for a standardised rig if I can get and accurate measuring device for flow and such of the water at different points in the system such as before and after the waterblock etc.

All I should need to do is install temp probes at key areas of the system such as the input and outputs of the radiator and waterblock for reference purposes (calibrated of course) to get an idea of the amount of heat being disipated.

Unfortunately we all know that in rig testing is sketchy at best, but is all most of us have.

Perhaps If I could get an idea of what is needed to produce an accurate die simulator for the various AMD and Intel chips I could see about adding a test bed rig into the cabinet and use it as a as close to real life setup as you could get with a die simulator by placing it inside a case with limited airflow and such or even in a sealed enclosure to remove airflow as a possible source for error.

I for one would be willing to help out with this if anyone may be interested in a discussion regarding the feasibility of such a setup.

[joemac]

Bigben this is a great idea!! There are many that say that this won’t work. It is because of this that there is no “testing alliance” of any type. Baby step are often talked about but no one has been willing to take them. One suggestion would be to run the “testing alliance” through member ship. It may start off small but that how many things start off big? Great idea about breaking it up into regions. I say go for it.

[pHaestus]

German waterblock makers have been doing this for a while: They all meet up and test out their blocks on a single testbed. Would this sort of testing meet with cheers or jeers in the US? Would people pay enough money to have a weekend meeting somewhere if they knew that engineers and representatives from many companies would be there and with their newest waterblocks to test? Maybe. Quakecon and many other big lan events started off similarly. Here are the impediments though:

1) Could a single testbed be agreed upon by all parties? By majority? Could test parameters be agreed upon?

2) Would mfgrs show up? Would they send blocks?

Would it be possible to have what amounts to a water cooling convention with testing of blocks being one of the main events I guess is what I am asking. Your thoughts? It would cost real money to make something like that happen.

[winewood]

Ok, I for one think GTA is absolutely right.
(preparing for flames)

If someone says, I can get the right numbers, and basically through intimidation, tech speak, using impossible standards and everyone else is irrelevant, they are the problem in this scenario. Many people are perpetuating this as well.

EVERY time I see someone attempt to get basic numbers, they are attacked here. Therefore no one wants, nor will post numbers AT ALL if they know whats good for them. They wait for someone to discredit them for self gratification and then all the assclowns come to pipe in on a blanket party. How does this help anyone? Other forums not filled with roadblocks are helping each other reach a good understanding of what THEY can do to be more accurate. In the meantime they find what is accessable and attempt to make it better. This one is stalemated completely (in case you haven't noticed) pH and Jaydee will fail, because i doubt neither will have an exact replica of Bills testbed, nor will they be helped to find the easiest and most fair way (possible to a normal individual) with anyone elses help.

1st of all there IS a way to get basic simple numbers that are predictors of how a block will work without flushing $20K down the drain for a test setup.
2nd No ONE person has all the answers, and cannot dictate to everyone else what is acceptable and what isn't.
3rd There HAS TO BE a way to find a good predictor of performance or a guage.
4th Sitting in one spot spinning your wheels is a sure way to accomplish nothing.

BillA stated he wouldn't "help" when he joined Swiftech? (I don't know or remember) Well thats exactly what is happening regardless! Attempt to convince the world that any result not blessed by ONE technique is completely invalid. That just isn't reasonable nor rational. He is accompishing his goal of self inflation and glorification by stating everyone else is CRAP, and as seen in this thread being a *sarcasm*-> great help in getting this moving toward an objective. A constant crap on everyones parade isnt the glue that joins anything together. Positive constructive feedback is.

Now BigBen2k wanted HELP on an objective. BillA wants the world to belive he is the messiah and standing alone on a platform and is making sure that BigBen will never reach a goal that doesn't leave Billa in the center of everything. BillA you are paid to do your job and use your equipment. I find it highly suspect that now, no one can measure anything anymore.

Frankly a good 'ol posting of I get X on my system with one block, and X on my same system with the other. There is direct and indirect meathods avaliable. SOMETHING is better than nothing.

Now is it BillA's job to help us? No. Is it our job or interest to find an answer? Yes. Ask yourself in honesty. Who is helping people get to a goal of a testbed and who is putting a stick in the spokes of the wheel? I appreciated your contributions Bill. I really did. But please stop this game of yours. Only your employer seems to gain since frankly they are the only ones we will never see any critisims from you on.

Now show me someone whom I can respect who will be willing to make testing accessible to the masses, and not be intimidated. If I make people mad with this, at least I may see some action on BB's resolution..

If you disagree, feel free to post. I have been here for long enough to speak my peace and make up my own mind on my observations. Flame at will.

update: just saw pHaestus' post. This is a classy move. A simple orchestrated testbed is a great idea.

[pHaestus]

winewood you don't know what you are talking about regarding Bill. He privately has helped me a great deal, and now that he is at swiftech and not testing blocks for overclockers he has offered to cross calibrate with my test system and JoeC's at overclockers. In theory, that would put Bill's results, mine, and JoeC's all on the same page. What more do you want?

I HAVE acceptable gear for testing and just cant get the time to put up results. And when I get close with using a CPU I always fry something. Sad but true. I cannot come forward and sing campfire tunes and tell everyone they can get meaningful data with a $20 compunurse and a bucket for flow rates. It just can't happen. There are plenty of people who are doing qualitative waterblock testing, and for example Cathar has gotten by quite well with that. If you want truly quantitative work though then simple overclocker tools aren't going to get you there. No matter how much you want to make it so.

[winewood]

Very well stated man. I just get ticked off sometimes. I see so many roadblocks, but no one solution oriented.
If Bill has talked to you privately thats cool. Frankly I wouldn't know anything about that.

In theory, that would put Bill's results, mine, and JoeC's all on the same page. What more do you want?

I want more than 3 people who can have access to some numbers or the ability to produce them.
(No slant on your efforts at all!)

I'm not really stating that we need 100% accurate measurements. I'm just wanting some measurements that are good enough to tell apple from orange with a higher reproduceable occurance to the average person. Right now the average person cant be served at all.

[bigben2k]

GTA: you're in

Arcsylver: sorry, you're out. This is about test benching, not system testing.

pHaestus: I guess I got a little carried away, originally, because I ultimately want to provide user testing. I believe that there is room for each of us to build our test benches any way we see fit.

Bill: Howdy! How's California treating you?

This Alliance is about cross-calibration, to give each member some backing to our results. As I explained to JD (privately), if I test a block and come up with a C/W of 0.20, with a +/- 0.01 error margin (yeah right, dream on!), and pHaestus tests the same block and comes up with 0.21 to 0.23, then I know that my rig is running right (almost), and that it's running on the high side. If the results were off, then either I or pHaestus (or both) have problems, and we have to fix them. By cross-calibrating each other, we'll be closer to a true mark, and we'll each know how valid our test bench testing is.

The Alliance is also about sharing the technical aspects of testbenching. I may not be willing to build a die simulator, so I might ask JayDee to make me one (and pay him for it, of course). I might forego the pleasure of designing a good thermal probe, so I might query pHaestus for a diagram, or other sources.

As for Bill, he's already invested more into testing than any of us ever will (together!), and his contribution (if it becomes available) would be his test results of his calibration block. With it, we can cross correlate our results even further, and tweak our test benches even further. No one should doubt Bill, he's just naturally grumpy. Right now, I have no hope of attaining the level of accuracy that Bill had.

[jaydee]

Originally posted by winewood
Very well stated man. I just get ticked off sometimes. I see so many roadblocks, but no one solution oriented.
If Bill has talked to you privately thats cool. Frankly I wouldn't know anything about that.

I want more than 3 people who can have access to some numbers or the ability to produce them.
(No slant on your efforts at all!)

I'm not really stating that we need 100% accurate measurements. I'm just wanting some measurements that are good enough to tell apple from orange with a higher reproduceable occurance to the average person. Right now the average person cant be served at all.

Easy on Bill guys. I would expect nothing less than him pointing out the flaws in this and in ourselfs. That is pretty relevant to the thread even if it comes out sounding harsh. At least I thought so anyway.

Solution oriented? It isn't like we havn't been looking. We just are not finding one. I have pulled my hair out for a year now trying to think of a way. Just like everything else, quality costs. You get what you pay for. I bought a couple $20 TC with multimeters and I get that qulaity of results. It works for me sure, but that is pretty much where it ends. I don't think I would feel comfortable writing a review on my results. Although I think it would be considerably superior to some of the crap I see daily. Still doesn't make it right though.

Anyway...

Is there more than 3 people up to doing it? Problem I see is people get all gun ho about doing it and then they slowly start realizing how difficult it really is (myself included). Not only money and equipment wise but TIME and EFFORT wise.

Sure there are plenty of people out there that have no issue taking a quick 30minute test and using the onboard probe and writting a review of it. We see that shit every day and I think for the most part the members here are quick to pick up on the flaws. BUT what about the thousands that do not know any better that read those reviews and then make a purchase on it only to come to these forums and get told they been had?

I got banned arguing this very thing recently. This one site was hell bent and trying to prove you don't need anything better than the onboard probe and "we make our tests repeatable for our readers". As long as there are idiots like this running sites and posting reviews there really is no hope as they out number sites like this 100+ to 1. Hell just look at [H]. They been doing it that way how long now? And they are still one of the major sites there is.

All we can really do is improve what we can do on this site (and/or sites we are apart of) and hope it starts trickling over to others. Although I really do doubt it because it really takes more effort than most are up to doing.

I was more than ready to dump the money in a decent test bed with $300 magnetic flow meter, $300 Lab quality thermoprobes and logging station, $200 for a good variable power supply with qulaity gauges that have been calibrated, probably another good $50 to finish the die sim, pressure gauges, water temp controller (very expensive) etc...

I then started to add up all the TIME it would take to do a decent test on a block with this stuff or ANY stuff. That is pretty much what kicked my ass. It would be like working a second full time job. Time I decided would be better spent on other things (if I had the time in the first place).

Well I had a rough evening helping my brother move. Wore out and time for bed..

[jaydee]

Originally posted by bigben2k
As I explained to JD (privately)

I didn't get your e-mail yet Ben. My mail server is dead it seems... Will try and respond tomorrow.

[winewood]

Bill, Man I can't sleep. There is no way you are to blame for as much as I heaved on you on that earlier post. That was a personal attack if there ever was one. I'm having a world record bad day.
One man can not account for my percieved flaw in communication on a worldwide accessable server. I am sorry if I was unfair and out of line. :shrug:

[pakman]

Kudos to you all for getting this started and I hope at some point completed...

Phaestus's suggestion is excellent if we could get all the waterblock manufacturers to agree... a true waterblock wwf style smackdown .

Now that being said, I am very curious to know ball park range, what figures are we talking about to get 1 testbed setup that would please you masters of watercooling... JD threw some numbers out, but roughly are we talking US$1K, or more like US$10K or more to get a proper watercooling test bed setup?

[bigben2k]

A test bench can be built at various levels, but the cheaper it is, the less accurate it will be. ~1'000$ would be a minimum, really. Bill's rig was more than $20'000.

We might be able to come up with cheaper ways to do things, I don't know yet.

[BillA]

Whinewood
your diatribe reveals only your relative ignorance on this topic
if you wear long trousers, I suggest you edit your carping

and here I was going to refer you to a 3yr old thread on the hard forum - but alack, it is no more
be that as it may, I have been helping those with an interest for a long time
- at this moment I am 'assisting' a tester in New York, and another in Calif
these are individuals who contacted me and, in the context of their budget and acumen,
I'm helping to get the best bang for their buck

so you are unhappy that the bar has been 'set too high' ?
wake up, I did not set it there; have you heard of ASTM, ASME, etc. etc.
the scientific community sets these standards, NOT yours truly

pakman
make a list of the US mgfrs, now put a check next to all those that have - in any fashion - revealed performance data on their products

some have, some have not - why ?
for starters, who wants to be 3ed ? or 5th ?
and the test setup, who has a 'technically acceptable' one ?
(yes Winewood, here I have done some substantial damage - and will do more !)
and apart from the intrinsic capabilities of the test bench, the more real difficulty is the time
it takes me 20hrs to test a wb,
for while a single point may provide a basis for comparison, such says little about the wb's actual characteristics

the $s have been discussed repeatedly, do a search
or perhaps someone bookmarked them
the 'high roller' approach can be seen in the equipment list on my old site (sig)

[Blackeagle]

Winewood,

You must have had a terrrible day.

I've learned more from this site, with it's high standards, than from ALL OTHERS COMBINED !

This is not a accident.

I have been slloooowwwly buying parts to build a new rig. I have the cooling components. Not one is the same as it would have been a year or so ago before I found this site.

Back when I first came here this was my idea of a truly great loop:

Maze 3
BlackIce Xtreme
Ehiem 1060
Lines would have been 1/2" clearflex and barbed fittings.

What I now have:

Cascade (copper)
2-342 (the one BillA tested along with one from Sence87)
Iwaki MD-15
Lines now will be 5/8" copper air conditioning lines with a small amount of 5/8" & 1/2" clearflex right at the blocks where I'll need it.

That doesn't even touch on the differances in the way I will now route the lines. Or the improved DIY res with 3/4" line into the pump's inlet.

With all the differances involved I'd say the system I'll now put in use will beat that first one by over 10C. I think you have to argee that is a huge improvement. Well worth the time invested.

BillA, Cathar & pHastus had a great deal to do with saving me from myself, although that is a very short list of those who have helped me learn.

[Since87]

Originally posted by bigben2k
GTA: you're in

Arcsylver: sorry, you're out. This is about test benching, not system testing.

So Ben, you've appointed yourself the head of this organization?

Might I suggest, Secretary/Treasurer would be a more appropriate role.

As for me, I'd like to apply for the position of Lead Vaporware Designer.

Edit: BTW "vaporware" has nothing to do with phase change.