Testing will begin today.

[nikhsub1]

Sigh, finally got everything setup the way I want it. I moved the whole test rig into the kitchen area (which is not a kitchen, it's a dark room). See my office is a studio apartment - the kitchen was gutted in favor of a darkroom. A custom stainless steel counter was made - it is made to have water rinsed on it, it is about 8 feet long with a built in sink, nothing more perfect for the water cooler :D . Anyway, there are no windows in there (well there are but they have been boarded up) so there is no draft and no 'sun' heat to deal with. It is a pretty ideal room in that way. I finally received an MCW 6002 which has NEVER BEEN USED! w00r! And for $20 you can't beat it with a stick! Blocks in which I currently have in my possession; Storm G4, D-Tek Fuzion, G5, Apogee GT (will be tested with both o-rings)and the MCW 6002. I do hope to get an MP-05 LE at some point and I had a brief discussion with Swiftech about getting a GTX top (don't need the whole block). All blocks will be tested with 3 mounts and at 2 different pump speeds. The Iwaki RD-30 will be run at 18v for one hour and then at 24v for one hour per mount. Some may be thinking why bother, just pick a point and use it but I want to see how the blocks scale. Fans for all testing will be 12v - that is over 300 CFM of fannage on the HE 120.3. Here are some pics of the setup.

Chronos if you see this, could you calculate an estimate for my CPU's heat output? It's at E6600 that I'm running at stock speeds but with 1.55v pumped through it (1.55 set in bios). Im also using TAT to generate load which aparently goes ~20% ABOVE the TDP rating! So a Stock 6600 would consume ~78W or so.

Now for some pics:

http://www.anonforums.com/builds/teststation/setup/full.jpg

http://www.anonforums.com/builds/teststation/setup/angle2.jpg

http://www.anonforums.com/builds/teststation/setup/angle1.jpg

It looks like the probe is sitting on the towel, it is actually in the center of the rad.

http://www.anonforums.com/builds/teststation/setup/probe.jpg

http://www.anonforums.com/builds/teststation/setup/temps.jpg

[Hard H20]

Very nice setup, looking forward to the results!

Subscribed.

I've got my Apogee GT gathering dust in the closet now, swopped it out for a FuZion.

[Sideroxylon]

Cool. I look forward to seeing some of your results. Couple questions...

Did you calibrate those analog sensors and with what? I've found my analog sensors off by as much as 10c and average 5-7c.

Just block testing or do you plan to expand to rads and whatnot?

I'm not sure you can say you live in a studio apartment. Doesn't having your other rig in a MM case qualify that as an extra room? ;)

[flou]

I hope you have enough courage to spend that time to do all the tests, it should take to you too much time!

[R B Customs]

Thats a pretty neat test rig youu have there and nice to ee you testing a big range of blocks too.
any chance of running say an RBX or TDX through its paces to see how it keeps up with the big boys today ?

good luck, and happy testsing.
Rick B

[nikhsub1]

Cool. I look forward to seeing some of your results. Couple questions...

Did you calibrate those analog sensors and with what? I've found my analog sensors off by as much as 10c and average 5-7c.

Just block testing or do you plan to expand to rads and whatnot?

I'm not sure you can say you live in a studio apartment. Doesn't having your other rig in a MM case qualify that as an extra room? ;)
Just block testing, have no interest in testing rads. The sensors are not calibrated to something known to be accurate no, but they are calibrated with one another meaning they read the same temp in the open air or with both in water.

[thunderstruck!]

I assume you'll also be logging idle and load temps via TAT?

[cadaveca]

Nice sink!

[Bun-Bun]

*jealous*

[nikhsub1]

I assume you'll also be logging idle and load temps via TAT?
Idle temps are useless and will not be measured. All loaded CPU temps will be read by TAT and load will be generated by TAT as well.

[NickS]

w00t, I can't wait to see how the 6002 does. I'm using a 6000 w/doubled up tube for my Conroe and it does pretty well... but I can't wait for a direct comparison. :woot:

[SiGfever]

Sweet rig.

Subscribed.

[Linux_Box]

Sounds great. Need any items to help? Maybe we can help get stuff together.

[MaxxxRacer]

Scott, it might be helpful to measure the actual cpu voltage under load to get a more accurate C/W value.

If you want to borrow my differential pressure meter or thermometer drop me a pm.

[JoeBar]

Eagerly waiting...

[nikhsub1]

Well it takes me about 4 hours to do one run of each block... My procedure is to mount the block, pump at 18v, load the system with TAT for 1 hour prior to taking any readings. Start of hour 2 is when I record the 18v run. After 1 hour I save the TAT log then I crank the RD-30 up to 24v. I then let the system run for 30 minutes. After 30 minutes at 24v I start to take readings for the next hour. Unmount, mount and repeat. Fun times. FWIW I got a much more consistent air/water delta with the setup where it is now. I am logging ambient and water temps but will use Water/CPU deltaT as 'it'. It seems realistic that I can do one pass of one block per day. That means it will take me 3 days to complete each block. 10 blocks would take me 30 days IF I did it every day, which I do not. 1 Block (3 mounts) per week is my minimum. Run 1 of G4 is done, w00t! Tomorrow if there is time I will do run 2.

[MaxxxRacer]

I take it the ambient temp in that room is very constant? (ignore my horrid grammar)

[qdemn7]

Subscribed, neat set up, looking forward to the results. Nice to see you using that HSPC Test Station. :toast:

[One_Hertz]

I take it the ambient temp in that room is very constant? (ignore my horrid grammar)

Why does the ambient of the room matter when he has the water temps logging? Or does he? I reread it twice and couldnt find the answer. Its getting late though...

[Doubage]

Nice setup.
Thanks for the effort. Look forward to the results.

[Quazi]

Subcribed. Looking forward to these results. Appreciate your efforts m8!

[MaxxxRacer]

becuase if he was in the middle of testing a block and the ambiant or room temp went up, it could cause the temp of the cooliant to rise, thus the CPU temp. Skews the results.

The more constant the ambiant temp, the more constant and fair the test results will be.

andyc

what he said.. the more stable the ambient, the more accurately you can compare temps between different test sessions and waterblocks. It just eliminates any issues with compensating for fluxuating air temps.

[Etacovda]

er, if hes doing delta cpu to water, it doesnt really matter too much, as long as its not wildly variable.

[MaxxxRacer]

er, if hes doing delta cpu to water, it doesnt really matter too much, as long as its not wildly variable.

I know. But at certain times of the year my room temp can vary as much as 15C throughout the day and night, which with all certainty will jack up the test results.

Now if the room stays within a 3C or less range all day, thats perfect (well not perfect, but considering Scott isnt doing lab testing here its good).

[nikhsub1]

I know. But at certain times of the year my room temp can vary as much as 15C throughout the day and night, which with all certainty will jack up the test results.

Now if the room stays within a 3C or less range all day, thats perfect (well not perfect, but considering Scott isnt doing lab testing here its good).
I'm looking at my first test's air and water temp log. Ambient temp greatest variation throughout the 18v test is no more than 1.4C. Water temp variation throughout entire test is no more than 1.1C. I think we gonna be ok :D For 24v test air variation max is 1.1C and water is .3C (if you can believe that!)

For the 'unscientific' testing I'm doing, we're gonna get very good results me thinks. At least the results will be gathered on equipment that everyone can easily get (or already has). I kept hearing people cry for 'real world' results, this testing is exactly what these will be.

[SiGfever]

nik,

We appreciate the amount of time and trouble it will take you to do these test, not to mention the expense. :toast: :toast: :clap: :clap:

Edit: Is that "Zym???" the polish that you use after you lap the CPU?

[stormshadow]

the wires going up to your temp display in conjunction with the bottles of liquid on your top shelf maka me nervous :stick:

[nikhsub1]

the wires going up to your temp display in conjunction with the bottles of liquid on your top shelf maka me nervous :stick:
Perhaps you need to seek help? :D

[Graystar]

Just wish we could tighten up the temperature accuracy. +/- 1 C on each sensor is +/- 2 C, which is a lot.

[nikhsub1]

Just wish we could tighten up the temperature accuracy. +/- 1 C on each sensor is +/- 2 C, which is a lot.
What exactly do you mean? I'm using the same sensors for every test. Who talked about accuracy?

[Graystar]

Each sensor has an accuracy of +/- 1 C, right? Two sensors...that's an aggregate accuracy of +/- 2 C.

[nikhsub1]

First of all im going to be using Water/CPU delta, I'm logging air for the heck of it. The resolution of the devices is .1 but accuracy is not known. I would suggest for this test that it makes little difference being that the same conditions will apply to all tests.

[Graystar]

First of all im going to be using Water/CPU delta, I'm logging air for the heck of it.
Gotcha. :thumbsup: I would think Air/CPU would be more important, but hey it's your test. :)


The resolution of the devices is .1 but accuracy is not known.
All the specs I've seen for CompuNurse devices give the accuracy as 1 C. So I'm simply assuming your sensors are 1 C sensors as well.


I would suggest for this test that it makes little difference being that the same conditions will apply to all tests.
Well, ya...unless results end up being within 1 C of each other. Then you don't really know.

[stormcloud]

What size of tubing are you using? It looks like the flumes at my local swimming pool:D

[Okda]

gonna be watching ur threads for results :)

[Grinch]

any results?

[Cyprio]

Subscribed and 5 Stars Nik. This is what XS is all about imo not all that flaming.








... maybe someone could send you a Cuplex Di? :stick: :D

[NaeKuh]

Ohhhhhh.... im gonna be waiting for this.

Going to see if all that noise is really worth it @24V. What i saw when i ran mine was i get better temps @20V.

But lets see!

[NickS]

I wanna see the MCW6002 bust out some pwncakes!

[ranker]

That's why I love XS over other forums like [H]: Lots of testing and collaboration. I'm looking forward to seeing the results.

By chance, do you know if the 3rd pin in the power plug for the RD-30 is an RPM meter? Naekuh and I have been wondering... this way I can create a wire and hook it up to the mobo/BigNG for monitoring.

[NaeKuh]

That's why I love XS over other forums like [H]: Lots of testing and collaboration. I'm looking forward to seeing the results.

By chance, do you know if the 3rd pin in the power plug for the RD-30 is an RPM meter? Naekuh and I have been wondering... this way I can create a wire and hook it up to the mobo/BigNG for monitoring.

i have???

But now that you bring this up. I am... :yepp:

[NickS]

Guys, seriously, cut it out. Don't hijack this thread. Ranker, it's time to drop it, NOW.

[Grinch]

any results?

was curious to see any results...

[Bun-Bun]

Yeah what is the progress on these tests? I am looking forward to seeing the results :D