WTH is this for WC setup?

[Pisklink]

Some clanmember of me thinks he knows a lot of computers and watercooling and het made the following setup.

http://i39.tinypic.com/2s8psbc.jpg

http://i40.tinypic.com/dtfzm.jpg

http://i43.tinypic.com/2569t7t.jpg

http://i44.tinypic.com/2ikurdv.jpg

http://i40.tinypic.com/96ko4n.jpg

http://i41.tinypic.com/2ut71aa.jpg

I'm completely clueless at what he has done. He says that he has setup all components in parallel instead of serial and claims it provides him with a substantial temperature drop. Furthermore did he claim that using a direct connection between his pump and rad will gain another temp. drop.

According to me he is completely nuts and sure as hell has to deal with restriction problems (in his head maybe?). What's your opinion and did you ever saw a parallel setup? And if so, did it gain that substantial temp. drops?

Please be honest since this guy is always someone that tries to know everything better and I hate that!!!!

[coolmiester]

snip
I'm completely clueless what he has done.

Me to as all the pictures aren't working!!!

[NaeKuh]

pictures not working either.

[nlancaster]

picture failure.

[sirheck]

It is the new Swiftech Stealth cooling set up.:D

[Captain_Harlock]

Personaly, I've never found the little red x's to be very efficient at cooling, and they can be quite restrictive to boot.

[shazza]

Personaly, I've never found the little red x's to be very efficient at cooling, and they can be quite restrictive to boot.

lol ... I'm seeing blue question marks - I think that's good for another few degrees temp drop.


Seriously, I'd be interested in seeing this setup ...

[sniperbob]

No one can see the pictures, as they are hosted on your clan's website and permission is needed to access. I suggest going to photobucket.com and uploading them there.

[Pisklink]

@ ALL

LOL!!!! Damn already 7 replies. You guys are all to curious or just plain fast. I think I can get in the Guiness Book for getting the most replies in 30 seconds. :D

I noticed as well but haven't had the time to correct my mistake. Fixed it now, so let those comments rollin'. :D

[NaeKuh]

he branched off out the outlet of the ehiem.

Where all those tubings goes i have no idea or clue.

But its a horrible paralell.

If he got better temps, i blame it due to the cleaning.
There can be no other explanation.

[sirheck]

Ok i see the pics now, and all i got to say is
What the... :eh:

I see a bay res with a Ehiem style pump and a triple rad
with what looks like un-needed ports and different size tubing.:shrug:

[Pisklink]

he branched off out the outlet of the ehiem.

Where all those tubings goes i have no idea or clue.

Me neither. I asked him for more pictures or at least a schematic to getter a better idea. If I have it tomorrow or the day after I will post it af course.

But is it even smart to set up parallel loops with only one rad and one pump?

[coolmiester]

Hey, as long as he is having fun and nothing got fried who are we to condemn or condone his fun :shrug:

Its a prime example as to why people don't try different things (and post here) if its going to be posted on XS only to have everyone give the guy grief for trying something different.

All the OP wants IMO is ammo to go back and ridicule the kid which i won't be part of :shakes:

[Delirious]

looks like a terrible cluster fudge. That 3 way fitting is probably killing what little flow he had from that ehiem

[shazza]

Hey, as long as he is having fun and nothing got fried who are we to condemn or condone his fun :shrug:

Its a prime example as to why people don't try different things (and post here) if its going to be posted on XS only to have everyone give the guy grief for trying something different.

All the OP wants IMO is ammo to go back and ridicule the kid which i won't be part of :shakes:

While I see where you're coming from, it's a bit too easy to paint "everyone at XS" with the same brush, and maybe the OP just wants to know if it could be true.

I like to see people experiment - but there is a valid question if the builder is saying his loop layout gives him a significant temperature improvement. If it does, then we should learn why - if it doesn't, then :shrug: I don't really care how he sets up his system, so more power to him.

[Pisklink]

All the OP wants IMO is ammo to go back and ridicule the kid which i won't be part of :shakes:

First of you don't know the kid, who is actually 45 years old and therefore expect him to have some common sense. Second I'm curious if what he says can be or is true because most of the fime he is full of sh*t. Third I can only encourage him to try something different as well but what if his experiment could be a danger for his components or he is wasting his money on something that's not going to work at all. Don't you think it would be fair to let him know?

Moreover he is telling me about his setup as if that is the only true reality. I would like to face him with some good arguments why that is not necessarily the case.

Another thing he was convinced about was that I had to use 1/2" ID and 3/4" OD tubes instead of 7/16" ID and 5/8" OD tubes because those tubes won't disperse that much heat into the case as compared to tubes with thinner walls. Isn't that a bit weird when you first make a different loop setup to get rid of as much heat as you can and at the same time want tubes that isolate as much heat as possible. I would rather suggest to get thinner wall tubes and let the airflow in the case take care of that. Overall I think it's all "bullsh*t" for our purpose of gaming. It would be a different story maybe when going for a world record OC.


While I see where you're coming from, it's a bit too easy to paint "everyone at XS" with the same brush, and maybe the OP just wants to know if it could be true.

Indeed!!!!

[coolmiester]

While I see where you're coming from, it's a bit too easy to paint "everyone at XS" with the same brush, and maybe the OP just wants to know if it could be true.

I like to see people experiment - but there is a valid question if the builder is saying his loop layout gives him a significant temperature improvement. If it does, then we should learn why - if it doesn't, then :shrug: I don't really care how he sets up his system, so more power to him.

For all i would love to agree with you shazza, this is a systemic trend that has been going on here on XS for years now.

The main players didn't just find another hobby or somewhere else to post for the good of their health but like myself, maybe got a little miffed at the stifled outlook here but i didn't want to turn this thread into a rant though not to see it is blinkered or purposeful IMO.

Anyway, i also would like to see and hear more of what this guy is up to but let the guy post his own thread :shrug:

I have absolutely no agenda at all other than to let the guys that think outside the box, post outside the box without being scared off at the first hurdle.

[Rise]

i can see what he did, and i don't get what's weird about it... if you spend 2 minutes looking at the pictures you'll see that it's a perfectly viable setup albeit a very messy tubing job vs. what we're used to seeing here on XS in the build log sections/WC gallery. it's not about flow here, the changes he talks about are relative to what he did before he put it in parallel. if he had all of that in series, he was carrying the temperatures from the CPU to GPU1 to GPU2 (or vice versa is even more likely) and then to the rad. But if his rad is able to cope with the heat dump requirements (we have no idea what CPU/OC he's got, if any nor do we know anything about the gfx cards), he is now instead delivering cool water to all three locations instead of the first only. He also seems to have a good fan setup with the air being pulled in from outside the case + shrouded through the radiator and not knowing anything about the fans, they could be doing a great job of removing the heat. the only recommendation i would make to him is to put screens over the fans since he won't see his radiator clog up with dust due to the positioning.

for your "furthermore" if he is trying to say that he will hook the pump up to the rad before going to the blocks, then yeah he'll see another minor temp drop if his rad isn't maxed out already because it will also get rid of the heat dumped into the loop coming from the pump.

on XS, you often see highly over clocked components putting out massive heat and requiring massive radiators to remove all of the heat. again, here we don't know any of that info and its certainly plausible he saw temp drops, especially if the loop went from his pump to GPU2 -> GPU1 -> CPU because the GPUs tend to run at much higher temps than the CPU, especially 2 of them in a row, so then yeah - its very obvious the parallel will give his CPU better temps. it's also most likely that he was talking about his CPU temps because people tend to care about them the most so this scenario would make sense.

[coolmiester]

First of you don't know the kid, who is actually 45 years old and therefore expect him to have some common sense. Second I'm curious if what he says can be or is true because most of the fime he is full of sh*t. Third I can only encourage him to try something different as well but what if his experiment could be a danger for his components or he is wasting his money on something that's not going to work at all. Don't you think it would be fair to let him know?

Moreover he is telling me about his setup as if that is the only true reality. I would like to face him with some good arguments why that is not necessarily the case.

Another thing he was convinced about was that I had to use 1/2" ID and 3/4" OD tubes instead of 7/16" ID and 5/8" OD tubes because those tubes won't disperse that much heat into the case as compared to tubes with thinner walls. Isn't that a bit weird when you first make a different loop setup to get rid of as much heat as you can and at the same time want tubes that isolate as much heat as possible. I would rather suggest to get thinner wall tubes and let the airflow in the case take care of that. Overall I think it's all "bullsh*t" for our purpose of gaming. It would be a different story maybe when going for a world record OC.



Indeed!!!!

Cool, he is the same age as me so why not let the kid / guy try his own methods instead of pushing yours (and anyone else's you can find to jump on the bandwagon) onto him :shrug:

[StAndrew]

Another thing he was convinced about was that I had to use 1/2" ID and 3/4" OD tubes instead of 7/16" ID and 5/8" OD tubes because those tubes won't disperse that much heat into the case as compared to tubes with thinner walls. Isn't that a bit weird when you first make a different loop setup to get rid of as much heat as you can and at the same time want tubes that isolate as much heat as possible. I would rather suggest to get thinner wall tubes and let the airflow in the case take care of that. Overall I think it's all "bullsh*t" for our purpose of gaming. It would be a different story maybe when going for a world record OC.


I dont think the tubes conduct all that much heat. I deff dont think they make a difference either way.

His approach is unorthodox, however that isnt necessarily bad. As long as each side of his loop are roughly equal in restriction, he should be fine. As stated above, he should probably optimize his tubing a bit better. This design introduces lots of extra tubing into his loop which can be a bit detrimental to his flow rate.

[Silverhawk]

Having temp drops is insignificant if we don't know how much it actually dropped from and what his system is running at. If he was going at 70C and dropped to 60C, that would be a good temp drop, but wouldn't be a good temp for a stock system.

[prava]

For all i would love to agree with you shazza, this is a systemic trend that has been going on here on XS for years now.

The main players didn't just find another hobby or somewhere else to post for the good of their health but like myself, maybe got a little miffed at the stifled outlook here but i didn't want to turn this thread into a rant though not to see it is blinkered or purposeful IMO.

Anyway, i also would like to see and hear more of what this guy is up to but let the guy post his own thread :shrug:

I have absolutely no agenda at all other than to let the guys that think outside the box, post outside the box without being scared off at the first hurdle.

I see where are you getting. XS still stands by its name...but in some senses it has lot the "science" to back everything up. What I mean is that many times nothing is explained, just a weird comment about something being wrong but not WHY that is wrong. There are still people who make detailed explanations, though.

So, what I want to say is that the guy who designed and created that circuit is total and absolutely wrong, and I'm going to explain it by rebating what other people in this very thread have written.



i can see what he did, and i don't get what's weird about it... if you spend 2 minutes looking at the pictures you'll see that it's a perfectly viable setup albeit a very messy tubing job vs. what we're used to seeing here on XS in the build log sections/WC gallery. it's not about flow here, the changes he talks about are relative to what he did before he put it in parallel. if he had all of that in series, he was carrying the temperatures from the CPU to GPU1 to GPU2 (or vice versa is even more likely) and then to the rad. But if his rad is able to cope with the heat dump requirements (we have no idea what CPU/OC he's got, if any nor do we know anything about the gfx cards), he is now instead delivering cool water to all three locations instead of the first only. He also seems to have a good fan setup with the air being pulled in from outside the case + shrouded through the radiator and not knowing anything about the fans, they could be doing a great job of removing the heat. the only recommendation i would make to him is to put screens over the fans since he won't see his radiator clog up with dust due to the positioning.

for your "furthermore" if he is trying to say that he will hook the pump up to the rad before going to the blocks, then yeah he'll see another minor temp drop if his rad isn't maxed out already because it will also get rid of the heat dumped into the loop coming from the pump.

on XS, you often see highly over clocked components putting out massive heat and requiring massive radiators to remove all of the heat. again, here we don't know any of that info and its certainly plausible he saw temp drops, especially if the loop went from his pump to GPU2 -> GPU1 -> CPU because the GPUs tend to run at much higher temps than the CPU, especially 2 of them in a row, so then yeah - its very obvious the parallel will give his CPU better temps. it's also most likely that he was talking about his CPU temps because people tend to care about them the most so this scenario would make sense.

First of all, loop order: unless you are talking about massive overclock with, specially, massive VGA's it's pointless to put any special care in loop order but to take care about how long the hole circuit is. Every little inch of tubing adds pressure drop to the loop, which means that the shorter the circuit, the better. Also, 45º and 90º add pressure drop. You know, 1+1 = 2. By the way, the heat a component puts into the water is measured by 0.X ºC, which means it's not much at all and, in most of the cases, we can even put it out of the equation.

Second, parallel loop. Running a parallel loop is not a problem at all...if you split it where its needed to be. Many components in a LC loop doesn't care much about flow, mainly the rads...but if there is one single block that does it is the CPU one. Now, if we take in mind that he is running and uber long, restrictive and complicated single loop the worst thing you can is to split it up between cpu and VGAs because you are halving (maybe even more than halving) the total flow the cpu gets, which is not good.

So, all in all, he has taken a wrong aproach, specially seeing the pump his using and the massive amounts of 90º he is using. He deffinitely can't have better temps this way than running a single loop, which would have no elbow at all (or one, maybe two) and would be deffinitely shorter. Take in mind he is not using i7 nor specially new VGA's, which means his heat load is probably modest (the cards are shorter than the mobo, which means they are under 10"), and a triple rad is more than capable of dealing easily with all of that.

[Waterlogged]

I sooo wanted to avoid posting in this topic but I couldn't, sooo

["jumping on" :p:]

"Significant drop" from air cooling maybe? :shrug:

Done right, parallel can work reasonably well. . .having said that, I don't think this is done right. In most cases, to properly balance the flow between the various lines that make up a parallel loop like his, you need flow meters and valves on each line, which is why you don't see it very often. It gets more expensive than water cooling already is real fast and the gains are rarely worth it. Short sections like what we've seen as of late with GPU blocks are fine without the extra stuff because they are already "equalized".

I can't even make out what those small tubes are going to and what blocks those are. :shrug:. . .and what Koolance CPU block is he using? If it's a 350, that could be the reason he saw something of a drop in temps (notice I didn't say "significant" ;)). It's a restrictive block and by switching to a parallel loop, that weak Eheim pump can improve the flow to it by diverting a smaller portion of the flow to the other stuff in the system as compared to pushing a full serial loop. In reality, this is where most sane ppl would use 2 loops. . .or real water cooling pumps.

Those crappy Triebwerk fans were tested by Martinm210 and were outworked by Scythe 1850 GT's on a rad that has less restriction than the one he is using. Yes, that's right, a 25mm thick fan beat a 55mm fan that was "designed for radiators"(:rolleyes::sick:), in static pressure. :ROTF:

[/"jumping on" :p:]

[Pisklink]

Cool, he is the same age as me so why not let the kid / guy try his own methods instead of pushing yours (and anyone else's you can find to jump on the bandwagon) onto him :shrug:

We are just having a discussion on our clanforum and he brings up this topic and tries to lay his methods upon me but at the same time not explaining why this could work or provide figures about how well it works. Due to my lack of knowledge about parallel setup I tried to gather information about it by posting a thread overhere. So I'm not pushing him or anyone else. To quote myself:

Please be honest since this guy is always someone that tries to know everything better and I hate that!!!!

I asked you to be honest when replying even if this means that I was wrong that a parallel setup is useless with regard to temps, flow, and costs. I'm sorry if you (and maybe others) interpreted that the wrong way. It's actually the other way around that he is pushing me with his methods and is claiming my setup is useless. And yes, that frustrates me after several years. I'm just asking for arguments so I can give him information for improvement or tell him that he could be right after all with his setup.


Anyway, i also would like to see and hear more of what this guy is up to but let the guy post his own thread :shrug:

He doesn't want to because he is convinced that what he does is always the best method. That's what I was trying to explain to you in my first post. So he doesn't see the importance of putting up information in here or asking for any. I thought stubborn is the correct word for it, isn't it?


I have absolutely no agenda at all other than to let the guys that think outside the box, post outside the box without being scared off at the first hurdle.

Me too and I love to see what creative people across this board are experimenting with. But if he doesn't ask it himself I will because I'm curious if what he claims is true and could work. Is that so hard to understand?!?


Having temp drops is insignificant if we don't know how much it actually dropped from and what his system is running at. If he was going at 70C and dropped to 60C, that would be a good temp drop, but wouldn't be a good temp for a stock system.

I asked him his results and will post them here together with system specs.

[StAndrew]

Those crappy Triebwerk fans were tested by Martinm210 and were outworked by Scythe 1850 GT's on a rad that has less restriction than the one he is using. Yes, that's right, a 25mm thick fan beat a 55mm fan that was "designed for radiators"(:rolleyes::sick:), in static pressure. :ROTF:
[/"jumping on" :p:]

Not to mention the price. I bet those fans cost almost as much, if not more then the radiator itself :rofl:.

[xslix]

what each has it own sep in's n sep outs that meet back to one into rad lol

[NaeKuh]

So wait CM your not saying thats a horrible Parallel?

He's using 2 different types of tubing first off.

There is no flow regulating the paths of parallel, not to mention, he probably didn't do a proper flow analysis for the parallel to work.

If he's getting better temps on the cpu its because the other blocks in the said branch of paralell is being starved.
There is no fanboy RD-30 bling bling.. EK FC ism going on.

Its a straight analysis of a parallel that the op posted, and it straight up sucks... O.o

I understand your situation tho, but, sometimes there is no nice way to be objective.
Sure you can say.. umm i think you need to rework your system... but being objective is one sided... thats how i always thought of it.

OP its a horrible paralell, if you want to get into why its horrible ontop of the main reasons i gave you above, i'll be glad to go at it with ya.
But like what CM said, id hate it if you took the comments we made, and took it over there to flame the Owner.

But if ur naturally curious, sure, i'll tell you all the places it fails.

[Pisklink]

OP its a horrible paralell, if you want to get into why its horrible ontop of the main reasons i gave you above, i'll be glad to go at it with ya.

I thought so as well and asked him for some more information about his rig (see information below).


But like what CM said, id hate it if you took the comments we made, and took it over there to flame the Owner.

Listen CM was insinuating that I was planning to do that. But we (the two of us) are always discussing computer related technical stuff since we are practically the only ones in our clan who read and tend to know more than average about computer stuff. For a good discussion you need arguments. When he told me about his setup I had a subjective opinion that I thought it was a bad setup. But subjective information, opinion or feeling is not enough to convince him (and should he? I think not).

Furthermore I'm 30 years old playing in a eldery clan, do you really think I'm still that childish to start flaming and ranting his methods / opinions. Although I get frustrated sometimes by his stubbornness, I still try to convince him with descent arguments and if he doesn't do anything with it I'm fine as well. It's still his machine and not mine. However I think it's my "moral" duty to prevent that he makes some vital mistakes and kills his machine. Besides that it would be ashame, we will have to play clanwars with one person less while we already having problems to win with all these grandpa's. :D


But if ur naturally curious, sure, i'll tell you all the places it fails.

So yes, I am really curious!!!! :)

@ ALL

I asked my fellow clanmember to post what he changed on his setup and what temps he got. Furthermore I will receive more information about his OC's and realised temp. drops. Below you can see that he changed much more than only his loop. He changed some critical components as well. I think it is legitimate to say that his undefined drop of temp., he's talking about, could come from this substitution.

Before
http://www.koolance.com/water-cooling/images/more/cpu-300-v10_p1.jpg

After
http://www.koolance.com/water-cooling/images/more/cpu-350ac_p0.jpg

Before
http://www.it-service.be/webshop/images/4640_0.jpg

After
http://www.it-service.be/webshop/images/hx-cu1020h_p0.jpg

Before
http://www.it-service.be/webshop/images/xspc-bm-silv.jpg

After
http://www.it-service.be/webshop/images/ddbayresdouble.jpg

Temps:
Room temp.: 22.5
System temp.: 27
CPU temp.: 27
Mobo temp.: 35
SSD temp.: 27
HDD's temps.: 34 and 35
GPU1 temp.: 37 (7950GX2 water)
GPU2 temp.: 53 (GeForce GTS250 air)

System specs:
Mobo: Foxconn Destroyer, bios versie: MCP72A01 79CF1P16 090618
CPU: QuadCore AMD Phenom X4 Black Edition 9950, 3000 MHz (13 x 231)
Mem: 8GB GeIL CL5-5-5DDR2800 5 - 64 bit
GPU: 7950GX2 watercooled and GTS250 air cooled