Zuhl3156

Methodical2
Zuhl, I saw your thread on the Swiftech forum and noticed the Swiftech to increase the motor speed to help bleed pump. How did you do that?  Is the pump speed adjustable externally on the pump or did you adjust it via MB pwm speed?

I have the pump hooked up to the CPU-1 fan header on the motherboard and have it running at full speed now as recommended by most watercooling users. I have the radiator fans hooked up to the CPU-2 fan header and have a custom profile controlling the fans according to CPU temperature. This is all done through the HW Monitor function in my motherboard's BIOS.

Why do they suggest it run at full speed as opposed to being controlled via pwm?  I had the pump installed on one cpu header the fans on the other.

I don't remember where I had seen the post about pump speed but the argument was made that if the water was moving slowly it was also passing some of the heat it removed back to the device being cooled. It was this reason that I run the pump at full speed although from the looks of that whirlpool in my reservoir I could lower the pump speed a little bit and still have plenty of flow. I have the fans on a temperature controlled curve because it is the airflow through the radiator that removes the heat. The hotter it gets the more airflow I need.
 
I was just thinking that a piece of clear aquarium hose would work nicely on that bottle for adding fluid too. I wrote that warning on the bottle in case I have another heart attack and my sister comes into my apartment and tries to sprinkle some of that green goo on her salad. LOL

HWMonitor is showing my pump at 3000 RPM while the Swiftech spec sheet shows 4500 RPM as the max speed for this pump. I would hate to think that someone at Swiftech soldered a dropping resisitor on my power input to slow the pump speed and lower the noise level. They wouldn't do that, would they? Nah, maybe?
 
http://www.swiftech.com/MCP50X.aspx
 
"Variable speed control via PWM thru the motherboard, from 1200 rpm for completely silent operation, to 4500 rpm for ultra high flow performance"
 

It's a pretty solid little unit. At least they didn't drag their feet for you like they did with my RMA. I've been thinking about the noise being really loud when the pump is defective is caused by the plexiglass housing acting like an accoustic chamber similar to a bass subwoofer uses it's housing to amplify its sounds. Just a thought.

Zuhl3156
I don't remember where I had seen the post about pump speed but the argument was made that if the water was moving slowly it was also passing some of the heat it removed back to the device being cooled. It was this reason that I run the pump at full speed although from the looks of that whirlpool in my reservoir I could lower the pump speed a little bit and still have plenty of flow. I have the fans on a temperature controlled curve because it is the airflow through the radiator that removes the heat. The hotter it gets the more airflow I need.
 
I was just thinking that a piece of clear aquarium hose would work nicely on that bottle for adding fluid too. I wrote that warning on the bottle in case I have another heart attack and my sister comes into my apartment and tries to sprinkle some of that green goo on her salad. LOL

This is a good short "pump speed" basics article. 
The theory made in this article is that too slow of a flow = laminar flow in the coolant and heat is not removed as well.
http://www.overclock.net/t/1108918/what-can-my-pump-handle-a-guide
 

Zuhl3156
.....It would be too complicated and time consuming IMO for someone to start fiddling with pump speeds and benchmark just to get the best cooling level although I can picture someone doing this. I think this is called being 'Obsessive Compulsive'.

HEY!!  I have done that before!  Calling me OCD sheesh. 
 
It is solid info but with modern parts until you are going insanely over the top, getting your loop into laminar flow is not a concern to most components now.  The pumps, tubing and blocks have seemed to find their equilibrium with each other and as long as you are building in the realm of what most people  on an enthusiast forum such as this would call a clocked SLI gaming rig, not something for competitive clocking/benching it can all be done on a single loop and have no worries.  Sure you could go and buy 3x 655 Strong's per loop in a dual loop config and have a major problem, but in general as long as you don't let the cooling shop take your WHOLE paycheck and don't have a bad habit of applying the logic of "If one is good, two is better and three is AWESOME!" (which imho is true for video cards) to all of your components you will be fine.
 
Personally I have a lot of hardware and a big case so I run dual loops, and I like it, I run a little cooler than I did with 1 loop, the big difference is for me, ease of maintenance and making it more modular.
 
The number cruncher in me loves reads like that, but the gamer in me says "you've built and benched it already, time to blow stuff up", which is where i am at now.

Ok, I'm back.  The new unit is working properly and my water loop is doing just fine.  The loop temps are 2* higher than ambient and the 980 Ti idles 2-3* above ambient (1* above the loop temps) and the cpu, well it fluctuates a lot and ranges from 29-35*, per HWMonitor.  
 
Initially, the pump was very loud and concerned me, however, I was able to dial down the rpms, via Asus Ai suite, as it was running higher than I wanted.  The noise level is right around when I had all the air cooling fans in the system.  The only problem I have with Ai Suite is it limits me to 20% rpms speed and doesn't allow me to go below that.  I would like to take it to 10%, so that I could lower the pump speed a bit more; pump is spinning at 1800 rpms, but I'd like to get it to around 1300-1400 rpms as I've found that spinning the pump faster (above 30%) only netted about 2* difference, but the noise level was not good.  
 
Btw, Zuhl my pump speed got 4411 rpms, per Ai Suite test.
 
I plan to build a gaming only rig in the near future and I'm seriously thinking about not using a PWM system that way I have complete control over the fan and pump speed - any thoughts on this?  
 
Again thanks for all the help and advice and photos...Al
 
Update:  I ran the speed test again with the pump connected directly to the CPU fan header and it reached 4623 rpms, so it seems there's some drop going thru the Swiftech connector.

I use RealTemp to shut down if my CPU exceeds 90°C and GPU at 70°C. Set the 'Alarm' temps and then select 'All Files' and choose the 'Shutdown.bat' file and it will shut it down if it gets too hot and sound a very loud klaxon type alarm. I setup a Task in my Task Scheduler to start RealTemp automatically whenever I logon. When creating the Task check the box to 'Run with highest priveleges' so you don't have to authorize Adminstrative Mode every time it starts. Just Copy and Paste the extracted RealTemp Folder to your C drive and direct your Task Scheduler to the .exe file there when creating the Task.

My CPU never goes over 72C when benching. Intel's thermal protection doesn't engage until 99C so I give it a few extra degrees of protection. If my CPU ever reaches 90C it means my H220 has failed. Since I put my video boards on their own separate watercooling loop they have never gone over 50C when Folding at Home 24/7 so if they ever reach 70C I know my pump has failed. My H220 did fail once and my CPU was running at 99C but everything was choppy. That was before I started using RealTemp and is the reason I started using RealTemp to shutdown if overheating occurs again. If I hadn't got up and needed to do something on my PC my CPU probably would've burned out.

EVGATech_DaveB

Zuhl3156
One thing that article points out that I didn't take into consideration is that the pump itself also contributes heat to the loop. I notice that most but not all loops send coolant directly from the pump to the radiator but my logic told me that the hottest components were the video boards. Not wanting to send hot coolant to the pump, I decided to run my flow directly from the pump into the video boards and then to the radiator to be cooled before going back to the pump. Any thoughts, Dave?

Honestly, I don't worry about it.  My GPU loop has dual pumps.  This is a bit overkill but it is basically res-pump-rad-GPU-rad-pump-gpu-gpu-res.  (I messed with several different configs, and this did cool best, granted only by a couple degrees, is it a big functional difference, NO, but it definitely does appease my OCD)  So I have hot (warmest water in the loop) hitting the res and pump (which are integral to each other) and have no issues.  Now since I have been working on other things I haven't clocked the GPUs yet, but so far all I have seen is a max of 3c variance between first probe after rad 1 and 4th probe on the return line to the res.  The temp difference is so minimal it will not effect anything.
I would tube it to be most convenient for you, and to leave you with the easiest maintenance options.
Sorry I didn't get back to you sooner, I was pretty tied up this weekend.

Mine is similar Res, pump, rad, cpu, rad, gpu, res, pump.  Dave, which case are you using?  
 
Zuhl, I will check out Realtemps.

I found a noisy defective Swiftech fan on my H220 while I was turning them around to blow air into my case instead of pulling it outwards. I felt that my radiator fans were being starved for airflow no matter how hard I ran the case fans since I had both rad fans blowing hot air out of the case. My CPU on the H220 is running 5° cooler already. Blowing the warm air from the H220 onto aircooled GTX-680 video boards caused them to heat up to 76° and throttle. When my CPU is under load at 55° the air coming from the rad was only warm while the air from my video board radiator was like a blast furnace when they were at 50° under load. We shall see how watercooled GTX-980 video boards do with that arrangement later on when I do some Folding at Home. Here's wishing myself luck.
BTW, I used Cougar Vortex fans on the H220. They're just a tad loud right now but not too bad. It's kind of like sitting next to the fridge when it's running. I can lower the RPM and tweak the speed later or buy quieter fans.

Yeah, it is slightly larger than something like the Hadron.  I got a lot of hardware, completely outgrew my old HAF932 by a lot, and I wanted to build for overkill cooling, also I don't see the best on small details and I got gorilla thick hands and arms..... so I got something with room to spare.... and I have nearly filled it.... it's kinda a challenge, here is a case that is way to big that you can never fill with things you will actually use... CHALLENGE ACCEPTED!
 
But it gave me the room to build one of the more overkill loops I have ever seen, so lots of headroom for whatever I want to add.  Got room for another 360 if I wanted to, but right now, I think im pretty good, as it takes Linpack to get my water more than 3c above ambient.
 
2 Loops right now.
Loop 1: XSPC Bayres+655 - BI360 Extreme - Apogee XT (thinking of upgrading it as it si good but could be better) - Mobo block by Nateman_Doo - BI360 Extreme - VRM block by Nateman_Doo - res/pump
Loop 2:  XSPC Bayres6+655 - BI360 Extreme - 980HC - BI360 Extreme - 655+EKtop(no res) - 980HC - 980HC - Bayres.  QD4s as well.
 
Will add QD4's top the CPU loop, don't have any yet.  Vent fans are on a NZTX grid, and the fans for the rads are controlled by Lamptron FC5v2's.  Tried V3's, advertised as a V2 with 2 more channels and a better lcd screen and voltage monitoring.  Well, caps are not to the spec the V2's were, I have blown almost all the caps off of 4 of them, v2's just keep working, the V3 LCD is WAY better, can read it at an angle better, more color options so you can make it fit your look and your eyes better, but the potentiometers were not of the same quality and tended to make my loop sound like it was breathing.  Oh yeah, and the rad fans are Panaflo's in Push/pull when at full speed you can feel them about 6-8 feet form the case, going full blast on those REALLY pushed the temps down.  Normally I run them around 1000-1200 RPM and they are really quiet that way but you can feel the pressure form the air, just not far form the rad, basically I got the power when I want it/need it otherwise it can sit there and be fairly quiet.  Granted 1 leg of my GPU loop fans seems to not be working gotta check that out.

EVGATech_DaveB
Yeah, it is slightly larger than something like the Hadron.  I got a lot of hardware, completely outgrew my old HAF932 by a lot, and I wanted to build for overkill cooling, also I don't see the best on small details and I got gorilla thick hands and arms..... so I got something with room to spare.... and I have nearly filled it.... it's kinda a challenge, here is a case that is way to big that you can never fill with things you will actually use... CHALLENGE ACCEPTED!
 
But it gave me the room to build one of the more overkill loops I have ever seen, so lots of headroom for whatever I want to add.  Got room for another 360 if I wanted to, but right now, I think im pretty good, as it takes Linpack to get my water more than 3c above ambient.
 
2 Loops right now.
Loop 1: XSPC Bayres+655 - BI360 Extreme - Apogee XT (thinking of upgrading it as it si good but could be better) - Mobo block by Nateman_Doo - BI360 Extreme - VRM block by Nateman_Doo - res/pump
Loop 2:  XSPC Bayres6+655 - BI360 Extreme - 980HC - BI360 Extreme - 655+EKtop(no res) - 980HC - 980HC - Bayres.  QD4s as well.
 
Will add QD4's top the CPU loop, don't have any yet.  Vent fans are on a NZTX grid, and the fans for the rads are controlled by Lamptron FC5v2's.  Tried V3's, advertised as a V2 with 2 more channels and a better lcd screen and voltage monitoring.  Well, caps are not to the spec the V2's were, I have blown almost all the caps off of 4 of them, v2's just keep working, the V3 LCD is WAY better, can read it at an angle better, more color options so you can make it fit your look and your eyes better, but the potentiometers were not of the same quality and tended to make my loop sound like it was breathing.  Oh yeah, and the rad fans are Panaflo's in Push/pull when at full speed you can feel them about 6-8 feet form the case, going full blast on those REALLY pushed the temps down.  Normally I run them around 1000-1200 RPM and they are really quiet that way but you can feel the pressure form the air, just not far form the rad, basically I got the power when I want it/need it otherwise it can sit there and be fairly quiet.  Granted 1 leg of my GPU loop fans seems to not be working gotta check that out.

Dave do you think it's best to control the fans manually vs using PWM (CPU header)?  The Swiftect pump gets loud when it gets over a certain rpm.  I wanted my fans to go higher, but keep the Swiftect pump rpm slower and steady, so I moved my fans off the CPU header and put them onto my FC5v2 that way I can spin them how I please.  Like you, when I crank them up I can drop 4-5* and the noise is not an issue because I can't hear them over the game.  I plan to do an overkill system (liking the Caselab SMA8 case as it kind of reminds me of my HAF 932, which I really like the looks) and was thinking of going with a non-pwm system for total control.  What's your thought on this?
 
Btw, that's good information about the Lamptron V3s.  I have 2 FC5v2 and can move one over to the new build if I have to.
 
Thanks

Zuhl3156

Methodical2
 
 
That's not good.  So, it sounds like the air flow was fighting each other.  I had a similar issue when I was running the Swiftech 220X; the 980 Ti was blowing hot air up into the case and into the H220X, but instead of pulling air into the case, via the H220X, I pulled it out the top to keep the hot going in it's natural path; up.  As a result, I think the temps were hotter than it would've been with a reference cooler which blows hot air out the back of the case and not into the case.  I debated pulling air through the H220X and into the case, but it would've been fighting the 980's air.  But now that's history with the full water loop.  I am officially hooked.

I was hoping that blowing cool air into the case from the front and the back would work but I could never get enough flow even at full speed. I could never blow the air upwards from my video board loop into my case. On the other hand, one 50° CPU on a loop equals warm air that I can blow into my case. Two 50° video boards on a loop equals fire breathing dragon.

Have you considered using some of the high pressure fans as case fans to get things moving?  I like having all my fans on the fan controller, so that I can run them as I see fit.  I initially had them on the CPU header but since moved them to the fan controller.  Now, I can run the fans at max, but let the cpu run the pump at a slower speed - this keeps the pump noise down quite a bit and I can knock off 4* and the fans are not that loud at max.

Zuhl3156
I have two headers for the CPU, CPU1 and CPU2. I can set the pump speed using CPU1 and run the fans on a custom profile on CPU2 both controlled by the CPU temp. I posted at MSI trying to find out what temperature controls the smartfan curve for the motherboard's SYSFAN headers to control my case fans. It doesn't look like anyone knows for sure if there is a sensor on the motherboard or it uses the CPU temp. I tried the Cougar Vortex high static pressure fans but they were too loud.
I got a pair of these for the front: http://www.amazon.com/gp/product/B00F6S13DE?psc=1&redirect=true&ref_=oh_aui_detailpage_o00_s00
I got one of these for the rear: http://www.amazon.com/gp/product/B00F6S0XK8?psc=1&redirect=true&ref_=oh_aui_detailpage_o01_s00
 
They are quiet and are moving a ton of air through the case now.

Sounds like you have a nicer MB that allows you to better control the pwm fans.  My CPU headers are the same.  I guess pretty much all the newer MB have this capability.  I placed those fans on my wish list and will take a closer look at them when I start the new build.
 
Thanks