Extremley high cpu temps with water cooling

[Jonboy_Fantastic]

I've been getting extremely high water cooling temps for years now and have triad a myriad of things in order to resolve this issue to no avail. I've bought premium blocks I've purchased great pumps. I've set up a dual pump system I've changed the orientation of my cpu block to it's recommended specification and as I have the New EK Evo block I've put in the "best" jet plate as demarcated by several overclocking sights. I've switched TIM's more times then I can remember I've delidded and added liquid ultra pro to the die. I've used two radiators one that goes to my graphics cards ( serial which isn't ideal but it gives me good temps ) and through my motherboard and set a separate dedicated loop with it's own radiator exclusively for my CPU. Yet my temps are still very high. My room is a little warm however I don't get anything near what people post, I don't even get near the posts regarding High CPU temps. Right now at idle overclocked through MSI's proprietary software ( again not the best ) to only 4.4k it's idleing at 43C-64C and the only thing I"m doing is writing this post and monitoring my computer ( Minus background programs ). Rainmeter is clocking my cpu usage at about 1-5pct. I'm doing something Horribly wrong and I have no idea what. Following will be the specs for my Rig.

CPU: I7 6700k Overclocked to 4.4ghz static ( note this cpu is NOT delidded )
CPU Waterblock: EK Supremcy Evo Elite Full Nickel CSQ Jetplate 2 The orientation, as I understand, is the best is to have the EK symbol on the bottom right hand side of your motherboard and that's where it is.
TIM: Shin-Etsu X23-7783D Silicone Thermal Compound
PUMP-RES COMB: EK-XRES 140 REVO D5 PWM ( Included pump )
Radiator: I don't remember the exact specs it's a XSPC 360 rad that came with a kit a couple years back
FANS SETUP AND ORIENTATION:Corsair Air Series SP120 High Performance Edition x6 in a push pull configuration.
OTHER FACTORS: I use primochill 1/2inch tubing and both of my reservoirs have kill coils I use no dyes are additives other then deadwater in distilled water and my loop from res/pump to cpu block is about 20in long, same for block to rad.

I know there are some things I could do differently. I have a better Radiator, a Black Ice GTX Gen Two Xtreme 360 Highest Performance Radiator that is connected to my mobo/gpu loop which used to be connected to my cpu but didn't offer better temperatures . I should purchase high static pressure fans which are fairly new due to the rise in people water cooling, I should probably purchase a better radiator for my CPU loop. I could also purchase all new barbs and reduce the diameter of the tubing ( not sure if that would help much but it might be worth a shot ) yet I'm hesitant to poor more money into this project as nothing as worked thus far.

The moral of the story is I feel like an idiot. I've been water cooling for quite some time and I'm not getting anywhere near the temperatures I see posted. I've spent a lot of time and money into this and I'm a fairly clever guy there's just something that's not adding up. I've got a premium setup and i"m not getting premium performance. Any and all suggestions advice are welcome. At this point I'm thinking of simply keeping my gpu's water cooled and buying a premium air cooled setup for my cpu; due to the fact that my water cooling setup is epicly failing. Help me please. I've never posted about this before probably because of pride and vanity and the creeping thought that I could figure out what was wrong by myself however it's been literally years and I haven't so I'm swallowing my pride and asking for help from the knowledgeable community. Thanks for your Help and I can post pictures if the information I provided isn't enough.

 

[LogicalProcessing]

I've been getting extremely high water cooling temps for years now and have triad a myriad of things in order to resolve this issue to no avail. I've bought premium blocks I've purchased great pumps. I've set up a dual pump system I've changed the orientation of my cpu block to it's recommended specification and as I have the New EK Evo block I've put in the "best" jet plate as demarcated by several overclocking sights. I've switched TIM's more times then I can remember I've delidded and added liquid ultra pro to the die. I've used two radiators one that goes to my graphics cards ( serial which isn't ideal but it gives me good temps ) and through my motherboard and set a separate dedicated loop with it's own radiator exclusively for my CPU. Yet my temps are still very high. My room is a little warm however I don't get anything near what people post, I don't even get near the posts regarding High CPU temps. Right now at idle overclocked through MSI's proprietary software ( again not the best ) to only 4.4k it's idleing at 43C-64C and the only thing I"m doing is writing this post and monitoring my computer ( Minus background programs ). Rainmeter is clocking my cpu usage at about 1-5pct. I'm doing something Horribly wrong and I have no idea what. Following will be the specs for my Rig.

CPU: I7 6700k Overclocked to 4.4ghz static ( note this cpu is NOT delidded )
CPU Waterblock: EK Supremcy Evo Elite Full Nickel CSQ Jetplate 2 The orientation, as I understand, is the best is to have the EK symbol on the bottom right hand side of your motherboard and that's where it is.
TIM: Shin-Etsu X23-7783D Silicone Thermal Compound
PUMP-RES COMB: EK-XRES 140 REVO D5 PWM ( Included pump )
Radiator: I don't remember the exact specs it's a XSPC 360 rad that came with a kit a couple years back
FANS SETUP AND ORIENTATION:Corsair Air Series SP120 High Performance Edition x6 in a push pull configuration.
OTHER FACTORS: I use primochill 1/2inch tubing and both of my reservoirs have kill coils I use no dyes are additives other then deadwater in distilled water and my loop from res/pump to cpu block is about 20in long, same for block to rad.

I know there are some things I could do differently. I have a better Radiator, a Black Ice GTX Gen Two Xtreme 360 Highest Performance Radiator that is connected to my mobo/gpu loop which used to be connected to my cpu but didn't offer better temperatures . I should purchase high static pressure fans which are fairly new due to the rise in people water cooling, I should probably purchase a better radiator for my CPU loop. I could also purchase all new barbs and reduce the diameter of the tubing ( not sure if that would help much but it might be worth a shot ) yet I'm hesitant to poor more money into this project as nothing as worked thus far.

The moral of the story is I feel like an idiot. I've been water cooling for quite some time and I'm not getting anywhere near the temperatures I see posted. I've spent a lot of time and money into this and I'm a fairly clever guy there's just something that's not adding up. I've got a premium setup and i"m not getting premium performance. Any and all suggestions advice are welcome. At this point I'm thinking of simply keeping my gpu's water cooled and buying a premium air cooled setup for my cpu; due to the fact that my water cooling setup is epicly failing. Help me please. I've never posted about this before probably because of pride and vanity and the creeping thought that I could figure out what was wrong by myself however it's been literally years and I haven't so I'm swallowing my pride and asking for help from the knowledgeable community. Thanks for your Help and I can post pictures if the information I provided isn't enough.

First of all, I'd like to thank you for actually taking the time and making a really big detailed post, it definitely helps out a lot when the contributors are trying to give their 2 cents on the subject. Now a lot of people here will tell you that air cooling is the way to go, but there is a reason they don't make air cooled cars anymore, because with the increased amount of power you need an increased amount of cooling. I love your water cooling setup, I can picture it in my mind and it makes me really happy thinking about it.

SO here comes the questions. What case do you have to house this monster cooling setup? Do you have adequate airflow - exhaust and intake - to keep the radiator supplied with a fresh supply of air?

 

[synphul]

What are the ambient room temps? Those idle temps seem high even for an inexpensive aio dual rad (240mm like the h100i) cooling the cpu. Is there a chance there's a problem with the mounting as in not getting enough pressure/contact between the cpu and water block?

The xspc 360 is a fairly thick radiator so it should benefit from push/pull unlike much thinner radiators. The sp120 fans are corsair static pressure optimized fans however there are fans with better static pressure out there. Not trying to be insulting, just covering the bases here but the fans are oriented the same direction to move air through the radiator one side to the other right? Not potentially both blowing inward or both trying to exhaust?

In terms of air vs water cooling, in theory water cooling is better as a concept as evidenced in car engines. However there's a large difference between dissipating heat from a cpu and a car's engine where the engine coolant temps while running are often 100c and other components of the engine can run much higher. Not to mention the much larger surface area producing heat vs a cpu, it's apples and oranges really. Air coolers are capable of keeping an i7 6700k within safe operating limits for temps even when oc'd and will run out of headroom in terms of vcore before temps literally can't be handled by an air cooler. If going to extreme overclocking for competition where core voltages are pushed beyond reasonably 'safe' limits they quickly ditch water cooling for ln2 out of need for sub ambient cooling which water cooling is incapable of.

That doesn't mean there's anything wrong with water cooling, just that it's not the only solution. It would also help to manually set the overclock in the bios rather than use 'ez' software/settings. The preset oc levels provided by the motherboard can often be too aggressive and set the vcore higher than needed in an attempt to target compatibility rather than efficiency. Using the lowest stable vcore for your individual cpu reduces temps as low as possible without crashing but it may not be enough vcore to keep someone else's cpu from crashing so they give it a bit 'extra' to avoid crashing. They're basing it more off a worst case scenario rather than typical or efficient voltage for your cpu specifically.

Also is speedstep still enabled so the cpu can actually throttle down when under light/low loads? If not it will run hotter than necessary which is why often times temps shown in the bios are quite a bit higher than in windows. Speedstep isn't operating in the bios and when the cpu is idling it should drop down to 800mhz rather than the full 4.4ghz.

Going by what you stated, something is likely going wrong with the install or something. Having used different radiators, different pumps, different cpu's, multiple installs/thermal paste applications etc. It doesn't seem likely to be a hardware issue, the same issue over and over. How do your water cooled gpu temps compare to others? Are they also high or is it only your cpu temps that have been plaguing you?

Have you tried running the loops without any restrictions as in with the side panel removed from the case, the radiator set someplace where it's not forced to intake or exhaust air through the case? (basically a free/open air type test isolating the radiator cooling from the case construction). Have you tried using a flow meter to verify there isn't something wrong with the pump itself on the cpu loop?

 

[Jonboy_Fantastic]

I know there is a forum etiquette and I"m completely oblivious to it. I'm not sure how to quote certain sections of certain posts so please bear with me. In the interest of readability in my limited knowledge of forum rules I"m going to answer each question in a linear fashion, please excuse any breaks in coherence. I've probably only posted on forums 3 times in my life and i"m 30.
First of all, in regards to Logical's question; my rig is housed in a Corsair 900D super tower case. As of now the thing is all over the place, not out of laziness though. I'm currently using a MSI Gaming z170A M9 ACK motherboard and to be honest I hate it, it does the job but I've had. . . . issues . . . with MSI. If this was a private post I would go into it but since it's a public one I will anyways; they hacked me ( yep, I know how it sounds believe me ). Since this incident I HATEEEE MSI and have been wanting to swap out my mobo, yet what's the price of principles, am I right? Both side covers are off. I had to RMA my Corsair AX1200I PSU which is currently outside the case because if I do cut the water line I don't want any leakage to fall on my psu in it's mounted position which is placed directly under the motherboard tubing ( when the line is cut even while the pumps are not running there is back-pressure in the loop and water will leak out of it unless you add positive pressure the opposite direction of your flow into the line and cut it in just one place ). Also the mounted motherboard blocks are notoriously hard to empty completely while my monster of a case is upright.
As far as case orientation of my case which is admittedly a hot mess, both of the Side panels are off. I have 19 sp120 corsair high performance fans and 1 sp140 corsair high performance fan throughout my rig. 3 120's are located in the front pulling air in ( This is impeded by the two hard drive cages I have installed in my case ), 1 140 is located in the back as an exhaust. My black ice radiator is located on the top of my case on the inside with three fans mounted on the top of the radiator blowing air out the top and three on the bottom blowing air in the same direction (push pull). This radiator handles the cooling loop of my 2 graphics cards as well as my motherboard. The panels on the basement of my case on both sides are removed as not to restrict airflow, as well as the mesh on the left side to allow me to mount the XSPC rad again with a push pull configuration 3 fans on both sides pushing into the case. As my fan configuration stands now my internal case pressure ( assuming I ever put the sides back on ) would be close to neutral. I should have turned them to be blowing out like they are on the opposite side of the case to achieve negative air pressure however the air blowing in my face ( my computer is fairly close to me and at about waist level while sitting ) was more then bothersome so I reversed the orientation to blowing air in for comfort. I know there's heated debate about the orientation of fans in a system ( in or out ) however I've deduced scientifically that adding positive pressure to a enclosed space with only moderate ventillation does not provide a significant convection of air exchange which ideally is what you want for heat exchange. With too much air blowing inward you're only relying on pressure in order to ventilate the space and exchange the air. With negative pressure you're relying on the fact that the air in the case ( which only needs to be moderate in a semi-enclosed space; which a computer case is ) to create a significant exchange of air. Negative air pressure also reduces on particulates accumulating in your rig ( i.e. dust ). The res and res pump combo I have 140ml each I believe are mounted on the side of the hard drive cages which further restricts airflow however I couldn't figure out how to mount two reservoirs on the back of the case due to space restrictions as well as the fact that the 900D comes with a RIDICULOUSLY large amount of grommets for running cable and time down points reducing the flat spaces that can be drilled for reservoir mounts. There is also a Single fan located on the left side of the case not attached to the radiator. The mounts are built for a 480 and I only have a 360.

To answer Synphul's questions, the ambient temp of my room is about 75-85F I have a lot of electronics in it, 2 screens a 55inch LED, 550w receiver with 5.1 surround sound and a mini fridge ( mini fridge's kick out a lot of heat ). I do have an air conditioner which is located on a window directly at the level of my motherboard which blows directly onto the BACK of my CPU this reduces the temperature of my CPU to about 26 at idle to 45-55C under load ( without overclocking and with speedsteping enabled ). I run the AC at 69F however it's fairly loud so I tend not to run it all the time. Now these temps seem good however with direct cold air, which is supposed to bring down the temperature of a room, blowing at my CPU it should be pretty good however without it I'm not running even close to ambient room temperature which makes me believe my water cooling setup is highly ineffective. The XSPC rad is 1in thick the black ICE is 2 they are both blowing in a push pull configuration. Also I don't wanna get snarky or make a thing about it but I had realized when I bought the sp120's that they were built with static pressure in mind ( which is why I purchased them ) however I was referring before to the fact that their are fans now being made specifically for static pressure to move air through radiators with relatively low airflow in CFM but high static pressure.
To answer your second question I've realized for a long time that MOST proprietary overclocking software and MSI's in particular are set way to high for a stable overclock. The reason probably being is that a crash of such proprietary software would appear like a failure for said software and therefore any auto setting with the software would be set with a wide buffer to make it seem "reliable" ( re-reading your post you stated basically the same ). That being said I haven't attempted it manually with skylake yet, as I've found it to be only moderately helpful and my problem isn't in increments but rather in gigantic steps. I have done it with Sandy Bridge, Ivy Bridge and the proprietary overclocking software for the Gigabyte z97 motherboard I had with my Haswell and then Devil's Canyon worked with moderate success. Now that you've brought it up I"m going to try to manually set the overclock for my Skylake because ignoring the fact that I hate MSI I've found their OC software sucks very badly. ( this is the 2nd MSI board I've purchased and never again ).
Speedstep is not enabled, mostly because I find I get more reliable, better and smoother performance when it isn't. I've found it isn't intuitive or quick enough to throttle appropriately and in a timely fashion for the tasks I'm doing.
My GPU temps are great as well as my motherboards. My gpu's idle at anywhere between 27-32C and they never get above 60C ( I don't believe I've ever stress tested them however I'm hyper aware of my temps at all times as I keep most of my diagnostics on my second monitor ) are overclocked to the highest stable OC. Scratch that I ran Catzilla a couple times and I think it peaked at about 70C.

To conclude thank you both for the positive and constructive feedback. You've given me a lot to think about Synphul. I'm interested in the prospect of setting up my rad outside my case and butted up directly to my AC unit, as well as manually tweaking the OC of my processor, I won't deal with speedstep ( personal preference ) however lowering my Vcore may have a significant impact in this case as I know for a fact MSI sets their software OC's too stupidly high levels. I'm also thinking about replacing the TIM. I have . . . well. . . more of them then I should; with either Gelid extreme, liquid pro ultra, or IC diamond ( which is a pain in the ass but if it'll help ). I'm also thinking about reseating my block, maybe even switching out the jetplate, and making sure the screws are tightened enough to give good contact. I'm still worried as hell it won't work. I've lapped my cpu's I've used all the aforementioned TIM's besides IC diamond mostly because it's a pain in the dick to work with and it still seems every time I read a post about someone's CPU temps using water cooling setups they are far lower then my own. At one point I got experimental and for some reason unbeknownst even to me I purchased a number of Delta 120mm X 38mm fans which run about 20 dollars a piece and kick out 240cfm a piece which did significantly cool my setup CPU and all however they run at 50db which is WAY to loud. The point being is I've tried a LOT of different things and I'm definately looking for out of the box solutions. Scratch that, <language edit> the box I'm looking for out of the box that would house the box that would contain the box people would usually want to think outside of.
I would appreciate continued suggestions comments and general observations. Thank you.



I broke this up into paragraphs, however it doesn't seem to be registering. As I stated before I'm a forum newb so if someone could post about that shit I'd appreciate it.

 

[synphul]

When it comes to styling posts in the forums, this may help some.
http://www.tomshardware.com/forum/id-2083471/read-forum-rules-styling-posts.html

Nothing special needed to format paragraphs, just use 'enter' on the kb and drop down a space as if writing in a text editor. To quote someone there's a text link at the bottom of each post with an arrow and reads "Reply to (username)". When you click it will give you a text entry field to respond with the message you quoted already there for you. You can always edit out bits by deleting the quoted text, just make sure to leave the info contained in the brackets alone since that's what formats the reply and tells the forum page which post you're quoting etc. It has an opening and closing tag, both need to be there or it 'breaks' the reply format.

Ok so your gpu temps seem fine, that loop (including your mobo blocks) seems to be functioning ok. It's just limited to your cpu. You might try speedstep, enabling it in the bios even just enough to test it out and see what your temps do in that regard. If there's no significant improvement then just turn speedstep back off again. For now it might be worth turning off the oc, reverting the bios back to stock at least until the temps are sorted out. That way there's less to rule out. I know it's only a mild overclock at 4.4 but it's worth a shot.

Have you run a program like hwinfo64 to read what your core voltage currently is? I know you used auto-oc but it would help to know how high your vcore is. That might not be helping your temps.

Did you double check that all the components of the water block were reassembled correctly when swapping out jet plates? I don't know if you still have the instructions or not, I found a copy here. Just to make sure your top plate with the inlet and outlet weren't mounted 180* off or anything. Double checking to make sure the feed and return coolant hoses are properly attached to the in/out on the block.
https://www.ekwb.com/shop/EK-IM/EK-IM-3831109800171.pdf

I'm not sure exactly which configuration you're using with the included parts, they recommend insert I1 for skylake (115x) paired with jet plate J2. It's also been a problem for some people using on the 2011 socket, not sure if that extends as a potential problem for the 115x socket or not. Check the screws on the water block that hold the base on and make sure they're flush or recessed a bit and not protruding. Someone had an issue with it, sent it back and got the same thing then EK sent them new screws. Said the screw shipment was incorrect and the screws being a little too long were preventing good contact with the cpu. May not be the case with your setup at all but worth a look.

Beyond that I'd be looking for something like a flow meter to verify that the flow is ok and the pump is working correctly. You can probably see the coolant flowing back into the pump/res combo to verify coolant is moving, does it look similar to the return flow coming back into the pump/res used for the gpu or does it look like it's returning more slowly? (Possibly indication a restriction somewhere).

 

[TMTOWTSAC]

Shot in the dark here, but where do you have everything connected in regards to the mobo fan headers? Also, what shows up in your BIOS, and does it let you control the pumps and fans?

 

[LogicalProcessing]

@synphul Is it possible his radiator cannot handle the Btus his 2 GPU's and CPU creates?

@Jonboy_Fantastic What are your 2 GPU's that you use? How many ounces of fluid does your water cooling system hold?

@synphul I am confused, if he has a EK-XRES 140 Pump, then he can pump 1500 liters an hour, then the radiator should be able to dissipate that heat. But that has to be determined by Q = Ww CP (T - Ti). ((Heat Extracted = Btu/hr = **I need to know how many ounces of water he uses** x flow rate in gpm (Exit Temperature - Initial Temperature). For example, 1 gallon of water equals 8.33 lbs, times that by 60 minutes is 500 and the pump moves 6.6 gallons of water a minute and say the discharge temperature is 140F and the initial is 113F, 140-113=27F (27F temperature change), 500 x 6.6 x 27F = 89,100 Btus/h or 1485Btus/m.

I can't find out how many Btus/h or /m that Black Ice Radiator is designed for.

Am I over thinking this or am I just exploring all avenues?

 

[TMTOWTSAC]

@LogicalProcessing

He has the GPU's in series on one loop, and an entirely separate loop (pump, radiator, reservoir) for the CPU.

 

[LogicalProcessing]

@LogicalProcessing

He has the GPU's in series on one loop, and an entirely separate loop (pump, radiator, reservoir) for the CPU.

Ahh I see that. I thought for some reason he was using it all in one.

@Jonboy_Fantastic How big is the room that you have your computer in?

 

[Jonboy_Fantastic]

Shot in the dark here, but where do you have everything connected in regards to the mobo fan headers? Also, what shows up in your BIOS, and does it let you control the pumps and fans?

None of my fans are 4 pined, therefore I can't control any of the RPM's of my fans; they are also controlled with fan splitters purchased separately connected to 12v molex cables. The D5 pumps I own are all Non-Variable speed pumps, as I wouldn't see any reason to turn down the speed of your pumps, and they run, at a their highest speed, at lower RPM's than their non-variable counterparts.

 

[Jonboy_Fantastic]

@synphul Is it possible his radiator cannot handle the Btus his 2 GPU's and CPU creates?

@Jonboy_Fantastic What are your 2 GPU's that you use? How many ounces of fluid does your water cooling system hold?

@synphul I am confused, if he has a EK-XRES 140 Pump, then he can pump 1500 liters an hour, then the radiator should be able to dissipate that heat. But that has to be determined by Q = Ww CP (T - Ti). ((Heat Extracted = Btu/hr = **I need to know how many ounces of water he uses** x flow rate in gpm (Exit Temperature - Initial Temperature). For example, 1 gallon of water equals 8.33 lbs, times that by 60 minutes is 500 and the pump moves 6.6 gallons of water a minute and say the discharge temperature is 140F and the initial is 113F, 140-113=27F (27F temperature change), 500 x 6.6 x 27F = 89,100 Btus/h or 1485Btus/m.

I can't find out how many Btus/h or /m that Black Ice Radiator is designed for.

Am I over thinking this or am I just exploring all avenues?

I really have no way of knowing how much liquid each system holds specifically other then to drain the entire system of both loops meticulously remove the motherboard and radiators and cpu block as well and to empty them all into a bucket and measure how much I get out of each. I mean I guess I could meticulously crunch the numbers and find out the water capacity for each of the radiators the reservoirs, the blocks on my motherboard, gpu, as well as my CPU block and calculate the specific water capacity of 1/2 inch ID tubing and multiply it by length but that seems like a lot of work to find out something of which I really don't see the purpose of. Understandably water would have a higher cool down rate the further it gets from the heat source and the longer it has to travel that distance however that's looking at an extremely passive form of cooling and dissipating the heat capacity of water, which may help in fractions however my heat problem isn't in fractions it's in spades. The only way I could think of something like that playing a factor ( and I have thought of doing this ) is creating a very deliberate and specifically placed flow restriction right before the radiator as you want all the water to move quickly over the CPU block as water has a very high heat capacity, higher then air which makes it such an effective cooling medium; however ideally and adversely you'd like it to spend as much time traveling through the radiator as possible because along with the fans and the specific materials that make up a radiator that is the only form of ACTIVE cooling in your loop. However to pull this off would be a feet of engineering to say the least you'd need a very specific flow restriction to reduce the flow and slow down the movement of the water right before the radiator as well as a seperate ( and extremely specific pressure outlet in order to keep the flow and speed of the water traveling through the radiator at a sluggish enough pace to make a difference as well as a big enough reservoir in order to house enough liquid in order to account for the decrease in pressure. Well I guess if you split the line right before the radiator and lowered the ID of the tubing going to the radiator with the second fork going straight to the radiator that might do it but it would be a mild increase at best. I'm sorry I'm long winded and thinking out loud.

About the 1500 liters per hour you're right and wrong those specs are for operation of the d5 non variable pump at 12v ( the variables run at 1200 liters if memory serves me correctly ) however mine don't run at 12v. I've modified dc to dc buck converters used in automotive's and crimped pins, and 12v molex heads onto them in order to step up each of my 3 pumps which are rated from 8v-24v to the maximum 24v. I've yet to find reliable data on the maximum flow rate, pressure head, or RPM's on a D5 pump running at 24v.

I'm also not sure what you're driving at, you got a lot of math their and it's impressive. However ANY 360 radiator should be enough to dissipate the heat from a cpu.

In summation, "Well sir, I don't agree with what you have to say but I'll defend to the death your right to say it".

 

[Jonboy_Fantastic]

When it comes to styling posts in the forums, this may help some.
http://www.tomshardware.com/forum/id-2083471/read-forum-rules-styling-posts.html

Nothing special needed to format paragraphs, just use 'enter' on the kb and drop down a space as if writing in a text editor. To quote someone there's a text link at the bottom of each post with an arrow and reads "Reply to (username)". When you click it will give you a text entry field to respond with the message you quoted already there for you. You can always edit out bits by deleting the quoted text, just make sure to leave the info contained in the brackets alone since that's what formats the reply and tells the forum page which post you're quoting etc. It has an opening and closing tag, both need to be there or it 'breaks' the reply format.

Ok so your gpu temps seem fine, that loop (including your mobo blocks) seems to be functioning ok. It's just limited to your cpu. You might try speedstep, enabling it in the bios even just enough to test it out and see what your temps do in that regard. If there's no significant improvement then just turn speedstep back off again. For now it might be worth turning off the oc, reverting the bios back to stock at least until the temps are sorted out. That way there's less to rule out. I know it's only a mild overclock at 4.4 but it's worth a shot.

Have you run a program like hwinfo64 to read what your core voltage currently is? I know you used auto-oc but it would help to know how high your vcore is. That might not be helping your temps.

Did you double check that all the components of the water block were reassembled correctly when swapping out jet plates? I don't know if you still have the instructions or not, I found a copy here. Just to make sure your top plate with the inlet and outlet weren't mounted 180* off or anything. Double checking to make sure the feed and return coolant hoses are properly attached to the in/out on the block.
https://www.ekwb.com/shop/EK-IM/EK-IM-3831109800171.pdf

I'm not sure exactly which configuration you're using with the included parts, they recommend insert I1 for skylake (115x) paired with jet plate J2. It's also been a problem for some people using on the 2011 socket, not sure if that extends as a potential problem for the 115x socket or not. Check the screws on the water block that hold the base on and make sure they're flush or recessed a bit and not protruding. Someone had an issue with it, sent it back and got the same thing then EK sent them new screws. Said the screw shipment was incorrect and the screws being a little too long were preventing good contact with the cpu. May not be the case with your setup at all but worth a look.

Beyond that I'd be looking for something like a flow meter to verify that the flow is ok and the pump is working correctly. You can probably see the coolant flowing back into the pump/res combo to verify coolant is moving, does it look similar to the return flow coming back into the pump/res used for the gpu or does it look like it's returning more slowly? (Possibly indication a restriction somewhere).

I took a lot of your insight to heart. I did reset the bios back to factor settings just to get a base reading. I don't have a thermometer in my room measuring ambient temp however at idle when speedstep is throttling to 800mhz I get temps of about 30-32 degrees which considering from my best estimate it's about 28-30 in my room is good. However under load at 4200mhz it's running at about 60-62. This seems like an acceptable range however it isn't really what i want.

Before I reset to stock settings i didn't check the Vcore of my MSI OC software and I don't plan to use it again so that's irrelevant. I do though upon completing this post intend to tweak the settings and try to achieve a stable overclock of 4.6ghz at around 1.35v and if that works not in temperature threshold but in stability I'm going to decrease by .01v until I get a stable overclock with hopefully the lowest temperatures I can achieve with the highest OC I can. If the Temperature is good at a Vcore of 1.35v I will step up the OC gradually to try to get to where I want to be which is about 4.7-4.8ghz. Any advice about adjusting the bclock as opposed to the multiplier would be appreciated, as stable bclock changes are supposedly more viable now with the new z170 chipset flagship motherboards ( all of them advertise it, Gigabyte, Asus, MSI ) Gigabyte calls it Turbo b-clock not sure if the others have different names for it.

I Run CPUID HWinfo and cpu-z to check my v-core and temps most of the time. I have used hwinfo64 however I format a lot and just never downloaded again the last time I did. I know the inlet/outlet of my evo is setup correctly there are no 90 degree angles and they are set up in the correct orientation. I haven't yet dissasembled my block to check the plate i know i used the correct configuration of the plate for my CPU though as the box has all the other ones except those. As for if I set it up inside the block correctly well we'll have to wait and see.

I don't think it's a flow problem as I've owned a number of D5 pumps in my lifetime and the lack of moving parts makes them hard to malfunction as well as the setup I have now has been using the same three pumps in different orientations and configurations. However investing in a flow meter wouldn't be the dumbest way I've spent money ( not that it is just I've spent a lot of money on dumb things ). I will buy one when I reassemble my loop which is going to be when I get rid of this MSI mobo ( which I hate more then I can explain in words ). I had an ASUS ROG maximus extreme VIII however I returned it as I've always been a fan of gigabyte's software and motherboards, not a fan of the new black and white color scheme on their flagship G1 but I'll take the hit for a company who's yet to let me down.

When I dissemble my loop and rebuild my comp I'm going to disassemble the water block as well and make sure everything is copacetic. Including the screws on the bottom which if I remember correctly are flush the orientation the jet plates the works.

My Final thought is the restriction problem you had. I'm not real sure if you are an avid water cooler however flow restrictions are a part of water cooling and especially and very specifically with CPU water blocks, the reason being that you are forcing a large amount of water to flow over a specific place. The only way to really achieve this is the limit the amount of volume inside a CPU block and create small channels which siphon heat away from the IHS. The Jetplates and inserts in the EVO illustrate this perfectly. Flow restriction isn't very noticeable with most other blocks/rads etc for instance the flow leaving my loop from the GPU/Motherboard isn't visually noticeable as the water you see exiting the tubing from the mobo ( the last block in my circuit ) is solid as it is when it enters. The CPU however is a different story the line moving up to the CPU is solid however the line moving out of the block is initially almost akin to a leak and there is condensation forming on the exciting tubing from the CPU block showing me there is space inside it. I'm not really sure how to remedy this, the only thing I could think of would be to put the two d5's I have connected with a bitspower dual d5 mod top to my CPU block. Or scraping the pump res combo keeping the pump and purchasing another D5 pump with a another dual d5 mod top to increase the pressure. Then again there's only so much you can force something to make it fit. I've also heard stories of putting too much pressure on a water cooling system and the weakest point gives being the tubes pop off the barbs which would be an unmitigated disaster. All the CPU water blocks I've owned have restricted water flow, the Evo though with the way it's made has been the worst. The xspc raystorm was the second worst the standard xspc block ( can't remember the name ) was the third. The only one that wasn't really that bad but still was visibly noticeable was the Apogee HD with it's 4 holes. As well I don't see the detriment in a flow restriction that happens only after it's passed through the block. Wouldn't spending longer time in the radiator increase the cooling efficiency of it? Air pockets are believed to be a huge detriment to cooling efficiency however the only real way that they are is if you get one stuck inside your block and or your rad. Other then that they are relatively harmless as an air pocket won't dissipate the heat as well however water will just flow around it.

 

[Melonious]

Your idle temperatures should stay about the same regardless of how overheated the system is. It is unreal to think that your cooling system is not bringing the heat down under idle as well.

You either have a bad motherboard or else your voltage regulators are overheating.

 

[Jonboy_Fantastic]

Your idle temperatures should stay about the same regardless of how overheated the system is. It is unreal to think that your cooling system is not bringing the heat down under idle as well.

You either have a bad motherboard or else your voltage regulators are overheating.

You're going to have to rephrase that, I don't understand what you're trying to say.

 

[TMTOWTSAC]

Shot in the dark here, but where do you have everything connected in regards to the mobo fan headers? Also, what shows up in your BIOS, and does it let you control the pumps and fans?

None of my fans are 4 pined, therefore I can't control any of the RPM's of my fans; they are also controlled with fan splitters purchased separately connected to 12v molex cables. The D5 pumps I own are all Non-Variable speed pumps, as I wouldn't see any reason to turn down the speed of your pumps, and they run, at a their highest speed, at lower RPM's than their non-variable counterparts.

I was hoping maybe the pumps got set to run slow so I guess that's out. I did have a sudden thought though while reading another thread. How big is the room, small enough to get warmer when you're stressing the computer? And what height is the computer located at? If you measure temps at floor level vs ceiling level it's not uncommon to see a 6c+ difference in a room with rising heat.

The numbers you were getting at 4.4Ghz looked ugly, but that could potentially be explained by MSI overvolting their OC. When you're running at stock the numbers look pretty good. Excellent in fact, given an ambient temp of 28-30. A lot of people with ambient temps in the 20-25 range struggle to stay below 65 with big air. With water you're within 32-34 of ambient under load at 62.

Take a look at this review here:

http://www.kitguru.net/components/cpu/luke-hill/intel-core-i7-6700k-i5-6600k-skylake-cpu-review/9/

They ran a 6700k at stock and at 4.7Ghz (1.4v) in a 25c ambient room with Corsair H100i. Their numbers at idle are substantially worse relative to ambient. Under stock load slightly better, but under OC load they hit 76c, 51c over ambient. That does make it seem like MSI was supplying more voltage than necessary.

 

[synphul]

Hard to say since you mentioned you don't have a thermometer in your room. 30-32c ambient is quite warm, equates to 86-90f. The theory of water transferring heat better is accurate, the larger radiator surface area for a decent sized radiator does indicate more efficient heat dissipation but it still relies on air cooling. The air removing the heat from the radiator and even though your system has reservoirs eventually the coolant will reach equilibrium with room temps.

It's going to hamper cooling performance blowing 86-90f air over the radiator and hoping for it to absorb the heat from it. Cooler air such as 22c/72f air temps being a higher differential compared to your coolant temp at the radiator would make the heat exchange more efficient. Might not be the entire problem but not helping matters.

 

[Melonious]

Your idle temperatures should stay about the same regardless of how overheated the system is. It is unreal to think that your cooling system is not bringing the heat down under idle as well.

You either have a bad motherboard or else your voltage regulators are overheating.

You're going to have to rephrase that, I don't understand what you're trying to say.

You don't have to understand I guess, if you don't know a lot about it you won't. That is just the symptoms that happen when your volt regulation is not working properly. And that means most of the time you need a new motherboard, but sometimes if you are overclocking it just means they are overheated.

If it were an issue of not enough cooling it would ALWAYS be hot.

 

[Jonboy_Fantastic]

Your idle temperatures should stay about the same regardless of how overheated the system is. It is unreal to think that your cooling system is not bringing the heat down under idle as well.

You either have a bad motherboard or else your voltage regulators are overheating.

You're going to have to rephrase that, I don't understand what you're trying to say.

You don't have to understand I guess, if you don't know a lot about it you won't. That is just the symptoms that happen when your volt regulation is not working properly. And that means most of the time you need a new motherboard, but sometimes if you are overclocking it just means they are overheated.

If it were an issue of not enough cooling it would ALWAYS be hot.

I feel like you're getting sore over the fact that i just requested you to elaborate on the question itself. It was short and poorly worded. I would never expect my cooling system which is based on air movement to be below ambient temperatures. However when you say "temperatures should stay the same no matter how overheated your system is" that doesn't make sense to me at all. In fact the opposite of that statement is true, no temperatures should remain the same when my system is overheated. When under load the processor heats up heating up the ambient temperature of my room the voltage regulators and a handful of other things.

My original questions was never regarding not enough cooling but rather a problem in the setup somewhere. As I stated before and if you looked up the motherboard I"m using then you'd see it has a water block built in over the mosfet-based voltage regulators of which I have good temperatures and a separate closed circuit water cooling circuit running through it. Therefore you posturing that the voltage regulators are overheating is unrealistic as according to hwinfo64 my temperatures are good under load, as well as at idle.

I suppose technically it could be that I have a "bad motherboard" however it's 5 months old and I had similar problems with my old motherboard so the probability of that being the case is unlikely. If so do you know of anyway I can find out if my voltage regulator is malfunctioning? If it was it would have to be a manufacturer's defect as opposed to overclocking damage the temperatures have always been water cooled and in extremely good condition as it stands my MB temp is 27C. And my CPU temperature is 63C. My motherboard is not overclocked but it's throttled to the max turbo boost at 4200mhz. The ambient temperature in my room is 69F.

 

[Jonboy_Fantastic]

Hard to say since you mentioned you don't have a thermometer in your room. 30-32c ambient is quite warm, equates to 86-90f. The theory of water transferring heat better is accurate, the larger radiator surface area for a decent sized radiator does indicate more efficient heat dissipation but it still relies on air cooling. The air removing the heat from the radiator and even though your system has reservoirs eventually the coolant will reach equilibrium with room temps.

It's going to hamper cooling performance blowing 86-90f air over the radiator and hoping for it to absorb the heat from it. Cooler air such as 22c/72f air temps being a higher differential compared to your coolant temp at the radiator would make the heat exchange more efficient. Might not be the entire problem but not helping matters.

I'll go into more later however I feel like information is being said by one person then telephoned down the line which often happens in long posts and forums with several things being stated. It is hot in my room however I said 30-32 was the temperature of my CPU at idle throttled to 800mhz. The ambient temperature in my room without my AC on is about 28-30C. Which again is an estimate as I don't have a thermometer in my room.

 

[LogicalProcessing]

Hard to say since you mentioned you don't have a thermometer in your room. 30-32c ambient is quite warm, equates to 86-90f. The theory of water transferring heat better is accurate, the larger radiator surface area for a decent sized radiator does indicate more efficient heat dissipation but it still relies on air cooling. The air removing the heat from the radiator and even though your system has reservoirs eventually the coolant will reach equilibrium with room temps.

It's going to hamper cooling performance blowing 86-90f air over the radiator and hoping for it to absorb the heat from it. Cooler air such as 22c/72f air temps being a higher differential compared to your coolant temp at the radiator would make the heat exchange more efficient. Might not be the entire problem but not helping matters.

I'll go into more later however I feel like information is being said by one person then telephoned down the line which often happens in long posts and forums with several things being stated. It is hot in my room however I said 30-32 was the temperature of my CPU at idle throttled to 800mhz. The ambient temperature in my room without my AC on is about 28-30C. Which again is an estimate as I don't have a thermometer in my room.

This has puzzled me all day long. I've thought about to great length throughout the day and the only thing I can think of is its your ambient temperature and environment your computer is in. Hear me out, the reason why I can get away with only 4 exhaust and 4 intake fans running a solid overclock at 4.8ghz and still only be running idle temperatures of around 31C with a small radiator is because of my environment. The computer is in the basement, its surrounded by 3 concrete brick walls that are not exposed to sunlight and a concrete floor with a rug that covers 60% of the bare floor, the room behind me is where the central air system is so I always have a fresh supply of conditioned air and my ambient temperature is always 20C or 68F. When I stress my CPU, it will go up to a maximum of 65C.

The key is airflow, if you have a steady supply of cool air coming into your room, keeping your ambient temperatures down, then your radiator's heat energy output will not be noticeable or as noticeable. If like during the winter and the heat is on but its not getting warm enough in my room, I close the door to the central air room, and put a rubber draft protector underneath the doors entering this room and I just let the computer heat the room for me. I tested this once just to see how hot and how quickly it could get. My starting temperature was 62F and within 10 hours, my room had risen to just under 83F. I heated my room by cutting off all the cool air supply from the rest of the basement.

If it is not a problem with your hardware, then you must examine the environment that your computer is in.

 

[Jonboy_Fantastic]

Hard to say since you mentioned you don't have a thermometer in your room. 30-32c ambient is quite warm, equates to 86-90f. The theory of water transferring heat better is accurate, the larger radiator surface area for a decent sized radiator does indicate more efficient heat dissipation but it still relies on air cooling. The air removing the heat from the radiator and even though your system has reservoirs eventually the coolant will reach equilibrium with room temps.

It's going to hamper cooling performance blowing 86-90f air over the radiator and hoping for it to absorb the heat from it. Cooler air such as 22c/72f air temps being a higher differential compared to your coolant temp at the radiator would make the heat exchange more efficient. Might not be the entire problem but not helping matters.

I'll go into more later however I feel like information is being said by one person then telephoned down the line which often happens in long posts and forums with several things being stated. It is hot in my room however I said 30-32 was the temperature of my CPU at idle throttled to 800mhz. The ambient temperature in my room without my AC on is about 28-30C. Which again is an estimate as I don't have a thermometer in my room.

This has puzzled me all day long. I've thought about to great length throughout the day and the only thing I can think of is its your ambient temperature and environment your computer is in. Hear me out, the reason why I can get away with only 4 exhaust and 4 intake fans running a solid overclock at 4.8ghz and still only be running idle temperatures of around 31C with a small radiator is because of my environment. The computer is in the basement, its surrounded by 3 concrete brick walls that are not exposed to sunlight and a concrete floor with a rug that covers 60% of the bare floor, the room behind me is where the central air system is so I always have a fresh supply of conditioned air and my ambient temperature is always 20C or 68F. When I stress my CPU, it will go up to a maximum of 65C.

The key is airflow, if you have a steady supply of cool air coming into your room, keeping your ambient temperatures down, then your radiator's heat energy output will not be noticeable or as noticeable. If like during the winter and the heat is on but its not getting warm enough in my room, I close the door to the central air room, and put a rubber draft protector underneath the doors entering this room and I just let the computer heat the room for me. I tested this once just to see how hot and how quickly it could get. My starting temperature was 62F and within 10 hours, my room had risen to just under 83F. I heated my room by cutting off all the cool air supply from the rest of the basement.

If it is not a problem with your hardware, then you must examine the environment that your computer is in.

Okay so I've done some trouble shooting and seemed to have at least stabilized the problem if not solved it. Some of it was me being <language edit> stupid, some of it was problems I still need to address. For one I was getting very high temps because I didn't have the fan splitting hooked up on my radiator ( duhhhhhh ). That one makes me feel like a moron. However I have now replaced my block and re-applied shin etsu micro not in the pea drop method that I used before, but in using the X method. I've tightened the screws on my Evo. I hooked up the booster to my EK pump res combo to run it at 24v. I looked through the box that the Evo came in and through process of elimination I've determined ( since I have I 2 and J1 in the package ) and it comes pre-installed with I1 and J2 which is what is recommended for 115x processors. I customized my OC to run at 4.6ghz at 1.285v Vcore. Through XTU I've clocked 85C which is a little higher then I'd like but a lot better then it was. The thing that really bugs me and that I believe is decreasing the cooling capacity of my CPU loop is that HUGE air pocket right after my CPU block.

My theory is that my one D5 even modded to run at 24v is not significant enough to run enough water through the circuitously engineered EK Evo, with all the inserts and jet plates constricting flow; as the water leaving it is only a trickle and I can see space in the loop up until it hits the radiator where it pools right before it. I'm lacking the pressure to get significant cooling capacity and it's bugging the <language edit> out of me. I'm thinking of switching my setup so my bitspower dual mod top with two swiftech d5's runs through my CPU block to see if that has enough pressure to solve this problem. I need to bleed the air out and I don't have a bleed valve or anything ( which I'm considering as well ). The only solution I can think of is raising the pressure.

A second problem I believe is even though air floats on water the hole for the intake and the outlet are located on the bottom of the res/pump combo which I believe allows for air to be recirculated through the loop. If it was a setup like my graphics card/motherboard loop which has an outlet at the bottom and an intake at the top which allows air to be bled out of the system a lot easier. Just a theory.

Thank you very much to Synphul and logical processing for your sage advice a lot of it if not directly helping me in this specific problem got me thinking in the right direction i guess? Any advice about this huge air cavity and getting rid of it would be appreciated. Thanks.

I've also thought a great deal about my environmental conditions. It gets very hot in my room like 85-90f when I'm not running AC. I'm thinking of moving my computer to the basement as underground there is a consistent temperature of 50f all year round. I do think that would help a great deal. Which got me thinking about phase change systems, and the ability to build one form scratch. It seems like an easy concept and with an enclosed space and basically the parts from an old mini fridge why not?. This would solve the problem of having to constantly run my window mounted AC unit even when it gets too cold in my room. I've also thought of peltier chips however after enough research I've determined that they are not a viable option. Isn't it an odd thing that to make something hot is a relatively easy process however to makes something cold is by comparison extremely difficult.

Also your posted temps are exactly what i"m talking about. People post temps like that all <language edit> day long and it drives me nuts because as you know I've invested a great deal of time and money into my setup and I can't get the temps others are getting, not even close. It drives me crazy. There's this little annoying 2 digit error post on my motherboard and if there isn't an error it posts the temperature of my cpu. It drives me <language edit> crazy I look at it ALLL the time and it upsets me. For instance my computer right now is running at moderate load OC"d to 4.6ghz and I'm establishing the block chain to mine ethereum, watching futurama on netflix and typing this post and my temps are hovering at 65-67c. Rainmeter is clocking my CPU usage at 17pct.

What the hell. I wouldn't say I"m a Moron, I do dumb stuff but my learning curve isn't that bad and I understand things rather quickly yet I can't make an effective CPU water loop.

Please watch the language in the forums

 

[LogicalProcessing]

Well it looks like your post got touched by synphul.

Im glad to hear that you might be moving your computer out of the scorching blazes of the a room it is in right now. Remember that you are pushing your pump past the point where it was designed to work. If you are supplying it with 24v and its rated to operate at 12v, that is the voltage the creators set it at to work efficiently. They put it at 12v because it cooled efficiently at that speed. So whatever you do now to is really can't be blamed on the pump because you have now made it your own creation...if you understand what I mean.

A phase change system would be a great way to keep it cooled. I know a lot of people go straight to liquid nitrogen but that stuff is very hard to store and very hard on the components if used for a long period of time.

Perhaps you are overthinking this, it's very easy to do. Did you try setting up this water loop to the creator's specifications?

 

[Jonboy_Fantastic]

Well it looks like your post got touched by synphul.

Im glad to hear that you might be moving your computer out of the scorching blazes of the a room it is in right now. Remember that you are pushing your pump past the point where it was designed to work. If you are supplying it with 24v and its rated to operate at 12v, that is the voltage the creators set it at to work efficiently. They put it at 12v because it cooled efficiently at that speed. So whatever you do now to is really can't be blamed on the pump because you have now made it your own creation...if you understand what I mean.

A phase change system would be a great way to keep it cooled. I know a lot of people go straight to liquid nitrogen but that stuff is very hard to store and very hard on the components if used for a long period of time.

Perhaps you are overthinking this, it's very easy to do. Did you try setting up this water loop to the creator's specifications?

I don't quite know what you mean to "the creator's specifications". You'll have to come a bit more specific than that. it's at 4.6ghz and stable and that's fine but with the hardware I have it should be putting out cooler temps. I still look nervously at that temp display on my motherboard all the time and when it hits 80 85 I just get upset. I just bought two GTX 980 ti's so I'm removing the water cooling loop from my graphic cards so I'm going to set it up so I remove the second radiator ( the xspc one ) just use the black ice radiator and only have my loop going through my mosfet/voltage regulators and my cpu block and see if that has enough pressure to get rid of that gigantic air pocket on the other side of the loop. I know the phase change systems you can buy are loud and connected straight to the CPU. My plan is to connect a smaller custom phase change unit to the radiator. Hopefully condensation isn't a problem.

I got what you meant about "creator's specifications" however. 12-24v is the operating voltage of D5 pumps. The only reason they aren't set to run at 24 volts in your computer is you'd need a voltage step up converter ( like the one I have ) to do it, because 4 pin molex cables are 12v, sata cables are 12v most of the connectors coming off a computers PSU besides the 24pin mobo connector and the cables to your graphics 6 or 4 pin cables going to your motherboard and graphics cards. Which I believe are a higher voltage but don't quote me on that. The fact of the mater is ALL D5's are made by a company called Laing as well as DDC pumps and are used for a myriad of industrial commercial and residential applications. The operating voltage of D5's is 12-24v so I am running it at creator's specifications. There's nothing that says you "should" run the pump at 12v it's only the different brands that resell the D5's with molex connectors who are assuming you are running them at 12v. Inevitably 24v will give me higher flow and more pressure. I've switched it off too and it doesn't help actually using the step up converter cools it by a couple degrees.

 

[LogicalProcessing]

Well it looks like your post got touched by synphul.

Im glad to hear that you might be moving your computer out of the scorching blazes of the a room it is in right now. Remember that you are pushing your pump past the point where it was designed to work. If you are supplying it with 24v and its rated to operate at 12v, that is the voltage the creators set it at to work efficiently. They put it at 12v because it cooled efficiently at that speed. So whatever you do now to is really can't be blamed on the pump because you have now made it your own creation...if you understand what I mean.

A phase change system would be a great way to keep it cooled. I know a lot of people go straight to liquid nitrogen but that stuff is very hard to store and very hard on the components if used for a long period of time.

Perhaps you are overthinking this, it's very easy to do. Did you try setting up this water loop to the creator's specifications?

I don't quite know what you mean to "the creator's specifications". You'll have to come a bit more specific than that. it's at 4.6ghz and stable and that's fine but with the hardware I have it should be putting out cooler temps. I still look nervously at that temp display on my motherboard all the time and when it hits 80 85 I just get upset. I just bought two GTX 980 ti's so I'm removing the water cooling loop from my graphic cards so I'm going to set it up so I remove the second radiator ( the xspc one ) just use the black ice radiator and only have my loop going through my mosfet/voltage regulators and my cpu block and see if that has enough pressure to get rid of that gigantic air pocket on the other side of the loop. I know the phase change systems you can buy are loud and connected straight to the CPU. My plan is to connect a smaller custom phase change unit to the radiator. Hopefully condensation isn't a problem.

I got what you meant about "creator's specifications" however. 12-24v is the operating voltage of D5 pumps. The only reason they aren't set to run at 24 volts in your computer is you'd need a voltage step up converter ( like the one I have ) to do it, because 4 pin molex cables are 12v, sata cables are 12v most of the connectors coming off a computers PSU besides the 24pin mobo connector and the cables to your graphics 6 or 4 pin cables going to your motherboard and graphics cards. Which I believe are a higher voltage but don't quote me on that. The fact of the mater is ALL D5's are made by a company called Laing as well as DDC pumps and are used for a myriad of industrial commercial and residential applications. The operating voltage of D5's is 12-24v so I am running it at creator's specifications. There's nothing that says you "should" run the pump at 12v it's only the different brands that resell the D5's with molex connectors who are assuming you are running them at 12v. Inevitably 24v will give me higher flow and more pressure. I've switched it off too and it doesn't help actually using the step up converter cools it by a couple degrees.

Well I am glad you understand and I appreciate that information too. In your basement where it is always cool...is it also damp? You might want to run a dehumidifier down there to keep the humidity out of the air and away from your sensitive components.