Higher than expected temps on water

[sazyario]

Hello All!!!

I need some help with my water cooling loop. My temps are higher than expected.
This is a single loop.

The build:
Phanteks Enthoo Luxe
ASUS X99-A
Core i7 5820k @ 4.5GHz (1.25v)

nVidia Titan X @ 1456MHz (110% Power Target +112mV)
16 GB (4GBx4) HyperX Predator
Corsair RM750

Samsung 840 EVO 250GB (OS & Key software)
WD Green 4TB (Plex)
WD Black 2TB (Steam/GOG/Origin)

Watercooling Components

Alphacool 420mm UT60
3x Noctua NF A14 PWM
Alphacool 240 UT30
2x Noctua NF F12 PWM
EK Supremacy EVO CPU Block
EK Titan X Nickel/Acetal GPU Block
EK 150mm X3 Res
Monsoon 3/8ID 1/2OD Fittings
(Enermax Cluster Advance 120mm on rear fan mount as intake)

1L Distilled Water

1L EK Ekoolant Clear

My loop flows counter clockwise.

Res -> D5 Vario EK XTop -> 240mm Rad -> CPU -> 420mm Rad -> GPU -> Return to Res

Ideology
GPU Pours Hot water into res then flows to 240mm rad to cool it, then to CPU which flows to 420mm rad to return cool water to GPU.

I'm seeing the following temps
40C on Titan X During Idle, 55C under Valley benchmark load
50C on 5820K During Idle, 75C under Intel Extreme Tuning
This is with all 6 fans running off of the CPU Header utilizing Noctua's splitters. I've tried running the fans from 300 RPM to 700 RPM and notice no improvement. If I increase fan speed over 700RPM I can hear the fans noticeably during normal gameplay/work. I purchased the Noctua's to have a near silent experience.

Increasing fan speed to 1000RPM I see the following

37C on Titan X During Idle, 47C under Valley benchmark load
43C on 5820K During Idle, 68C under Intel Extreme Tuning



(As a side note... on both the ASUS X99-A and my previous MSI X99s SLI Plus when monitoring CPU temps they fluctuate +-10c every few seconds under XTU. I did not experience this on my previous 3570k)

I just don't feel like I"m getting the temps I expected. I plan on adding a 2nd Titan X. However I have no more room for any additional radiators.

[shypin]

you have to run the loop for about 24 hours to bleed out all the air before the temperature can stablize

tilt the case too to help it

[Arty]

you have to run the loop for about 24 hours to bleed out all the air before the temperature can stablize

tilt the case too to help it

this

and if those temps are MAX and stable thats still fine.

[W-L]

-SNIP-

 

Did you just fill your loop give it time to bleed out air since that can heavily affect performance, also I noticed one thing in your list you said you put 1000mL of premix and another 1000mL of disitlled water, if you got premix you cannot dilute it more to make more fluid.

 

BTW you load temps are good but the idle temps are what seem to be higher than normal, what's your ambient temps.

[soaringchicken]

Yeah keep trying to get the bubbles out, make sure the pump is working properly. 

[Enderman]

Ideology

GPU Pours Hot water into res then flows to 240mm rad to cool it, then to CPU which flows to 420mm rad to return cool water to GPU.

 

Thats not how watercooling works

The order of radiators and components does not matter at all

Only the total heat output and heat dissipation of the entire loop matters

[sazyario]

Ambient temp is 71f to 74f. So, 24c or so. (Thermostat controlled window AC unit across the room. Thermo is in the middle of room)

The reason it got dilute was the leak test I did with with 2l of distilled. I drained... but not enough I found out after. (Bleeding and draining the UT60 is a PITA)
All bubbles are out. I've seen these temps for about 3 days now with the pc under a load for about 50% of that time.
Is the 50/50 mix that large of a problem?

I'd like to keep my fans around 700rpm for quiet (not totally near silent) use.

BTW my d5 vario is on setting 5. I'd like to lower it. I hear mixed reports that anything over 3 doesn't help with heat transfer.

EDIT***
I had KBoost turned on in EVGA Precision X 16... Turning that off dropped the Titan X to 33c instantly during desktop idle with CPU at 40c idle with fans at 1000rpm....

[W-L]

 

-SNIP-

 

The fluid is less of a temperature problem but a chemistry problem, your currently have half of the concentration of the biocidal agents and chemical needed in the loop.

 

The UT60 do give great higher speed performance especially with high load but even sub 1000RPm you should get fairly cool idle temps, and for the D5 you might see a degree of two of difference on high to a low setting.

[sazyario]

 

 

 

The fluid is less of a temperature problem but a chemistry problem, your currently have half of the concentration of the biocidal agents and chemical needed in the loop.

 

The UT60 do give great higher speed performance especially with high load but even sub 1000RPm you should get fairly cool idle temps, and for the D5 you might see a degree of two of difference on high to a low setting.

 

 

Well... Should I drain the loop and order more Ekoolant? I have another 1l bottle of it on hand as I knew my loop needed 2.

I will say this. in V1 of my loop. (Cpu + UT60 only) I found out that a silver kill coil will cause corrosion with the nickle in an evo supramacy. So I learned the hard way with corrosion. Now my loop is nickle and copper only. (Why do people recommend silver kill coils at all??? or not give a warning about Silver+Nickle corrosion?)

I think my temp problem was solved by turning Kboost off. I must have been flooding the loop with heat from the Titan X needlessly.

[W-L]

Well... Should I drain the loop and order more Ekoolant? I have another 1l bottle of it on hand as I knew my loop needed 2.

I will say this. in V1 of my loop. (Cpu + UT60 only) I found out that a silver kill coil will cause corrosion with the nickle in an evo supramacy. So I learned the hard way with corrosion. Now my loop is nickle and copper only. (Why do people recommend silver kill coils at all??? or not give a warning about Silver+Nickle corrosion?)

I think my temp problem was solved by turning Kboost off. I must have been flooding the loop with heat from the Titan X needlessly.

 

EKWB does give a warning about using silver with Nickel plated products since they act as borderline dis-similar metals in "harsh environments", kill coils are great with distilled water as long as you use all copper waterblocks.

 

For the coolant Ideally it would be best to have the proper premix but since you do have some I would change it out quicker instead of pushing it to the max of around 12 months, maybe 6 months at most. 

[sazyario]

EKWB does give a warning about using silver with Nickel plated products since they act as borderline dis-similar metals in "harsh environments", kill coils are great with distilled water as long as you use all copper waterblocks.

 

For the coolant Ideally it would be best to have the proper premix but since you do have some I would change it out quicker instead of pushing it to the max of around 12 months, maybe 6 months at most. 

I guess around Christmas I'll buy a 2nd 1l bottle of Ekoolant and do my best to flush all 2l in the loop out.

I'm beating myself up on this loop. Fan Speed vs. Temps. I wish I could add more radiators, but room does not allow me.

I very much so wanted to add a 120mm in the rear spot, but the fittings from the UT60 get in the way. And flipping the UT60 around is not an option as I mounted my 2 HDDs in the ODD bays to remove the HDD cages. Mounting an additional 120mm/140mm on the floor is not possible due to PSU shroud. Seems a standard ATX sized tower like the Enthoo Luxe isn't meant for as an extreme build that I chose. 

Push/Pull is not an option on the UT60. I could add a 3rd 120mm to my 240, but the pump is in the way of doing a full push/pull on the UT30.

Do you think adding a 2nd Titan X will dramatically increase temps? I could always use just 1 Titan X until the Pascal based Titan product launches in spring of 2017.

[W-L]

-SNIP-

 

You have a pretty good setup as of right now with a triple 140 and a dual 120 which is more than enough for a the CPU and single GPU, since the Titan X's are fairly high power you might start to be pushing it but still with a loop like that it should be able to manage.

 

Push pull on a UT60 isn't that big of a difference I run Push pull but only see a few degrees at most on load, It's better than have in general longer rads than thicker one since longer ones give more flow distance for the fluid to travel.