Jump to content


Boinker

Member Since 18 Nov 2008
Offline Last Active Feb 18 2019 10:53 PM
-----

Topics I've Started

Lian Li PC-011 Dynamic thermal testing.

20 January 2019 - 02:02 PM

If you guys and gals had seen that I recently got an upgrade In my PC life. You would know that I am toying with my 011 Dynamic.

 

After using this case for a while the question came to me. Is there a better way to configure the cooling for my setup? Searching the web I had mixed data from various youtubers and review write ups. This painfully concluded that I was not sure how I wanted to Configure the system. And I wasn't going to take on every suggestions as gold. So that meant one thing, time to test this myself. 

 

For all intense purposes the case layout looks a bit like this. 3 groups of 3 located at the bottom, top and front side facing away from the glass. Here are two reference photos. please ignore the pair of 960's in the setup photo. Those have been out for a while.

 

Attached File  20181230_235010.jpg   105.17KB   4 downloads  Attached File  Lian li fan layout.png   5.61KB   4 downloads

 

Testing setup: 

 

Case: 

         Lian Li pc-011 Dynamic

Case cooling:

         X9 Scyth slipstream PWM fans 0-1600 RPM w/pwm (rated @ 74cfm)

Motherboard and its minons :

         Intel 3930K processor on a Intel DX79si Motherboard supporting 16 gig Mushkin DDR3 2133

Note: The Ram is running at its rated speed in XMP and the Processor is OC'd to 4.20 Ghz from 3.20-3.50 turbo in single core turbo.

Graphics:

         EVGA RTX 2080 Black edition (Stock cooling and stock clocks)

PSU:

         EVGA 1KW PSU

Processor cooling:

         I am using a liquid cooling loop that remained static in the Front side fan locations. Watercool Heatkiller VI basic waterblock, XSPC EX360 rad, XSPC Photon tube res, Swiftech MCP655 and 1/2-3/4 tubing. 

 

Testing solution:

 

The test suit I used was 3dmarks Time Spy. For its function of testing both the CPU and GPU under the most extremes. and its free for the moment to me.

Monitoring software used was EVGA Precision X1 for the GPU, Intel XTU for the board and its overclock, Core temp used for CPU temps and tracking. 

 

To keep things even more even Steven the GPU and the CPU were optimized into a Max performance profile so the Clock speeds are constant and Not skewing the results.

 

After 2 consecutive runs with time Spy, one for warm up and second for data gather, The Max load temps will be recorded and then the system was allowed to sit in windows with only the monitoring software up for 10 minutes to establish a fair Idle reading.

 

System setup: 

 

GPU was set to its automatic fan profile by Nvidia. This puts the fan efforts around 50% at 70c and 0 while the card is idling.

 

To skew the results slightly, cause it cant be an amateur write up if I didn't, the fan profiles were set to a constant 50% with the intel built in ramp up starting from 50c and all on being at 60c. During testing it was ranged from 800RPM to 1200 RPM when the cpu ramped above 50c. In my defense, I did this because that is how I use it day to day. 

 

Lastly I did not take ambient in a manner that proper delta could be established. For this I do apologize. But according to a thermometer it was roughly 70F' next to the case when before I tested and after all testing took place.

 

On to the Data: 

 

In the manner described above I went to it. Phillips in hand we shall see what happens. I was legit curious to see and apparently ready to delete 4 hours of life for the data gathering.

See Crappy picture of fan locations for reference above. 

 

The first configuration I chose, because that is how it was to begin with, to have the top and side front fans populated as exhaust fans. 9 fans in total with the radiator sandwiched.

 

CPU temp: Max 55c    idle 35c

Max Motherboard VRM temp: 54c

GPU Temp: Idle: 47c Load: 73c

GPU fan speed: Max 54 % Min 0%

 

Second went Bottom intake with side and top exhaust. 9 fans in total with radiator only having 3 fans as outlet.

 

CPU temp: Max 57c    idle 40c

Max Motherboard VRM temp: 54c

GPU Temp: Idle 55c Load 73c

GPU fan speed: load 50% idle 0%

 

Next was The Bottom and side intake with top exhaust. This is where things get interesting.

 

CPU temp: Max 55c Idle 39c

Max Motherboard VRM: 56c

GPU Temp: idle 45c load 75c

GPU fan speed: load 57% idle 0%

 

next was No bottom fans with side intake radiator sandwich and top exhaust. 

 

CPU temp: Max 53c idle 37c

Max Motherboard VRM: 58c

GPU temp: idle 48c load 77c

GPU fan speed: load 60% idle 0%

 

Lastly I did a brute force test with fans at 100% minus the gpu. Fan location was top exhaust with side intake and no bottom.... This was excessively loud but nowhere near as bad as bottom intake setups. 

 

CPU temps: Max 50c Idle 33c

Max Motherboard VRM: 48c

GPU Temp: Idle 35c load 65c

GPU fan speed: load 48% idle 0%

 

 

With that performed lets talk about this for a bit.

 

Concluding: 

 

What is a constant here is the fact that you just cannot beat brute force when it comes to cooling performance. While impressive this case has not much a way of dampening noise or vibrations. Its so loud that it will rattle your teeth. For this reason I did not test the system much further with bottom intakes. Between the noise levels and results being either within margin of error or worse I abandoned using them in that location all together. A pc on your desk that is running north of 50 DBA is a distraction not worth getting used to. 

 

Now with the watercooling setup being where it is I cant help but wonder. Would the system perform better with the radiator placed elsewhere. With these results in hand once I have more time I have considered moving the radiator to the top and redoing testing just to get the most from my system. Movement of the heat coming from the GPU is going to make a clear difference, and in addition to that having fresh air only being ingested front the side fans is going to change that game. But how will that effect the CPU and VRM? Will a vertical gpu bracket improve this as well? Now that I have said that I will need to add the results as soon as I can allocate time to perform them. Likely next week. (January 26th tentative) 

 

With the testing performed it does prove one thing. Its an issue of give and take. With the fans as exhaust only and no bottom fans fitted we did see the best all around results. While with noise blown , haha, out of portion you can make the system cool better. What bothers me the most is that data clearly shows that bottom fans really do not serve a purpose other then blasting thy ears. You do see improvement with all fans pegged. But will suffer from hearing loss at the gain of only 5c on the CPU and 8c on the GPU. 

 

Other then that. Anyone that has taken the time to read this I greatly appreciate.

 

Finally to answer my initial question.

Is there a better way to configure the cooling for my setup? As usual the answer is, That depends.


Boinker's long awaited system update.

27 December 2018 - 11:20 PM

So its not much of a system update. More of a minor upgrade.

My family was good to me this year. Guess I was a well. Still the parts I received was much more the I deserve; But I digress..

So I will be migrating my current build from my old current HAF932 to the Lian li PC-011DW and subtracting my current video cards (x2 gtx 960's) and putting in a EVGA RTX 2080 Black edition. I still can't belive I received it and am forever thankful.

To kick off this build I decided to fit the chosen radiator that I had in store for it and the Primary motherboard That I entend to use until it expires. Cant beat intel boards. when they make them. Just wow.

With the goal of this system to bring the PC back to the desk and in a classy fashion I set off. It will take time as I still have year end stuff to accomplish and Much to take apart before I can get all things operational. We shall see if I can get it together before end of the year. But no promises.

To kick things Off I took it apart. In typical fashion It had its disappointments. To start with I installed the radiator (XSPC EX360). I had to compromise a bit. Turns out the spacing of the radiator water fittings wouldn't allow it to me mounted Behind where I want it. But this is OK. I compromised by using push pull for a slight bump in performance. To hide the radiator it would require some modifying. However I am sure this will not be the first rendition I will do. In short, its gonna get hacked up eventually.

Attached File  20181228_005018.jpg   243.55KB   7 downloads

Cant forget about the Intel skull

Attached File  20181228_005027.jpg   259.4KB   7 downloads

Lastly the back end. Will soon be unclustered only to be reclustered by my EVGA 1KW SQUID