Project R-TYPE | Watercooled Build
- Thread starter R-TYPE
- Start date
Ok here are some pics of the system filled up. It's all working well (I think!). Have ordered some sunset yellow pastel to add into the reservoir to make the fluid a little more orange to match the gigabyte mobo colours and cables. Whilst I like the red, it needs to match a little bit more to the components (noobo mistake..).
Last edited:
I didn't mention the water loop temperature sensors really, until now. Consists of three inline sensors connected to the main loop, like this aquaaero sensor.
The Bitspower sensor is part of the drainport assembly, as is a second xspc sensor.
The Bitspower sensor is part of the drainport assembly, as is a second xspc sensor.
Last edited:
The sensors are located just before and just after the CPU/GPU sequence. Temps remain about the same. The xspc sensor, display only, is in the base of the case to see temps when case is open. Was useful when I was setting things up and wasn't sure if pumps and aquaero were working properly.
Last edited:
Scariest part of building the watercooling loop was turning it on.
The hardest part was.... cable management! Kin ell this took me ages to sort and it's still messy... Tried to get it as flat as possible to get the case closure on.
The corsair black flat ribbon cables are actually quite difficult to get to lay flat as they're quite rigid.
The hardest part was.... cable management! Kin ell this took me ages to sort and it's still messy... Tried to get it as flat as possible to get the case closure on.
The corsair black flat ribbon cables are actually quite difficult to get to lay flat as they're quite rigid.
Last edited:
Preliminary testing: first Heaven benchmark and monitoring to test the system. This is after two runs at Heaven with max res 3840x2160
Some data logged from HWiNFO64 with the GenericLogViewer plug-in to pump out the graphs.
Heaven started about 30 seconds into the test, and ran twice, until about the 10 mins mark.
Some data logged from HWiNFO64 with the GenericLogViewer plug-in to pump out the graphs.
Heaven started about 30 seconds into the test, and ran twice, until about the 10 mins mark.
Associate
Hi nice list of parts but sort that loop out, you can make it much more effective and get more performance and use less parts which in the long term equals better reliability. less chances of leaks occurring.
Remember water takes the path of least resistance having two outlets means the flow in the upper card will very poor and all those 90 degree bends are killing your flow rate.
I do like that you have stuck a drain port tho in the loop (more people need to think this out) and your build is very pretty.
quick check of your flow rate and looking at your graphs and your fans are ramping up to max (i could be wrong) to cool the look but the temps are still climbing
have a read and a play
http://martinsliquidlab.i4memory.com/MartinsFlowRateEstimator.html
its a build to be proud of tho
Remember water takes the path of least resistance having two outlets means the flow in the upper card will very poor and all those 90 degree bends are killing your flow rate.
I do like that you have stuck a drain port tho in the loop (more people need to think this out) and your build is very pretty.
quick check of your flow rate and looking at your graphs and your fans are ramping up to max (i could be wrong) to cool the look but the temps are still climbing
have a read and a play
http://martinsliquidlab.i4memory.com/MartinsFlowRateEstimator.html
its a build to be proud of tho
Hi Karl thanks so much for the link... will def give that a good look at.
Re the cards, yes am a bit puzzled by the 2-3 degree differential between GPU1 and GPU2 temps... all the research I did on the net pointed to parallel = equal temps, and also that parallel reduced flow restriction compared to series. I thought that parallel plus dual D5s would offset the restriction additions of the 90 degree fittings and the QDCs. Not to mention use of 1/2inch ID tubing.
A few notes on the fans. Fan1 fans are the 5 EK Varder Furious F55-120 fans mounted to the front intake Rad1 (EX-360). These are minimum 1200rpm and maximum 3000rpm (ie the graphs below are scaled, not min/max). So on the test above these basically went less than 50% of their capacity.
The Fan2 fans are WingBoost 2s (Rad 2, 2x push), 500 min rpm and 1500 max rpm. I didn't log Rad 3 (also WB2, but 4xfans in push/pull). So neither seemed overly pushed.
However, what I have learned is that the 1200 min rpm of the EK Varders is too loud for me when considering a system that is "silent" when at rest - as the minimum this fan goes to is 1200, I can't get PWM control to reduce the Varder below min speed of 1200, so am considering ordering some replacements in the form of 5x Wingboost 2s (which can spin to 0rpm under aquasuite), subject to stability testing to see how far the Varder RPMs go under a full system stability test, to see if the WB2 maximum speed of 1500rpm is appropriate for a high load situation (ideally need a fan with 500rpm min, 2000rpm max!)
Speaking of stability tests, here is a 20 minute AIDA64EXtreme stability test for CPU / FPU / GPU / RAM.
Test started just after the log began, and ran for 20 mins. Temp1WC Inline is the water temp. CPU Package is the package temp, not the CPU temp. Max Package obtained was 79C peak, average 57.9C.
Here is the AIDA64 statistics page:
And a Firestrike rating: http://www.3dmark.com/fs/4691516
Will run some longer term stability testing, as you say temps are still climbing. However, I cannot yet match these temps in a gaming situation as of yet - ie stability testing is pushing the system far more than games have to date - I will post some results from gaming in due course.
Thank you
Ps- flow rate is not displayed on the graphs, it's a fixed setting at 2.5 on the d5 varios
Re the cards, yes am a bit puzzled by the 2-3 degree differential between GPU1 and GPU2 temps... all the research I did on the net pointed to parallel = equal temps, and also that parallel reduced flow restriction compared to series. I thought that parallel plus dual D5s would offset the restriction additions of the 90 degree fittings and the QDCs. Not to mention use of 1/2inch ID tubing.
A few notes on the fans. Fan1 fans are the 5 EK Varder Furious F55-120 fans mounted to the front intake Rad1 (EX-360). These are minimum 1200rpm and maximum 3000rpm (ie the graphs below are scaled, not min/max). So on the test above these basically went less than 50% of their capacity.
The Fan2 fans are WingBoost 2s (Rad 2, 2x push), 500 min rpm and 1500 max rpm. I didn't log Rad 3 (also WB2, but 4xfans in push/pull). So neither seemed overly pushed.
However, what I have learned is that the 1200 min rpm of the EK Varders is too loud for me when considering a system that is "silent" when at rest - as the minimum this fan goes to is 1200, I can't get PWM control to reduce the Varder below min speed of 1200, so am considering ordering some replacements in the form of 5x Wingboost 2s (which can spin to 0rpm under aquasuite), subject to stability testing to see how far the Varder RPMs go under a full system stability test, to see if the WB2 maximum speed of 1500rpm is appropriate for a high load situation (ideally need a fan with 500rpm min, 2000rpm max!)
Speaking of stability tests, here is a 20 minute AIDA64EXtreme stability test for CPU / FPU / GPU / RAM.
Test started just after the log began, and ran for 20 mins. Temp1WC Inline is the water temp. CPU Package is the package temp, not the CPU temp. Max Package obtained was 79C peak, average 57.9C.
Here is the AIDA64 statistics page:
And a Firestrike rating: http://www.3dmark.com/fs/4691516
Will run some longer term stability testing, as you say temps are still climbing. However, I cannot yet match these temps in a gaming situation as of yet - ie stability testing is pushing the system far more than games have to date - I will post some results from gaming in due course.
Thank you
Ps- flow rate is not displayed on the graphs, it's a fixed setting at 2.5 on the d5 varios
Last edited:
Associate
A very detailed and thorough build log, good work! Definitely a reference for when I get around to using water for the first time.