New Computer Part 2 – The Cooling

Picking up from where I left off with part one of the new PC build we move right into it with the tubes! Tubes everywhere! Well ok………. Maybe not everywhere, just here and there. So as I mentioned I went with a motherboard that had a water block over the VRM modules with the only reason being to add some addition visual flair to the build since I am rocking a glass case.

So what I did in order to prepare for the water cooling loop was take some string and roughly mock up the layout of the tubes to make sure I had a minimum count of fittings and angle connectors. This also gave me a roughly estimate of the amount of tubing I would need. So with this information in hand I went to 3 websites and a local brick and mortar store to get everything I need.

Microcenter:

  • -1x Bitspower male to male black connector
  • -1x Bitspower black ball valve
  • -2x Bitspower black plugs

FrozenCPU:

  • -1x Bitspower 90˚ female to female matte black fitting adapter
  • -12x EKWB Black 12mm rigid tube compression fittings
  • -1x EKWB 100mL replacement reservoir tube
  • -1x Mayhems orange dye

EKWB:

  • -4x 2x500mm PETG Rigid tubing or 4m of rigid tubing
  • -1x EKWB fancy segmented 90˚ fitting that swivels as 2 45˚ segments in black
  • -6x EKWB 90˚ rotating fittings in black
  • -1x Set of silicone inserts for bending

Amazon:

  • -1x Monsoon Hardline Mandrel Kit for ½” or 13mm tubing.

With everything in hand from my orders as well as the tools at home I was ready to go since the core of the PC was done.

I began the process of the water loop by watching a lot of YouTube videos about how to layout cooling loops as well bending rigid tubes. Having watched a lot of videos about the best way to bend tubes I decided to do at least a single practice bend before cutting and measuring for the runs in the case. The first (and only) practice bend I made ended up slightly wrinkled in the radius but I realized that I may have not properly heated the tube and bent it too quickly. So what I did was adjust the heat level of the heat gun down a couple of points and then continuously rotated the tube under heat even as the bend was being formed. What this did is prevent uneven material movement, like flat spots and wrinkling, from occurring when I pulled the tube over the mandrels. This practice resulted in near perfect looking bends every time (though of course I still had to scrap a fair bit of tubing in the end).

Now the first bend I went for was to be the simplest and shortest of them but ended up being the most difficult. This was the bend from the CPU block outlet to the VRM block inlet. I had wanted to do this as a 90˚ fitting off of the CPU and then a 90˚ bend into the VRM block. But after 3hrs of work and wasting something close to a meter of tubing I realized that the radius was going to be too tight and create a possible weak seal over time due to increase pressure on the o-rings. So I went the route of the more complex 180˚ tube bend from CPU to VRM block. This ended up being the best solution both visually and from a time sink standpoint as I got it done on the first try for the radius, just not the height, but since I made the tube long it was cut to fit.

20170605_202038

From there I made the run from the VRM outlet to the radiator inlet which given the height difference for the openings resulting in some more interesting bends as I wanted to add some visual flair to the top of the case as I felt that a straight tube would be a little bland. I then moved on to the radiator outlet to the reservoir which was the second most annoying run of the build since I decided to run the tube to the top of the reservoir. The position of the openings was such that I had to make small bends in the X/Y/Z directions rather than just the X/Y directions. It took a couple of tries but I managed something halfway decent for the tube. After all of these I did the easiest of all runs. The CPU inlet to the pump outlet. This could have been as easy as a couple of straight runs with a 90˚ female to female fitting since the tubing was not long enough to do this run with a single piece. But because I wanted to add some additional flair to the build and to make the RAM more visible, I brought the tube down at an angle before running it over to the fitting and into the pump. I also placed the ball valve on the “inlet” side of the pump/res as a system drain.

With all said and down I made sure the system was empty of all but the motherboard/CPU/old GPU and the cooling loop and then laid down some papertowel before the fill. This was done as I wanted to minimize the risk of water on the good parts given this was my first time water cooling at all. I then mixed up the coolant concentrate that was in the kit with some distilled water, added some orange dye, and began filling the system. And after some nerve wrecking 10 minutes the loop was filled and running without leaks. At this point I let the loop run overnight and the following morning no leaks were present on the papertowel.

Now it was at this point that I had the brilliant idea to try and change the front case fans as I got new ones that had higher static pressure and so I emptied part of the loop only to find I could not fit them without a massive teardown. So I went back through and put the tubes I pulled out back in and realized a design flaw with my loop. Since I changed to a multiport top on the reservoir due to interference issues with the GPU cooler there is no leeway with changes to the top location… Basically the issue is that since there are 2 ports on top you have to be precise when removing and replacing the top otherwise the tube is useless and can create uneven tension on the tube. Thankfully I was able to work things back into place without issue and refilled the loop and figuring out how to remove the air foam that was present. I did not really do the tipping method that most use but I did monitor the system for about 3-4hrs and occasionally add more liquid and this worked well

A few days later I was able to install the hard drives I picked up as well as the GPU from the computer I was using at the time. Windows was installed and everything looked to be running smoothly although noise did become an issue as the radiator has a high fin density that can cause such an issue. Although then the slight noise issue the computer runs quick and I did not have to do anything special in the BIOS to install and boot from the NVMe drive.

Temps at idle where in the upper 30’s to low 40’s ˚C

Under AIA64 loads the CPU jumped to low/mid 70’s ˚C

I think these temps are due to the fact that Intel uses thermal paste between the processor die and heat spreader as well as the restrictive airflow on the case intake due to its design.

Overall I am pleased with how the build turned out, but there is still more to be done which I will cover in another short update.

20170607_210122

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s