[Build Log] NCTEKM1 - NCASE M1 Dual Radiator Watercooling

tekm

n00b
Joined
Apr 14, 2014
Messages
17
I recently completed my NCASE M1 watercooling build with dual 240mm radiators. I was planning to do this build for a while but only got around to completing it now. I also incorporated several quick disconnects in this build to easily dismount the side radiator and allow for easier bleeding of the loop. This way I can bleed one radiator and then insert the second radiator to complete the process.

wBB6KZ6l.jpgs

  • Case: NCASE M1 Rev 2
  • Motherboard: ASUS Maximus Impact VII
  • CPU: Intel Core i7 4790k (Delidded)
  • CPU Waterblock/Pump Combo: Swiftech Apogee Drive II
  • GPU: PNY GeForce GTX 970
  • GPU Waterblock: EK-FC970 GTX Full Cover Water Block - Acetal + Nickel & EK-FC970 GTX Backplate - Black
  • PSU: Silverstone SF600-G 600W
  • Optical Drive: Bluray Slot-loading drive
  • SSD: Plextor M6e M.2 256 GB
  • HDD: 3 x 2 TB Samsung Spinpoint M9T
  • Radiators: XSPC EX240 240mm (side), Coolworld 240mm Radiator (bottom)
  • Reservoir: FrozenQ M1 NCase Reservoir
  • Tubes: 3/8'' ID, 5/8'' OD Primoflex Advanced LRT Crystal Clear
  • Coolant: Mayhems Pastel Blue Berry
  • Fittings: Barrows, Bitspower, and Koolance (QDCs)

jL4yUiJl.jpgs

KI3o40Il.jpgs

K5hxiyTl.jpgs

psIy0kBl.jpgs


35Dyyrtl.jpg

LyItKKkl.jpgs

p8zn97ol.jpgs




xIQs33Dl.jpgs

DGFO3Ppl.jpgs



LKeZWt6l.jpgs

rRA7trnl.jpgs

KQJDVY8l.jpgs



ZJW4negl.jpgs

MwWqBixl.jpgs

kWemGvUl.jpgs

qjljwDcl.jpgs

VxMHfVOl.jpgs

2dEIf3Wl.jpgs

 
Wow, loop looks really slick. Good job. How do the Silverstone fans do?
 
Would 25mm fans fit on the top radiator with the quick disconnects? I was planning to get rid of the res on the back by using quick disconnects and filling the loop with some res and then removing the res from the loop. I just wonder how much space would there be between rad + standard fans and the psu?
 
I recently completed my NCASE M1 watercooling build with dual 240mm radiators. I was planning to do this build for a while but only got around to completing it now. I also incorporated several quick disconnects in this build to easily dismount the side radiator and allow for easier bleeding of the loop. This way I can bleed one radiator and then insert the second radiator to complete the process.

That is beautiful.
Also curious about the how those fans perform, as well as temps/noise/etc.
 
What an amazing build! Gorgeous colours, high practicality and clean tubing. I am in awe!
 
Wow, loop looks really slick. Good job. How do the Silverstone fans do?

They obviously won't match the static pressure of some of the more premium full size fans, but I think they are adequate. I did run one quick test with IntelBurnTest under a very high stress level at 4.4 ghz, and after 10 runs I had a max temperature of ~76 deg C. The CPU voltage is at ~1.33V so hopefully after optimizing the voltages I can get the temperatures lower or the overclock higher.

Would 25mm fans fit on the top radiator with the quick disconnects? I was planning to get rid of the res on the back by using quick disconnects and filling the loop with some res and then removing the res from the loop. I just wonder how much space would there be between rad + standard fans and the psu?

Upon closer examination I do not think 25mm fans would fit in with the quick disconnects. I tried wedging in another 12.5mm fan in the remaining available space to observe the remaining clearance and it would not fit in for me. Unfortunately, the clearance seems to be a couple of millimeters too narrow.
 
Upon closer examination I do not think 25mm fans would fit in with the quick disconnects. I tried wedging in another 12.5mm fan in the remaining available space to observe the remaining clearance and it would not fit in for me. Unfortunately, the clearance seems to be a couple of millimeters too narrow.

Wouldn't that depend on fitting choices? If you used a longer extension fitting between the rad and the 90 deg fittings it looks like there could be room. Or am i missing something?
 
A nice, clean build for a dual radiator setup. How are the quick disconnects with spilling ?
I can't but wonder if it was worth the trouble and money over a single radiator build with 25mm fans though.
 
Wouldn't that depend on fitting choices? If you used a longer extension fitting between the rad and the 90 deg fittings it looks like there could be room. Or am i missing something?

Yes, I think it could be possible to place the quick disconnects beyond the PSU and potentially fit in a regular size fan. The reason I placed the qdcs at the radiator inlets is so that I can easily connect the tubes together when I remove the side radiator and I just want to form a closed loop to bleed only the bottom radiator.

A nice, clean build for a dual radiator setup. How are the quick disconnects with spilling ?
I can't but wonder if it was worth the trouble and money over a single radiator build with 25mm fans though.

So far I am very satisfied with the Koolance quick disconnects and I think they work great. When I disconnect them I maybe have a drop or two of coolant that escapes but I wouldn’t call it spilling as it doesn't spurt out or anything. Just have a paper towel handy to dab the ends of the disconnects when you’re taking them apart and you should be good to go.

And yes, it might not be the most practical or cost-effective solution for cooling, but that wasn’t my intent. I saw this as a fun project to see how I could maximize the capabilities of the case with respect to water-cooling despite possible diminishing returns on cooling efficiency when adding a second radiator in an area with non-optimal air flow.
 
How is the build quality of the FrozenQ reservoir?

I know they had issues when they first came out and I'm curious if they've resolved them since.
 
How is the build quality of the FrozenQ reservoir?

I know they had issues when they first came out and I'm curious if they've resolved them since.

I bought mine half a year ago so I don't know how closely it reflects current batches, but I haven't noticed any issues with it.
 
I finally got around to measuring temps. Running the GeForce GTX 970 at stock speeds, I first ran Furmark and recorded a maximum GPU temperature of 36 deg C. For comparison, Anandtech measured 63 deg C when performing the same test using stock air cooling, so I think I achieved a decent drop.

KxnjVxls.png


I also finally optimized my overclock this time and have the system running at 4.7 Ghz / 1.25 V. I ran IntelBurnTest under very high stress levels again several times and recorded a max temperature of 78 deg C. This is a bit higher than I would have liked, but I think it will suffice.

XoakDZSs.png
 
i dont have any actual experience with watercooling but those temps seem reasonable given the numbers here:
http://www.overclock.net/t/1505481/summer-water-block-round-up-2014#post_22655918

friggin haswell...

though you do have it delidded. do you know what temperature the water was at when you measured 78C for the cpu?

Yeah, I forgot how bad the thermals can get with haswell. I am used to the lower temps that I get my other haswell build though that has a 540mm radiator.

I have no way of measuring the temperature of the water as I do not have a temp/flow meter. The CPU temperature reported here is after I already ran intelburntest three consecutive times (after ~30 minutes under load).
 
What's in between your ihs and core?

Tweaktown did a review of the same waterblock/pump combo and got a max 64c on their i7 2600k oc'd to 4.5ghz so I guess your temps aren't that bad. http://www.tweaktown.com/reviews/498...ew/index7.html

I'm using Coollaboratory Liquid Pro between the ihs and core, and Arctic MX-4 between the ihs and Swiftech Apogee Drive II.

I raised the maximum fan speed, and now get a slight drop in CPU temperature, with a maximum temperature range of 69-76 deg C between the cores. I'm tempted to swap out one of the side radiator fans with a Noctua fan to see how much the increased static pressure would help.
 
I'm tempted to swap out one of the side radiator fans with a Noctua fan to see how much the increased static pressure would help.

Slim fans with same number of blades tend to have a similiar P-Q curve, since blade attack angle will be similar. Static pressure is measured at 0 airflow is not always a useful indicator for the pressure needed to get through radiator fins. For the same reason max air flow isn't realistic for radiator use either. To make an analogy with motor vehicles: airflow is like top speed, downhill; static pressure is like a tractor pull; what you actually need to know for radiators is the torque per second over the whole range of rpm.

Just trying to warn that fans with impressive specs on the box may disappoint since the measurements quoted are not directly relevant to radiator performance.

Which exact model fans do you have now and which other fans are you intending to purchase? I can see if I can dig up real world P-Q data for you.
 
Slim fans with same number of blades tend to have a similiar P-Q curve, since blade attack angle will be similar. Static pressure is measured at 0 airflow is not always a useful indicator for the pressure needed to get through radiator fins. For the same reason max air flow isn't realistic for radiator use either. To make an analogy with motor vehicles: airflow is like top speed, downhill; static pressure is like a tractor pull; what you actually need to know for radiators is the torque per second over the whole range of rpm.

Just trying to warn that fans with impressive specs on the box may disappoint since the measurements quoted are not directly relevant to radiator performance.

Which exact model fans do you have now and which other fans are you intending to purchase? I can see if I can dig up real world P-Q data for you.

You are right that fan performance is a function of several variables over different operating ranges. I didn't mean to oversimply the performance of the fan to a single metric disclosed in marketing material. However knowing that the NF-F12 have generally have worked well for me in the past when cooling my radiators, and because I already have unused spares, I figured that I might as well try it in this rig since it only takes a minute or two to swap in.
I replaced the second fan in the side radiator (the only place that it would currently fit) and as before, I reran IntelBurnTest three consecutive times under very high stress levels. The maximum temperature range between the respective cores is now 67-75 deg C, so there seems to be a small improvement by just swapping 1 of the 4 fans with the Noctua one.
 
NF-F12 have generally have worked well for me in the past when cooling my radiators, and because I already have unused spares, I figured that I might as well try it in this rig since it only takes a minute or two to swap in.

Ah sorry, thought you were talking about a slim Noctua, but as I see they don't make any in 120mm size.
 
Back
Top