I have placed the probe 2cm out from the middle of radiator (Mo-Ra 3 3x3x120mm) which is covering most of the front of the PC. I'm guessing that's the best place as it is the actual temperature that is used by the radiator.

Anyone that has a different opinion?

I control the 2 pumps and the radiator fans from the difference of the air temperature in the room and the water temperature out of the radiator. This difference seems to be stable around 2.5C when the PC isn't doing anything. On hard use the difference can rise to as much as 8C or even more, but this only happends during stress tests with Aida64. In these cases the radiator will easily cool the water 6C or more, but the water and room temperature will gradually rise. Not sure how high it eventually will go, but I have a window I can open and an air conditioner I can turn on, in worst case. I have the PC placed under a large desk I made for the 3 monitors and a place for all the stuff I have on the desk, but that makes it difficult for the air around the PC to escape, so I placed 3x120mm fans under it, to pull cool air down there (actually the noiciest thing in the room, even if you hardly hear them heheh).

CPU Intel core 7 5960x (4x4.4GHz+4x4.5GHz)
GPU ROG Strix 1080GTX OC (2151MHz/2752MHz)
RAM Dominator Platinum 3200 (3200MHz)

This is stable OC.
The CPU can be OC'ed up to 4,7, but it will eventually crash; probably have to some serious work on the OC'ing if I want it that high.
The GPU will crash a lot in Unigine Valley if I go above 2151MHz, but I've been able to run it on 2164MHz. The GPU Memory clock seems to be easier to OC, but the effect doesn't seem to be high.
The RAM can not be OC'ed further unless I do some serious work in BIOS, I guess.

I did some of the same inside the cabinet. Put a 9mm heatshrink around the sensor and shrunk the lower part so it fit into a hole. Seems to work well as it's hardly disturb by the airflow. Will try the same here. Thanks.

I'll have to recalibrate all termometers over night. Seems things have changed since I moved the PC from the work bench, downstairs, and up here to the "office". Even with one side panel open and a blower blowing air from the same place as the intake for the radiator doesn't drop the temp diff below ~2C. I'll turn off the PC, close the door and leave it for 12 hours or so, then check with the Aquaero how the temps look.

A picture link I think

Until I get something better. A hose filled with foam and the sensor in the middle.

Ended up in a whole lot of ¤#%/&()(.

The tubes I used aren't good enough, they conduct heat and can't be placed in front of the radiator. Will have to find some stuff for that.

Turned out I have way too much power for air going into the cabinet. Was trying to test it out and managed to get negative delta between cabinet and room (ie, it was warmer outside the cabinet than inside). After some head scratching I found tht there was so much positive pressure inside the cabinet that it was forced out through the radiator, against the air flow made by the fans there. I knew I had gone over the top with all fans being push/pull, but that 4+4 fans in the rear are able to overcome 9+9 fans in the radiator and 2+2 fans pulling air out over the motherboard was more than I bargained for. Guess I have to take into accout that the filter, although a good one, has to much friction (or whatever you call it when it comes to filters). So, now it's all back to testing, trying to figure out the optimum way of cooling this monster. It was working, btw, with the temp probe in front of the radiator, but the flow out through the radiator destroyed it. Might just have to find the limit and set a max rpm.

Yeah, but I got a total of 15 120 holes and 4 80 holes on the top. As it is now I use the 2 outs behind the MB as outtake, the rest as in and let the over pressure get rid of the rest through the top 4 holes. This is obviously not enough (or too much, depends on hwo you see it), so I'll turn the in-fans at the back to blow air out. Will see what that does.

Changed direction on the 4+4 fans on the right side back and it did indeed help. It actually looks like I can't get negative pressure even with all the fans in the back at full speed. I've ordered 4 80mm fans to put on the top to see if that can help getting the air out.

Fans out, back is 12 fans in a pull/push configuration.
Fans in through the radiator, front is 18 fans in a pull/push configuration.
The radiator is a Mo-Ra 3 Pro (yeah, I mounted it in the front of the cabinet, got MountainMods to make a special front for me hehe).

With the front fans at 400rpm and the back fans at 1000rpm the water temp is about 1.4C above Room temp, and the cabinet temp is about 3.8C above the room temp.
With the front fans at 400rpm and the back fans at max (almost 2000rpm) the water temp is about 0.1C above room temp, and the cabinet temp is about 1.6C above the room temp, and even with these settings there's a healthy positive pressure inside).

The cooling of the radiator is about -2.8C. The only thing heating the water seems to be the Aquaero where the water rises about 2.2C (heavy use of 2 of the amps, but even so, they are hardly above 29C, half a degree higher than the 2 amps used for the radiator fans). The CPU is 0.2C and the GPU 0.4C. I removed the water blocks on the X99 and Mosfets, the MB looks better without them heheh.

The sensors was calibrated by leaving the cabinet open (sides and top removed) and letting it stand in a closed room for about 12 hours. I then used a normal termometer as basis, since the real temperature isn't as significant as the difference in the tempratures.

The room temp is meassure about 15cm below the desk and 15cm in front of the top part of the Cabinet, a little over where the radiator starts. The probe is placed in foam (to keep it from touching the sides) inside a small tube previously used for a refrigerator termomether. This should keep the sensor from getting disturbed by wind and also from fluctuating too fast.

When I get the new fans and have everything set up to keep the delta temps at 1.5C for both cabinet and water I'll make a video of the Aquasuite with all the meassurments while doing a stress test.

I ordered 4 SilenX, iXtrema Pro, IXP-52-14, 80x15. Have used them before and they are ok. What makes noise in the room is the 3 120mm fans mounted under the side of the desk, to keep cold air flowing under it (the power supply I use for them can be set to several different voltages between 3v and 12v, I'll reduce it from 12v later, when I know that the PC cooling is as I want it).

Oh, and the damned pumps (2 Aquastreams, 1 Ultima and 1 XT). I got my brother in law to make an acrylic plate wich I mounted on those yellow rubber screws, and I even put a 2,5cm stub of rubber hose after the outlet of the pumps (I use plastic tubes and they carry vibration), but even so, the vibration from the pumps, even at 3000rpm, is notcable, not bad, even at max they aren't too bad, but it's definately the noisiest thing inside the caboinet, on reasonable normal use (turning the fans up to max will make more noise, but mostly from the wind they generate).

The power cable from the XT is touching the side window as well, not enough room for the old Molex plugg. I'll by another Ultima later. The Ultimas free up 1 curve controller each, in the Aquasuite, which is sorely needed.

When it comes to flow, is it better to keep both pumps at the same rpm or is it ok to, as I do, let the first pump start and then let the second kick in later? I have set the first to run from 1,5 to 4,5C and the second from 2,5 to 5,5C, that way there won't be any "vacum", and the second one only kicks in after heavy or long use.