1. OCAU Merchandise is available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion in this thread.
    Dismiss Notice

O11 Air Mini

Discussion in 'Extreme and Water Cooling' started by woja, Jan 1, 2022.

  1. Spyder6052

    Spyder6052 Member

    Joined:
    Nov 15, 2006
    Messages:
    1,264
    Location:
    Woodroffe NT
    pm your address and i will send it off to you (although dont know how long it will take as AusPost are as useless as tits on a bull)
     
    juzz86 and woja like this.
  2. OP
    OP
    woja

    woja Member

    Joined:
    Feb 18, 2004
    Messages:
    680
    Location:
    Sydney
    You're an absolute champion!
    Thanks so much!
     
  3. OP
    OP
    woja

    woja Member

    Joined:
    Feb 18, 2004
    Messages:
    680
    Location:
    Sydney
    Spyder6052 I tried to mount my D5 into the FLT120 but the O-ring that I had from my other pump tops was a few mm too big to fit into the FLT slot, I've sent through a support ticket to see how I can get one that will fit, the O rings on the website are 60x3mm where as the FLT 120 is 52x3mm
    I did manage to mount it and run it but the o ring was squished out to the side and out of the groove which made the pump slightly off kilter and made a bit of a buzzing sound and yeah a bit too janky for me so I ended up putting in the DDC for now until I find an appropriate size o ring for it.

    Did some super rough testing today with regards of flow rate and temps and coolant temps.
    I've got 2 water temperature probes in my system at the moment (not in the most ideal position for this).
    Loop order is FLT 120 (Water Temp Internal) -> Bottom Rad -> GPU -> CPU -> GPU (Parallel) ->External Rad -> External Res/Pump -> Water Temp External -> Top Rad -> FLT120 (Water Temp Internal)
    Ideally I'd have the water temp sensor after the Bottom Radiator before the GPU to measure the coolest point and one after the GPU outlet before it hits the external radiator to measure the hottest point.

    With the testing I should have made the fan speeds static rather than on the fan curve that I currently run and I only waited 5 minutes or so for the temps to stabilise (which is probably too soon).
    I ran Unigine Heaven on a loop as well as Prime95 Small FFT non AVX as the load.

    65 litres/h - highest temps components wise and biggest difference in water temps which makes sense given water has more time to pick up heat from the component and lose heat in the radiators.
    I think the additional temp in the components is that due to the parallel loop GPU/CPU and the splitting of flow rate it's not playing nice with the microfins as well the CPU outputting heat back into the GPU water block but that's with the parallel loop in general regardless of flow rate
    65lph.png

    This picture is a bit of a combination of 101 litres/hour and 190 litres/hour
    at 100 l/h CPU and GPU temps were 2-3 degrees less than 65 l/h. Water temp delta between sensors was less as expected with slightly higher flow rate
    At 190l/h CPU and GPU temps are 1.5 -2 degrees lower again with an even smaller water temp delta
    100lph 180lph.png

    Finally at full bore 240l/h temps again drop across the board and the water temps are lower in general.
    240lph.png

    Final thoughts though, confounding factors are that fan speeds are tied with the Internal water sensor which is the cooler sensor since it's had most of the heat out from the radiators. Next time I test I need to set static fan speeds. I should also have more granularity in flow rates so I can plot some data points.
    Ideally I should move the sensor that controls the fans to after the GPU before it gets to a radiator, right now it's the one that has the big external radiator and the top radiator bleed out the heat. I've currently changed it to the one that's just after the external radiator for now but I'll move it when I can next be bothered.
     
    juzz86 and Spyder6052 like this.
  4. OP
    OP
    woja

    woja Member

    Joined:
    Feb 18, 2004
    Messages:
    680
    Location:
    Sydney
    Ooo, jsut found out that the O ring on the spare multitop that I have is is 52x 3 so I can use that in a pinch, the new o rings are made of EPDM where as this one is nitrile (NBR 50) but I think it should be fine since we're not dealing with excessive heat or UV or stuff like that. I'll try again when I don't have the kids around haha
     
  5. OP
    OP
    woja

    woja Member

    Joined:
    Feb 18, 2004
    Messages:
    680
    Location:
    Sydney
    Used the O ring from the X3 multitop that I had and never used,
    20220228_082923.jpg

    Ended up with a slightly lower flow rate than with a DDC and D5. Not sure if my loop is super restrictive with the quick disconnects.

    I taped a a spare 10K sensor to the outside of a fitting before it went into the external radiator and it says it's 3 degrees higher than the water temps which I highly doubt.
     
  6. Will S

    Will S Member

    Joined:
    Aug 15, 2003
    Messages:
    6,264
    Location:
    Coniston (Wollongong)
    not sure about the flt120 but flt240 came with both D5 and whatever the other pump is O rings
     
  7. OP
    OP
    woja

    woja Member

    Joined:
    Feb 18, 2004
    Messages:
    680
    Location:
    Sydney
    I had bought the flt120 with the DDC attached so it didn't come with the d5 mount and o ring
     
  8. Will S

    Will S Member

    Joined:
    Aug 15, 2003
    Messages:
    6,264
    Location:
    Coniston (Wollongong)
    ok, now I understand...thats a pain in the butt
     
  9. OP
    OP
    woja

    woja Member

    Joined:
    Feb 18, 2004
    Messages:
    680
    Location:
    Sydney
    Decided to get a more accurate hottest water temp, changed a 3 way cube to a 4 way (accidentally ordered a silver one lol) and added another barrow water temp fitting into the GPU output before it gets to a radiator.

    20220303_153402~2.jpg

    This time I set static fans and just varied the pumps to get desired flowrate. Again heaven looping and Prime95 small fft non AVX

    Surprised about the difference in water temps between the hottest and the coolest sensor (still not coolest point as its before the 3rd radiator) everyone states that's its always very similar temps throughout the loop. Looked like 2-3 degree difference.

    Screenshot_20220303-191341_Photos.jpg

    Screenshot_20220303-191115_Sheets.jpg
    Ran out to time before I could do the lower flow rates, need the lower data points. I'm thinking maybe I should use a static gpu stress test like furmark.
     
  10. The Beast

    The Beast Member

    Joined:
    Jun 27, 2001
    Messages:
    5,146
    Location:
    Gaufres. Bière. Chocolat. Moules Frites.
    I'd believe 2-3 degrees, there has to be some change across the loop.

    What people mean by not worrying about component positioning in the loop is that the effect (negative or positive) on cooling performance for a component is going to be next to zero.
     
    woja likes this.

Share This Page

Advertisement: