[WORKLOG] Meshify 3 XL Build

Hi All,

I finally got itchy feet and bought a new case to see how much difference it would make to my coolant temps. Original thread here: Running hot after rebuild.

New case is a Fractal Meshify 3 XL. I laboured over the decision between this or a Define 7 XL, because I have really loved my Define R6, and wanted a bigger case for a bit more room to move. Size difference is pretty apparent straight away:


My assumption was that this case should definitely improve the general airflow because of the 3x140mm fans on the front, and a single 140 at the back as exhaust. Both radiators as exhaust. Build was fairly simple and they are a nice case to build in. Quite well thought out. I am a bit bummed at no dust filter on the front, and no option for vertical GPU mounting, but oh well. I had a friend 3D print me an adjustable GPU support strut, which looks nice and functions well (sorry I forgot to take a photo of that one).

Mid build shot - showing the side mounted rad (X-Flow does fit!), and the clear airflow through the case from the front.


Finished build:


Funny, the clearance between the fans on the top and side radiators is zero. The rubber fan mounting thingies are touching and needed to squish a little (not tight) to get the top radiator bracket to screw down to the chassis. I also needed to change the tubing to use the opposite corners on the top rad to make it work.



Temps...

Well, I am not blown away by the results, but I perhaps had unrealistic expectations. I will provide some graphs tonight, from my early testing, but coolant temps top out at around 41-45c (depending on the ambient temps), GPU 55c top, CPU around 80c. Noise is the other thing; this is no longer a 'quiet' build. Its not bad really, but under peak load (BF6), it does make a bit of a hum. Kinda like one of those box fans on low speed.

Overall, I am happy with the build, and it was a fun exercise to transfer the old equipment to the new case. Just don't know if this is the what I should expect for this loop, and the noise levels. I will run a test with 100% fans to test the max cooling capacity (haven't done this yet because that's not going to be how its used in practice).

-Scott

 

having used a fractal R6 for years, its really good at noise deadening (solid metal and sound deadening) where as the meshify is a open mesh case, you also might be hearing the pump resonance as well.

do some noise testing and isolate if its pump or fan noise (fan control is a good free program to control fan and pump speeds)

temps are always relative to ambient while your water temps are a little high its not the end of the world, again check pump flow speed.

cpu and gpu temps of what your getting kinda make sense in line with your water temp.

Also you could tidy tubing and loop order up for a cleaner look for very little cooling difference. Pump->GPU->CPU->TopRad->SideRad->Pump

Flow rate overcomes benefit of loop order

Build looks good, great choice in tubing

 

Component temps look pretty good and your coolant temps are consistent with mine as I can see 43c on a hot day (2 x 360mm rads). Have also played around with fans and curves to get something satisfactory, but under prolonged gaming there is not much you can do as the heat soak will be the limiting factor. I have my fans at 100% for coolant temps above about ~38ish (can check when i get home). Whilst this is not quiet at all, it's fine with my headphones on.
Maybe the X-flow rads are less effective than traditional rads?
I don't recall you mentioning which model of Noctua fans you are using, so checking the air pressure ratings might be worth doing? I opted for TT Toughfans EX12 Pro for noise and performance optimisation, and value.

 

Pump flow seems to be nice and quick. The pump drains the res in a few seconds during the filling, and the loop is reasonably low resistance. I don't have a flow meter to be scientific about it, but I am happy with flow rate based on observing the bubbles passing through after filling. I am running the pump at 100% - no longer on a curve.

I deliberately deciced not to do this order so that I would give the water a chance to cool down some between each heat source, rather than dumping partially heated water into the GPU. Agree, it would look nice though. Maybe one day when I get bored I will test this scientifically.

I am running 3x Noctua A12x25's pushing on each rad, and an A14 on the chassis rear in exhaust.

For info, fan curves are currently mobo controlled, as Fan Control was not reading correct temps from the coolant temp sensor - NFI why not ; all my other monitoring software was consistent with the temps shown in BIOS.

I have the fan curve settings as follows (from memory - so mid point might be different):
Rad fans based on Coolant: 20% at 30C, 60% at 50C, 100% at 60c. Seems to keep it at a sensible noise.
Front/rear chassis fans based on MB/Chipset temps. Similar ramping curve.

 

Yeah you may be just saturating your 2x360's when i went from 3x280's to 4x280's i saw 1 deg improvement but heat soak took 30-45min longer

I have done testing where i put more rads between components vs gpu straight to cpu and it was 2deg best case.

Cpu's are far more erratic in temps in games as they don't do sustained loads like a GPU does so found its less detrimental to a cpu to be fed from a gpu even though that's the hottest component (usually) again this is only for neatening up tube runs lots of tube isn't an issue

When you flow rates are 150l/h + its moving so fast its moving all the heat it can i can see temp changes when at 100l/h or less (visibly slow) like you said 2rad 2block 1 block very low resistance

As mentioned I would grab Fancontrol (Free software) and i think you can bring the water temp in as a input then you can make fan speed chase delta or Ambient vs Water temp

This is how i control in aquasuite very simple but tied to aquacomputer hardware

 

This kinda explains your coolant temps as fans are not going above 60%. Means you have some head room for more cooling performance, but would be at the cost of noise. But at 60% on the noctua's, i'd be very surprised if they are making an audible noise? Would most likely be the pump motor that you're hearing, so maybe dropping that down to something tolerable and checking that performance is still ok?

 

Ah, that was the point I meant to answer earlier. The pump is essentially inaudible. I cant hear it running unless I put my ear up against it. The pump is the star performer of the bunch.

 

Pumps are a super mixed bag lucky you got a quite one i went through 5 before i got silent ones.

apart from mounting pressure, thermal paste, and lapping blocks not much else you can do for temps apart from increase fan speed or cooler ambient temps or a water chiller.

I never really go above 1000rpm of 60% on my Artic P14's cooling is better by a few deg at 1800rpm but i don't wanna listen to that.

at the end of the day even if you run at the temps you have its still more silent than air cooling.

 

Graph as promised.



I will note that the boost clock on the GPU is sitting mostly at 1965MHz. Specs from Asus: Base Clock: 1440 MHz, Boost Clock: 1905 MHz, OC Mode - 1935 MHz. So that's nice...