so 6mm extender coming up from the rad, 12mm from the distro?Old post...
...refers to the first prototype:
...the combo should be:
This:
This:
EK-AF Angled 2×45° G1/4 Black
EK-AF (Adapter Fitting) Revolvable angled adaptor (2×45°) with G1/4www.ekwb.com
an this:
EK-AF Extender 12mm M-M G1/4 - Black
EK-AF Extender 12mm M-M is a simple male-threaded extender interconnect for connecting two water cooling components with female G1/4 threads together.
EK-AF Extender 6mm M-M G1/4 - Black
EK-AF Extender M-M is a simple male-threaded extender interconnect for connecting two water cooling components with female G1/4 threads together.
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
Stalled DICE CASES - FLOW : expandable SFF case for watercooling enthusiast (built in distroplate)
- Thread starter diceboii13
- Start date
"should be so"...
...to be honest made that 1 year ago...
The combo look like this,
remeber that this is attached to 240 rad...
btw a bunch of extender and angled 2x45 always solve everything !
Take this opportunity to thank @diceboii13 ,
best package ever received!
(and the message inside is gorgeous)
...to be honest made that 1 year ago...
The combo look like this,
remeber that this is attached to 240 rad...
btw a bunch of extender and angled 2x45 always solve everything !
Take this opportunity to thank @diceboii13 ,
best package ever received!
(and the message inside is gorgeous)
Got the Aquacomputer Quadro so I am able to measure the water temps now. Here are some measurements after 10 minutes of Cinebench
Tdie remained steady around 68 throughout but I believe this is the threshold where the 5900x stops boosting so not sure if this behavior is good or bad. Water temps increased by 5 degrees over 10 min. Is that reasonable or should I revisit my CPU block installation?
Tdie remained steady around 68 throughout but I believe this is the threshold where the 5900x stops boosting so not sure if this behavior is good or bad. Water temps increased by 5 degrees over 10 min. Is that reasonable or should I revisit my CPU block installation?
The rad is not in a fixed position. If can be moved forward and aft. Almost like a rail, but a slot cut into the bottom/top frame. In other words, you could just move the rad to the fittings. Multiple combinations should work.
Finally leak testing overnight. 
+ my ghetto setup for powering the pump without messing with the PSU. It’s 12v Ac/Dc from an old fan to a buck converter hooked to the pumps power wires. I drop it to like 8v and let the pump run overnight.
+ my ghetto setup for powering the pump without messing with the PSU. It’s 12v Ac/Dc from an old fan to a buck converter hooked to the pumps power wires. I drop it to like 8v and let the pump run overnight.
Here It’s 3 dual ball bearing 40x28mm 18.5k rpm Delta fans attached to an Alphacool 1u triple 40mm radiator wrapped in sound dampening foam. Its crammed in between the GPU, exhausting out of the 3 slot GPU mount on the rear. I took the fans apart, lubed the ball bearings & spring so the fans are quieter. They don’t spin until the system is hot, and when they do the fans only spin to 15%. They only turn on when I’m gaming (with headphones) so I don’t hear it.Whoah, that's a pretty intense setup! What is that array of tiny fans behind the GPU for? You seem to have replaced the PSU fan with a noctua, but those things don't look quiet
That's some serious patience, fair play. I would break everything if I attempted something like that. 
BTW, how come your flow meter is not spinning in the video? I am pondering about getting a flow sensor because it's hard for me to evaluate how to translate the spinning of the wheel into a flow rate When I set the pump to about 1000rpm it spins very slowly, I can see the fins rotating clearly. Not sure if that's sufficient for good thermal capture.
BTW, how come your flow meter is not spinning in the video? I am pondering about getting a flow sensor because it's hard for me to evaluate how to translate the spinning of the wheel into a flow rate
When I set the pump to about 1000rpm it spins very slowly, I can see the fins rotating clearly. Not sure if that's sufficient for good thermal capture.Pretty sure my distro plate screws are too tight or the wheel is backwards. Can’t tell, but I’m leaving it until I feel like taking the front cover off...That's some serious patience, fair play. I would break everything if I attempted something like that.
BTW, how come your flow meter is not spinning in the video? I am pondering about getting a flow sensor because it's hard for me to evaluate how to translate the spinning of the wheel into a flow rate
I have a flow meter at the very end of my loop, before the distro input port for my flow rate. It lets you control the system much more precisely with a digital flow meter + aquaero. Add a temp sensor (I run 2) and you can do some really cool stuff with Aquaero software like calculate wattage output/dissipation in real time.
I'm a sucker for these types of metrics so I think I might pull the plug on a flow sensor. I'm already controlling the fan speeds with the water temps (with aquasuite) so this seems like the logical progression!
Yep, water temp really is the only sensible metric to control your fan speeds by.
Flow rate isn't quite as essential imo, I generally just setup my pump speed by letting my loop reach full load equilibrium. Then drop the the pump speed till the temps start going up, then increase it a bit again to have some margin.
But its nice to know for maintenance, if your flow rates start dropping significantly, its probably a good indication your loop is clogging up or the pump is failing.
I've managed to get it running. Took a while due to my motherboard RMA. Cabeling is well.. in progress. FYI it is possible to fit higher GPU than specified. At least as long, as you are willing to drill your brand new case. But taking the GPU supply into consideration, I really had no choice.
Finally got it running. Had a bad PSU which was quite unpleasant. Hit some quick Cinebench runs just for fun. 77c after 16 runs. VRMs don’t go over 53c and chipset is mid 40s. Nothing is tuned yet. Good luck to everyone still building.
I think my distro plate is too tight or the wheel is on backwards. Either way, it’s not getting fixed until I feel like taking the front cover off again which will be never. Doesn’t seem to obstruct flow much.
I'm not thrilled with the Optimus water block... Last night I tried Prime95 torture test and the CPU (5900x) started throttling above 80C while water temps were around 40 which seems like a large discrepancy. Increasing pump speed did nothing. I tried applying thermal paste twice, once with Kryonaut and the other time with Kingpin paste, but results were pretty much the same. I ordered a Heatkiller pro block and liquid metal compound. I will try to swap that in and also install the flow meter while I'm at it. Will report if the temperatures are any better with that combination.
Interesting... I have 2x240mm HWlabs radiators which are way more than needed for just a CPU. I could ramp up the fans, but I prefer silent operation and since I have norprene tubing I'm more than happy to let the water temps rise up to 50C. If the heat transfer between the block and water is effective I would expect water temps to rise well above 40C before the CPU hits its thermal threshold. So my primary suspect is the block.
Last edited:
But you have to understand that 40 degree delta between cpu and water is normal for zen 3. You wont find the block thats much better. You should focus on lowering water temp.
With 420 rad for 9900k and 3080 i have about 14 degree delta water temp(39 total water temp), but thats pulling 500watts, not 150 as in your case.
With 420 rad for 9900k and 3080 i have about 14 degree delta water temp(39 total water temp), but thats pulling 500watts, not 150 as in your case.
