UNSOLVED
C
coldfish_91
104 Posts
2
23788
November 22nd, 2019 21:00
Aurora R9, DIY project
Just got my brother a new Aurora R9 as a gift. Here is the specs: 9900K+ 2080Ti + 850W PSU+ Lunar light
We saw a few interesting posts in this community (very helpful and informative) and decided to do some DIY mods by ourselves before PC arrived. Basically, there are three phases in our project.
Phase 1: Replace top and front fans. Add push-pull setup for CPU cooling, and extra top front intake fan. Replace memory with 3200 ones.
front intake fans -> Gentle Typhoon 2150RPM 4-pin Using Noctua Y-splitter and connect to front_fan header
top fans-> ML120Pro (push-pull) Using Noctua Y-splitter and connect to top_fan header
Memory-> G. skill gskill trident z 2*16GB 3200 CL14
We never tested the stock fans. First thing we did after taking it out of box was replacing the fans. Now, idling is about 28-30 degrees on CPU, about 32 degree on GPU, and noise is very minimal after we adjusted the curve. Full load, CPU is about 55-60 and GPU is about 70.
For pushj-pull setup, l was considering to add one extra fan on top of radiator directly after removing top cover however, there is no enough space between top cover and top bezel to fit the pull fan. You have to remove the whole CPU cooler to add the fan. And it is quite tight so be careful not to damage any component when installing.
For top front fan, thanks to this community, I find the weather strip is the best way to secure the fan there. Otherwise, only 90mm fan can be fit in with the HDD cage.
However, there are a few problems here after replacing the fans:
1. AWCC only displayed % RPM reading. Even if we set it to max manually, it would only show about 59% for top fan and 45% for front, because the software is using stock fan max RPMs to calculate the %. ML120pro has 2400RPM which is about 60% of stock top fan max RPMs. We tried some other software but only Intel XTU can give you the RPM readings. So if you replace the fan, don't worry about the % reading.
2. If I manually set both fans at max before stress test, top fan would slow to min RPM. If I set top fan at 70% speed, then front fan will slow. If I set both fan at 70% speed, no issue. I also had both fans running at max for 30 mins while PC idling and no issue. There are two possible reasons, one is unstable voltage when CPU is consuming more power. Second is unstable PWM signal. IMO, this problem seems to be related with PWM signal since CPU pump fan is working perfectly under stress test all the time, which shares the same +12V source with top fan but without PWM control. If it was becasue of unstable voltage, then pump fan should also slow but it did not.
Here are a few pics
Phase 2: Upgrade 2080Ti blower to water cooled. To be continued。。。。。
2130
0
Posted December 4th, 2019 18:00
Nice work.......great temps!
Amazing how that 15mm fan performs pretty darn well.
I'm sure that machine will perform fine at stock CPU clocks. You can always tweak the fan speeds to reduce noise levels as real world gaming stress is a lot different from those 100% load benchmarks tests.
2375
0
Posted December 4th, 2019 18:00
This is what everyone wants to do. And you actually went through with it..
Which M.2 heatsink did you use?
How did you ever get the top of the case removed?
I personally would rather remove the mobo, and radiator assembly to add the push pull fans.
2167
0
Posted December 4th, 2019 18:00
@Bmwpower603 start from rear end, just pull the cover up with force. I have uninstall and re-install the top cover for 10 times, no issue so far.
Heat sink: ELUTENG M.2 Heatsink 2 Sets, NVME Heat Sink Alloy Aluminum Double-Sided M2 2280 SSD Cooler Fits for PCIe NVMe NGFF SATA with 4 Silicone Thermal Pads
2097
0
Posted December 4th, 2019 19:00
I ordered a Corsair H80v2 by mistake, 120mm *49mm thick rad, I am thinking about trying it up on CPU and see how much it can outperform stock 120mm*25mm rad,
That might just be what you are looking for to control the CPU OC temps..........wouldn't that be nice. Go for it.
2688
0
Posted December 4th, 2019 20:00
I ordered a Corsair H80v2 by mistake, 120mm *49mm thick rad, I am thinking about trying it up on CPU and see how much it can outperform stock 120mm*25mm rad,
That might just be what you are looking for to control the CPU OC temps..........wouldn't that be nice. Go for it.
The community did talk about max thickness of the top radiator going into R9 chassis a while back. I think @Cass-Ole summarized it best here:
You should be fine with 49mm radiator + 25mm fan. Looking forward to this new upgrade.
2457
0
Posted December 6th, 2019 05:00
So I succefully fitted H80v2 thick rad into the case last night, with 25mm push fan + 15mm pull fan, which is about 89mm, no issue. The max you can go is about 92-94mm based on my measurement.
H80v2 has not only larger rad but also much thicker tube (probly 100% thicker than stock). It is quite hard to find the right position for them in order to close the PSU. After spending hours playing around, I only find one particular position to make it work. However, I was running out of my thermal paste but I will finish it today and post results here later on. Here is the pics of the right position for tubes which can close PSU easily.
Anyone has knowledge about different fans used in push-pull setup ? Should I put stronger fan as pull or push ? Will that really matter ? Or shoud I just go with single push fan with ML120Pro ? Need some ideas here.
Edit: pics posted in next post
2441
0
Posted December 6th, 2019 06:00
2502
0
Posted December 6th, 2019 07:00
@coldfish_91 , looks good! I'd be cautious about that tubing in the second picture though. The one over the DDR. It looks like a pinch point but I'm sure you're aware of it and making sure everything is fine when you close the PSU swing-arm.
I'd go for homogeneous push-pull i.e. same fans but in this case, 25mm x2 is unlikely. You'll have to find the sweet spot of matching the fans, maybe throw in some creative play of LNA adapters.
We've been discussing whether the default thin radiator with push-pull beats thick radiator with single fan. Looks like you're best positioned to let us know the answer. Thanks!
coldfish_91
104 Posts
2157
0
Posted December 4th, 2019 16:00
Besides, I would highly recommend adding heat sink on nvme SSD, which can reduce the temperature for more than 10 degrees in my test. Constantly running nvme SSD on high temp will reduce its life. Also, With 240mm AIO being fitted, nvme SSD sits quite close with radiator (heat source). Without heat sink, temp can easily reach 70 degrees under stress test or gaming, but now it barely touches 50.
In our case, we need some low-profile heat sink otherwise, it may not fit with VGA card. Here is the one I am using:
ELUTENG M.2 Heatsink 2 Sets, NVME Heat Sink Alloy Aluminum Double-Sided M2 2280 SSD Cooler Fits for PCIe NVMe NGFF SATA with 4 Silicone Thermal Pads