The tube ports on that water block look pretty chunky. I would look at the EK Vektor. I've used EK Quantum Torque fittings, and yes, that's ZMT 10/16 tubing.
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
Log Yet Another NCase M1 (Zero compromise edition)
- Thread starter Kitsune
- Start date
The tube ports on that water block look pretty chunky. I would look at the EK Vektor. I've used EK Quantum Torque fittings, and yes, that's ZMT 10/16 tubing.
Riddle me this. Here is where it starts to become increasingly apparent that I don't know what I am doing. These (pictured) are the fan and pump i/o ports I have on my motherboard. I have 4 fans and a pump. The mobo has 2 fan i/o's and a pump i/o. I assume I need some kind of splitter. Can someone tell me what the defacto solution for this is?
Last edited:
Pump into CPU_FAN so you can take advantage of the rpm sensor in case the pump fails in the future. Then you would need to splitter either one of the headers and control them all together, or maybe splitter them both so you can have two groups of fans.Riddle me this. Here is where it starts to become increasingly apparently I don't know what I am doing. These (pictured) are the fan and pump i/o ports I have on my motherboard. I have 4 fans and a pump. The mobo has 2 fan i/o's and a pump i/o. I assume I need some kind of splitter. Can someone tell me what the defacto solution for this is?
If you decide to power all 4 fans off a single header, I would err on the side of caution and get a powered splitter, although even if it is a passive splitter powered by the header it should be fine with most fans like Noctua.
I use this guy in my NcaseM1 build.
Another question. The EK DDC Pump has power and PWM separated, unlike the 4 Noctua fans which have power and pwm in the same small header. Is this necessary? Can this be rewired to get power and pwm from the CPU_FAN header or does it need more juice direct from the PSU?
So I researched and answered this myself. The DDC draws 18W, and the motherboard headers output a max of 12W. That sucks.
Riddle me this. Here is where it starts to become increasingly apparent that I don't know what I am doing. These (pictured) are the fan and pump i/o ports I have on my motherboard. I have 4 fans and a pump. The mobo has 2 fan i/o's and a pump i/o. I assume I need some kind of splitter. Can someone tell me what the defacto solution for this is?
I bought this splitter for my case: https://www.ekwb.com/shop/ek-cable-splitter-4-fan-pwm-extended
Yeah, you reached the right conclusion, no go for header powered DDC pump. I think there is some varients that are limited to 14W or 18W or so, but otherwise you risk burning the header so I wouldn't recommend it anyway.
Currently I have the fan hub just above the GPU, underneath my rear intake fan. To make it fit I cut off the mounting plastic holes. This obviously blocks the iceman ports, so maybe in your case you would consider where the fan would normally be. It's just there by double sided tape
I used to have it on the PSU, but now that area is occupied by my routing of the 24-pin cable, that's why it was relocated.
Another advice: Flush that XSPC radiator at least 4-7 times.Look what came out of mine which arrived on Saturday. You don't want these metal parts in your pump.
Hi Kitsune, I'm really enjoying this build log, thank you for posting everything in such detail! This is my dream build so this thread is invaluableThe EK fittings and DDC pump arrived today!
Pump is mounted to the res, and I've had a play with the compression fittings and put the first tube in, from the CPU block to the res. Still putting off cable work...
I wanted to ask if the above 45 degree fittings wont cause issues with the thick side radiator as @Blubbie21 mentioned earlier? Also, I've done some research into pwm fan & pump headers and the aquatuning quadro comes up quite a lot as quality hub for pwm control. I believe it comes bundled with aquasuite which helps you control everything on the software side. I've thought about throwing a temp sensor in too in order to tie fan/pump rpm to water temp.Going 16/10 all the way. EK Tube ZMT 15.9mm OD/9.5mm ID. Using Alphacool fittings since they are easy to source in Germany.
Well good luck.
I had troubles with my project because in the end everything was to too tight - but mainly the GPU paired with the GPU-waterblock was to wide to fit the NCase ? ? ? ? - finally i had to drop my custom build and switch to an air-only-build.
I still will try another attempt with the custom loop - with the next / new GPU generation.
So you're fine with the 16/10 fittings? For me they felt like to big and the soft tubes were to hard to bend on the shorter parts of the loop. (like the part: Icemen-Reservoir => CPU-Block and bottom radiator => GPU-Block)
I had troubles with my project because in the end everything was to too tight - but mainly the GPU paired with the GPU-waterblock was to wide to fit the NCase ? ? ? ? - finally i had to drop my custom build and switch to an air-only-build.
I still will try another attempt with the custom loop - with the next / new GPU generation.
So you're fine with the 16/10 fittings? For me they felt like to big and the soft tubes were to hard to bend on the shorter parts of the loop. (like the part: Icemen-Reservoir => CPU-Block and bottom radiator => GPU-Block)
Hey guys, I am planing on building a similar build to most other Ncase M1 builds in this thread. I am thinking about which radiators to get and after some research the HWlabs Nemesis 240 GTS seam to be the best, why are you guys using XPCS rads?
Because it leaves a decent gap between the GPU and the fans to guarantee some decent airflow. Most thicker radiators will have the fans blocked by the GPU, with very limited thermal benefits. No airflow means no heat blown away from the radiator. For the side radiator there are more viable options available.
Well good luck.
I had troubles with my project because in the end everything was to too tight - but mainly the GPU paired with the GPU-waterblock was to wide to fit the NCase ? ? ? ? - finally i had to drop my custom build and switch to an air-only-build.
I still will try another attempt with the custom loop - with the next / new GPU generation.
So you're fine with the 16/10 fittings? For me they felt like to big and the soft tubes were to hard to bend on the shorter parts of the loop. (like the part: Icemen-Reservoir => CPU-Block and bottom radiator => GPU-Block)
Sad to hear about this. Having got so far into the build in your situation I would honestly have considered cutting a hole in the side plate. I've actually seen this done, perhaps on Reddit. If I find the picture again I will share it with you!
So August has been pretty busy with work, holidays etc which I'm citing for the lack of updates on my build. Quite a lot has happened, and is still happening.
I decided after spending god knows how much on MPDC-X cable sleeving, cables, terminals, crimps, heatshrink, more tools that cutting custom cables was holding me back from getting this little beast turned on, so I buckled and decided to go ahead and stuff in the stock PSU cables. The biggest road block was the large number of double cables in single terminals on both the PCI-E 8-6+2s and the 24 pin. Sleeving and terminating 2 cables in a single terminal is more of a PITA than I could be bothered with, and I honestly just wanted to get this thing fired up after staring at it's lifeless shell for literally months.
So here's some updates from the last few weeks:
Adding more tubing and figuring out positioning for the XPSC rad in the bottom. I decided at this point that I wanted stand-offs from this rad with 90º fittings to bring the tubing horizontal over the GPU. The PCI-E cables were only in at this point because I was figuring out how long my custom-never-to-be cables should be.
This was a pain to get right. I think I've spent as much on shipping costs as I have done on fittings. The right combo of stand-offs and connectors to get to a compression fitting and keep it super-close to the fan so it all fits in the void between the fan and the PSU.... Urgh. It looks mental. What I needed was a quick-release with a female G1/4 to go straight to the compression fitting, but whatever.
I switched the fans around on the rad so the PWM cables kinda meet at the same place.
By this point I was entirely unprepaired to wait any longer. After a lot of faff with tubing and fixings. This tubing it heavy-weight stuff. Getting the compression fitting attached at either end is finger-shredding work. Basically you have to attach one end, then push the tubing onto the other end in-situ. I decided to use a small transparent piece as that can be seen under the rad when you take the side plate off. Kinda fun.
I decided after spending god knows how much on MPDC-X cable sleeving, cables, terminals, crimps, heatshrink, more tools that cutting custom cables was holding me back from getting this little beast turned on, so I buckled and decided to go ahead and stuff in the stock PSU cables. The biggest road block was the large number of double cables in single terminals on both the PCI-E 8-6+2s and the 24 pin. Sleeving and terminating 2 cables in a single terminal is more of a PITA than I could be bothered with, and I honestly just wanted to get this thing fired up after staring at it's lifeless shell for literally months.
So here's some updates from the last few weeks:
Adding more tubing and figuring out positioning for the XPSC rad in the bottom. I decided at this point that I wanted stand-offs from this rad with 90º fittings to bring the tubing horizontal over the GPU. The PCI-E cables were only in at this point because I was figuring out how long my custom-never-to-be cables should be.
This was a pain to get right. I think I've spent as much on shipping costs as I have done on fittings. The right combo of stand-offs and connectors to get to a compression fitting and keep it super-close to the fan so it all fits in the void between the fan and the PSU.... Urgh. It looks mental. What I needed was a quick-release with a female G1/4 to go straight to the compression fitting, but whatever.
I switched the fans around on the rad so the PWM cables kinda meet at the same place.
By this point I was entirely unprepaired to wait any longer. After a lot of faff with tubing and fixings. This tubing it heavy-weight stuff. Getting the compression fitting attached at either end is finger-shredding work. Basically you have to attach one end, then push the tubing onto the other end in-situ. I decided to use a small transparent piece as that can be seen under the rad when you take the side plate off. Kinda fun.
No going back. What leak test?? I plugged everything in, had a large roll of paper-towel handy and started filling.
Turned it on and got a VGA error during POST. After a load of troubleshooting, sweating and 1 sleepless night I came back to it to discover I hadn't plugged in one of the GPU power cables. Text book. Plugged that in and....
Lift off! No RPM being reported by the DDC though, which is plugged into the CPU PWM header.
Was still open case testing at this point, gaining confidence in the loop.
Was pretty excited to see it coming to life. Love the small transparent tube with red coolant. No RGB on the GPU block though...The GPU block I didn't want any RGB on but ordered the wrong block and got it anyway.
After a litte more tinkering and plugging in the RGB to the right header the right way round the GPU waterblock sprung to life. Time to close up....
So this is pretty much it! The side plate doesn't fit absolutely flush because of the screws I've used to mount the side rad to the case, I'll deal with that soon. Debrief and lessons learnt coming soon.
Last edited:
I do however, have what you might call a slightly large problem. After getting the system set up, windows installed, Steam, 3DMark, buying Flight Sim 2020, I discovered that under heavy load my 2080ti is black-screening after, well, not very long. I've been around the internet with this, and there doesn't seem to be any single magic bullet. I've used DDU and done clean installs of latest NVidia drivers. I've messed around with temperatures, fans, even under-volting the card. I've just ordered a Corsair SF750 in the hope that it's power related. If it's not, it's going to be a drain and a tear-down to troubleshoot RAM, CMOS and everything else I can't reach which I obviously want to avoid if I can. However if I do have to, I will reconnect and fill the loop on an open board and continue troubleshooting and benching open before reintroducing everything to the case. There's a whole bunch of stuff I don't really understand yet. I wish I had got a water temp monitor in before I filled the loop. Fingers crossed on the power supply for now, but failing that, I'm going to need some help folks
Temperatures are pretty incredible. At idle the 2080ti ticks over at 30ºC / 86F.
Temperatures are pretty incredible. At idle the 2080ti ticks over at 30ºC / 86F.
So it turns out the PSU wasn't the problem. I'm now well on my way to having enough spare components to build another computer. After having a chat with the very helpful @Qzrx over on Discord I've come around to the understanding my GPU is running far too hot. I've drained and taken everything apart. Next, I'll disassemble, clean and reseat the GPU water block with (more) fresh TIM and hopefully get better contact.
Other small jobs will involve popping the CMOS. and testing the RAM. I'm interested in testing all this before putting it back in the case, but I need to learn more about doing that with a custom loop. Mainly, how does one drain a loop if all your rig is spread out over a table.....
Other small jobs will involve popping the CMOS. and testing the RAM. I'm interested in testing all this before putting it back in the case, but I need to learn more about doing that with a custom loop. Mainly, how does one drain a loop if all your rig is spread out over a table.....
