Black copper NCase M1

[confusis]

Epic work, proper SFF right here.

 

[iFreilicht]

Painted, vandal switch with added resistance installed

That is soo cool, do you have any pictures of that front I/O with the front panel attached? I guess that is a 12mm switch, is that correct?

 

[BaK]

Epic work, proper SFF right here.

Thank you, appreciate it!

That is soo cool, do you have any pictures of that front I/O with the front panel attached? I guess that is a 12mm switch, is that correct?

Not at hand but I will post such a pic soon np.
This is indeed a 12mm switch.
I guess I will have to mod the bracket for a 16mm RGB iFreiSwitch when it will be available! (just catching up on your amazing thread )
Or wait for the 12mm version...

 

[BaK]

Some more pics before final ones next week!

Almost done, tubing, pump and MCubed FanAmp added

As you have noticed, I have flipped the PSU, it is now placed with its fan on the outside of the case.
My first idea was to have it the other way, in order to remove its (loud) fan and use the air flow coming from the 120mm fan of the push/pull unit.
Cable management made me change my mind, especially the CPU 8pin cable

Coming from the Silverstone PPO5-E short cables set, it was just long enough to reach the PSU with that route under the mobo heatsink.
I could have made a custom cable (that's still a plan B if needed) but here each cable are attached to the next one and is more elegant that way I think.

Have a look at the zero tolerance fit of the GPU backplate and the mobo I/O

And another mutliple zero tolerance:
- 90° fitting in contact with the mobo heatsink under it and also touching the mobo I/O
- straight fitting (had to find a narrow one) touching the glass of the res
- 45° fitting touching the CPU block fixation

Foam has been added into the res btw.

 

[iFreilicht]

I love it when bits touch. It's looks so... efficient.

 

[|||]

Your reservoir will not act as a reservoir in that place in the loop. It needs to go immediately before the inlet of the pump for it to function as it should. Otherwise, the pressure head needs to be maintained through the reservoir and can't compensate for lower fluid levels. However, if a system is adequately de-aerated, it shouldn't need a reservoir and could work with only very occasional top-offs to compensate for permeability and any leaks in the system (looks like you have low permeability tubing, so shouldn't be that bad).

 

[TheDreamingMonk]

Your reservoir will not act as a reservoir in that place in the loop. It needs to go immediately before the inlet of the pump for it to function as it should. Otherwise, the pressure head needs to be maintained through the reservoir and can't compensate for lower fluid levels. However, if a system is adequately de-aerated, it shouldn't need a reservoir and could work with only very occasional top-offs to compensate for permeability and any leaks in the system (looks like you have low permeability tubing, so shouldn't be that bad).

For all intents and purposes, reservoirs in computer cooling systems are more for looks than they are useful. As long as the pump is at the bottom of the loop or at least the majority, order and placement does not matter. Even if a computer water cooling system isn't properly de-aerated, you still don't need a reservoir. In such a small closed loop, you will not have fluid levels fluctuate enough to cause any issues.

Though I will agree it is not efficient with the way it is. Without making sure the res is de-aerated and is properly pressurized, the next component in the loop will be getting less flow. Still not efficient. But not a detriment to the system.

 

[|||]

...which is why I mentioned the a system shouldn't need a reservoir, but it needs to be set up very well.

Even if a computer water cooling system isn't properly de-aerated, you still don't need a reservoir.

If you do not de-aerate and do not have a reservoir, you will have air bubbles travelling into the pump, causing massive cavitation on the low-pressure side of the pump impeller. This will significantly degrade the performance and efficiency of the pump, as well as cause excessive noise. (side note: higher temperatures of the working fluid can cause this as well)

The typical operation of a reservoir collects the fluid after it has traveled through the loop of components. However, with the fluid simply dumping into an unpressurized vessel, the pressure head of the working fluid is completely lost. The excess fluid allows for changes in fluid levels without changes to performance of the loop, as the pump is drawing fluid from the bottom, where the is supposedly only fluid, and not air, to draw upon. So there is a function to a reservoir; although, this function is usually over-emphasized and not critical to the performance of the loop itself.

In such a small closed loop, you will not have fluid levels fluctuate enough to cause any issues.

Actually, unless you have a pure metal set-up, plastic will have a certain amount of permeability, allowing fluid to leach through and leave the system. This should only happen after a significant amount of time (greater than a year) before it causes a need to top-off the loop to keep the system operating optimally.

 

[jeshikat]

Actually, unless you have a pure metal set-up, plastic will have a certain amount of permeability, allowing fluid to leach through and leave the system. This should only happen after a significant amount of time (greater than a year) before it causes a need to top-off the loop to keep the system operating optimally.

So are closed loop coolers using tubing with particularly low permeability since they're advertised as no maintenance?

 

[|||]

Yeah, you'd hope so with 5 year warranties and not too much fluid to begin with.

Take a look at vapor barrier's. Here's a test of silicone with different fillers and their affects to permeability.

 

[BaK]

Your reservoir will not act as a reservoir in that place in the loop. It needs to go immediately before the inlet of the pump for it to function as it should. Otherwise, the pressure head needs to be maintained through the reservoir and can't compensate for lower fluid levels. However, if a system is adequately de-aerated, it shouldn't need a reservoir and could work with only very occasional top-offs to compensate for permeability and any leaks in the system (looks like you have low permeability tubing, so shouldn't be that bad).

For all intents and purposes, reservoirs in computer cooling systems are more for looks than they are useful. As long as the pump is at the bottom of the loop or at least the majority, order and placement does not matter. Even if a computer water cooling system isn't properly de-aerated, you still don't need a reservoir. In such a small closed loop, you will not have fluid levels fluctuate enough to cause any issues.

Though I will agree it is not efficient with the way it is. Without making sure the res is de-aerated and is properly pressurized, the next component in the loop will be getting less flow. Still not efficient. But not a detriment to the system.

Hey guys, thx for joining!
Noob in watercooling here and willing to learn!

My loop order is
Res - CPU block - Pump - Rad 120 - GPU block - Rad 240 - Res
Fill port on the res and there is a drain port at the pump outlet.

I would of course have put the res right before the pump if I could, but believe me there is no other option with this setup.
Actually, as I was told wisely, having the CPU block in between doesn't seem to interact anyhow.
I had the drain opened while I started the filling, and water came out as soon as I put some into the res. Doesn't that mean there is pressure head?

Anyway filling/bleeding the loop was done without much trouble.
I am sure there was always water in the pump. Its gurgling sound decreased proportionally to the air going out of the loop.
It is now dead silent at lower speed.

The res did its job of removing bubbles. There are still some trapped in the foam, but from what I see there are slowly going up and gathering at the top. I will do a top off when most of them will be up there.

Other plus of having a res is the fill port and the ability to see the water level.

 

[|||]

The little bit air in there is a compressible fluid; think of it being like a spring acting on the water. Think about if you were to press on something, but it is spring loaded; that's the effect it has on the water in the loop. It's not super detrimental, but it isn't going to function like a traditional reservoir. It can still be used as a visual gauge to judge if additional water needs to be added to the loop.

 

[BaK]

That is soo cool, do you have any pictures of that front I/O with the front panel attached? I guess that is a 12mm switch, is that correct?

I should get a proper light for shooting pics, will do for my next build!

The little bit air in there is a compressible fluid; think of it being like a spring acting on the water. Think about if you were to press on something, but it is spring loaded; that's the effect it has on the water in the loop. It's not super detrimental, but it isn't going to function like a traditional reservoir. It can still be used as a visual gauge to judge if additional water needs to be added to the loop.

What is not making it a traditional res? Because it is not directly connected to the inlet of the pump or because the fill port is not on top?

While I understand you point, I don't see how it is different from a res like this one:

Isn't the air on top of the res also acting like a spring as you described it?

 

[EdZ]

For the extremely low flow rates and pressures in a PC watercooling setup, the effect of trapped air compression in operation is so low to be negligable. And unless you're running very short hardline runs, the flexibility of the tubing walls will be enough to account for changes in loop volume from ambient pressure/temperature changes (which are going to have to be small anyway, due to the tolerances of the computer working around them!).

The main purpose of a reservoir for PC watercooling is to make filling and bleeding easier. If you want to avoid installing one, then a removable T-line is a good substitute: add a T fitting at the highest point practical in the loop, and add a long tube extending from this with a funnel to fill with. Fill the loop until the water runs up the line above level with the topmost portion of the loop, clamp the line (either with a bulldog clip, or your hand if you want to live dangerously), run the pump for a bit to push any trapped bubbles around the loop up into the T-line. Uncap the T-line, drain it down the the T-fitting, then remove the filling line and cap off the T-fitting. Unless you expect to be draining and filling the loop often, then you trade a bit of hassle when filling for not having a big tank of water sitting in the case doing very little of value (except maybe making a nice stagnant area for things to grow in depending on port design and layout).
The only practical difference in performance adding a reservoir to a watercooling loop is going to make will be in increasing the time constant of the loop by increasing the fluid volume (i.e. increasing the time needed for the fluid to warm up to equilibrium for any given heat input). Water has a pretty excellent heat capacity, and unless you have a very small radiator or very little airflow you're not going to be running a WC loop whose equilibrium temperature is anywhere close to where it would be unable to sink heat from even a high power CPU/GPU, so there isn't a risk of 'overloading' the loop.

As for cavitation: no PC watercooling pump is powerful enough to cause cavitation. If it was, you'd know about it pretty quick due to the plastic impeller exploding violently. You could damage the impeller from the pump running dry, but if your loop is that badly bled then a reservoir is not going to help you!

 

[BaK]

For the extremely low flow rates and pressures in a PC watercooling setup, the effect of trapped air compression in operation is so low to be negligable. And unless you're running very short hardline runs, the flexibility of the tubing walls will be enough to account for changes in loop volume from ambient pressure/temperature changes (which are going to have to be small anyway, due to the tolerances of the computer working around them!).

Thx EdZ for the very detailed explanation!

Still wondering though why | | | thinks my res is not a traditional one...?

The only practical difference in performance adding a reservoir to a watercooling loop is going to make will be in increasing the time constant of the loop by increasing the fluid volume (i.e. increasing the time needed for the fluid to warm up to equilibrium for any given heat input). Water has a pretty excellent heat capacity, and unless you have a very small radiator or very little airflow you're not going to be running a WC loop whose equilibrium temperature is anywhere close to where it would be unable to sink heat from even a high power CPU/GPU, so there isn't a risk of 'overloading' the loop.

So adding the 120mm rad, if not helping the cooling due to its slim fan spinning slowly, will at least be helpful in regard of the time constant of the loop.

To stay on this topic, some pics of the filling process:

 

[BaK]

As I wanted to decouple the pump, I have put it on a foam and attached them to the case with a velcro tape.
Looking for the right piece of foam, I stumbled upon some old acoustic one I never used to make my rigs more silent (as they already were ). Thus came the idea to set a whole layer of foam on the bottom of the case, to put the pump on it but also to help for better cables management in hiding them under it.

I am still doubtful if I like it or not, tell me what you think!

Everything in!

Customized power cord

 

[|||]

Still wondering though why | | | thinks my res is not a traditional one...?

An analogy: you have a maze you're going to push marbles through. There is carpeting, so there is substantial resistance to the marbles moving (to emulate restrictive passageways of heat sink channels and radiator tubes). You push a marble into the maze, but it stops not too long after it enters. You push a second one in and it pushes the first one a little bit further, but they both then stop. You continue to do this and get the marbles pushing through the system up to a reservoir. At that point, if there are still other pathways the marbles need to go before getting to where you can take them to push them back into the loop, the marble will need to go into the reservoir and completely/mostly fill the reservoir before the will start pushing marbles through the subsequent pathways. If the reservoir is at the end, where you able to pick them to push back into the loop, then there is no need to fill up the reservoir for the loop to function, and the reservoir allows for different levels of marbles without affecting the ability to pick another marble that has gone through the loop to be put back in.

For a water loop with a reservoir immediately before the pump, the fluid has gone through all of the components in the system and is then dumped in the cylinder. The motivating force drawing the fluid into the pump in the negative pressure created by the pump and gravity (not much from gravity other than keeping the liquid toward the side of the pump intake) and can compensate for different fluid levels and allow the pump to always draw a constant amount of fluid (assuming no vortex or aeration issues). If you have the reservoir in the middle of the loop, then it will need to fill it up to maintain the pressure across it to continue the pressure through the subsequent components in the loop.

So adding the 120mm rad, if not helping the cooling due to its slim fan spinning slowly, will at least be helpful in regard of the time constant of the loop.

Adding additional water to the system increases the thermal capacity of the water. You need to think about it in units of energy. 1 calorie will raise the temperature of 1 gram of water by 1 degreeC. 1 calorie will only raise the temperature of 2 grams of water by 1/2 degreeC.

 

[BaK]

For a water loop with a reservoir immediately before the pump, the fluid has gone through all of the components in the system and is then dumped in the cylinder. The motivating force drawing the fluid into the pump in the negative pressure created by the pump and gravity (not much from gravity other than keeping the liquid toward the side of the pump intake) and can compensate for different fluid levels and allow the pump to always draw a constant amount of fluid (assuming no vortex or aeration issues). If you have the reservoir in the middle of the loop, then it will need to fill it up to maintain the pressure across it to continue the pressure through the subsequent components in the loop.

Thx a lot, got it now!

Anyhow I managed to fill the loop without problem.
So from what I see there was only a small pressure drop in the reservoir, or the negative pressure from the pump was enough to compensate the CPU block constraint/resistance.

 

[BaK]

Small update, CPU has been delidded!

Actually the 3V CMOS battery has been changed too...

I had an issue at cold starts. I was receiving an 'NTLDR is missing' error during the boot of Win 10, installed on a 256GB 960 EVO M.2 in the slot at the back of the mobo.
UEFI was reset to default.
A trick to make it boot again was removing the M.2 disk, try to boot, putting the M.2 back, success boot!

Because of the UEFI reset I put the blame on the CMOS battery and was glad to see it was in an easily reachable area! *caugh*
Do you see it?

Stucked to the I/O port mobo pack pannel, had to move the res to reach it

So I replaced it, had to made a custom one as Asrock used a laptop kind of attachment instead of a regular socket for the battery.
Asrock one was 3.0V, so I guess still good. New one is more like 3.2V

Delid was then done without problem thanks to Chri's tool.
Foam was placed inside the tool to prevent the PCB to hit it but it wasn't really necessary. The PCB was detached smoothly as I went slowly, letting time to the Intel silicon stretching itself at each 'turn' on the vice.

Four drops of liquid metal tape for relidding let me put the CPU back in its socket with the mobo in a vertical position.

CPU is now running 10-15°C colder and with all cores within a 3°C diff (10°C between two cores before delid).
I've used CooLaboratory Liquid Pro on the die and Liquid Copper on the CPU block.

Unfortunately the computer is still having a cold boot problem.
It starts but after 4-5 seconds reboots itself,the UEFI is set to default again but then and new error message appears. *need to write it down next time it happens*

From what I see it happens most when I didn't start it for 48 hours or longer.

New trick now to make it go is to load a saved UEFI profile and simply reboot.
What I don't understand is how come the error message has changed and why the UEFI is turned back to default while the saved presets are still available. And why loading one gets rid of the boot problem...
Heard cold boot problems can come from a bad peripherals sometimes. RAM, M.2, PSU?
If anyone got a clue?

Well I simply have to game every day!

 

[aquelito]

Don't know how I missed this terrific build...

How do you like your Tygon A60G Norprene tubing ?
According to the specs, it has a pretty low bending radius (7/8 inches vs 1 for B-44-3) but it is prone to kink quite easily.