Production FormD T1 Classic (READ FIRST POST)

[gwertheim]

I would like more tests)))))) if you can do something else all the users of this forum will tell you, as always, a million times thanks

@fabio has tested so many builds, he can't test every combination of parts out for sale.

There's a point where you just need to test yourself when you get the case.

 

[xkruz]

I've done some test with some combination
Since I don't have flow meter, I use reservoir to measure the flow ?
This is only for the flow, not yet tested on running computer

Test setup 1:
Alphacool Eisbaer Lt Solo Block
TPV Tubing
Magicool G2 240
Bykski reservoir

Result: flow are OK, I think this will work fine

Test setup 2:
Alphacool Eisbaer Lt Solo Block
TPV Tubing
TPV Quick Disconnect
Magicool G2 240
Bykski reservoir

Result: flow still OK

Test setup 3:
Alphacool Eisbaer Lt Solo Block
TPV Tubing
HSPC TX240
Bykski reservoir

Result: Flow are bad, so low I can't hear the water splashing in reservoir, and with this setup it's impossible to fill the loop just using the eisbaer dc-lt so I use syscooling ddc pump to fill the loop.

Video of the 3 setups
Flow test (sort of)
PS:just find out imgur changed video to GIF so there is no sound. My bad

So, I think tx240 need to be paired with more powerful pump.
I will test it on my 3950x soon

 

[Gerold]

Sure!!!
But do you think it’s enough just the last build?? ?

I think your latest build, where you said you will go with it in the future, should be sufficient for an update in your thread.
It seems like the most thought through and tested environment imo.

 

[ExplodingWaffle101]

Guys, this is my last setup, and I can say is incredible, I am keeping it!

D5 set up with a range of 60-80% to have a minimum flow of 100l/h
Noctua A12x25 25-90%
Scythe Slim 20-60% (I want it quiet)
PBO tweaked at PPT 145, TDC 100, EDC 130
GPU Undervolted 925mv @ 1920 MHz

IDLE: 27dB (front at 30cm)
CPU LOAD - multithread, 16 core: 30-31dB (front at 30cm) - Blender Rendering 30min
GPU LOAD+CPU Single thread: 33-34dB (front at 30cm) - Heaven 4.0 loop 20min
(@threestripevida)

GPU


CPU

did i miss how you got a d5 into the case?
a12x25 OP, i really cant see any build with 2 slim fans working the best

 

[Cronos83]

On the contrary, push/pull makes more sense if you are runing at low rpms

Yeah, but NOT on slim low FPI radiators.
So what is your point?

That is exactly what I meant.

Push-pull on slim low FPI rads makes absolutely no sense.

 

[fabio]

did i miss how you got a d5 into the case?
a12x25 OP, i really cant see any build with 2 slim fans working the best

Here you are! Is almost inside ?

On the right the Air-cooled Ghost for comparison!

 

[fabio]

BTW, I've checked the Eisbaer LT Pump to see if there was any issue because of the low flow with it in the loop.

Well, it turned out that the CPU block is super dirty and all the fins are blocked! That's why!.

BUT, that was the FIRST TIME I was using it, and all the other parts were flushed and cleaned from the last months of use.

Pretty bad IMO and bad QC, so, just to be sure, OPEN your Eisbaer LT block before to install it! ?

 

[Cruella Devill]

BTW, I've checked the Eisbaer LT Pump to see if there was any issue because of the low flow with it in the loop.

Well, it turned out that the CPU block is super dirty and all the fins are blocked! That's why!.

BUT, that was the FIRST TIME I was using it, and all the other parts were flushed and cleaned from the last months of use.

Pretty bad IMO and bad QC, so, just to be sure, OPEN your Eisbaer LT block before to install it! ?

Will you do the tests again after cleaning? How long was the pump without cleaning?

 

[fabio]

Will you do the tests again after cleaning? How long was the pump without cleaning?

No, I will not, I am happy with the actual setup. But I've tested that 1 DC-LT is fine with a loop with low restrictive components. And also Optimum tech showed that is fine as well on his videos.

Well, the pump was new/never used, so it wasn't supposed to be cleaned.

 

[F7IN]

I am, there were no problems beside having to sand some of the radiator down to make it fit (but I think this was fixed in the 1.1 version of the case). If you have tall RAM (I am currently using Corsair LPX) then there is a high possibility that you will run into problems getting the tubes to fit in the case so you can close the side panel (only in 3 slot mode).

The only time I ran into problems was when I flipped my PSU and the bracket was pushing the tubes into the fans. This is fixed in the 1.1 version of the case though, and that will be the one that will be available to purchase in June.

You can find my posts on the build in this google doc that explain my problems, thoughts, and other things.

Perfect, thank you. I too have corsair LPX so hopefully it should all be fine

 

[xkruz]

BTW, I've checked the Eisbaer LT Pump to see if there was any issue because of the low flow with it in the loop.

Well, it turned out that the CPU block is super dirty and all the fins are blocked! That's why!.

BUT, that was the FIRST TIME I was using it, and all the other parts were flushed and cleaned from the last months of use.

Pretty bad IMO, so, just to be sure, OPEN your Eisbaer LT block before to install it! ?

Oh my, thats nasty!
I've open mine, and it's clean..

So these are the result
Tx240 alphacool lt solo pump block @100%
Tx240 Syscooling ddc @100%

Tx240 syscooling ddc @50%

The setup

Edit: all of the test above are after 5 cinebench run, and screenshot taken in the middle of 6th run

 

[fabio]

Oh my, thats nasty!
I've open mine, and it's clean..

So these are the result
Tx240 alphacool lt solo pump block @100%
Tx240 Syscooling ddc @100%

Tx240 syscooling ddc @50%

The setup

There is basically 1 degree of difference. NICE!!

I've checked the max temp of course for CB.

If you do a longer test, like rendering for 30 minutes we can see and evaluate better

 

[xkruz]

@fabio can you give me a link to your blender benchmark? I'm not familiar with blender ?‍

Note: the test above are after 5 cinebench run, and screenshot taken in the middle of 6th run

 

[fabio]

@fabio can you give me a link to your blender benchmark? I'm not familiar with blender ?‍

Note: the test above are after 5 cinebench run, and screenshot taken in the middle of 6th run

Sure, just go on Blender website and download and install the software first, 2.82a version, and then download this scene:

https://download.blender.org/demo/test/benchmark.zip

And then setup like this. Press F12 for render it.

 

[xkruz]

Thank you @fabio
Ok I will test again, and will be back with the result

 

[fabio]

Edit: all of the test above are after 5 cinebench run, and screenshot taken in the middle of 6th run

There something in the temp that is a bit weird also with the DDC.

Or maybe is just the CPU block better? But I have also the GPU in the loop...

Here my temp INSIDE the case with the same test than your, so taken on the 6th launch

Maybe try to swap 1 slim noctua with the Arctic 25mm I see in the background

PBO Teaked 145-100-130 - CB Score 9200-ish


And here OC of 4250MHz on all core

 

[xkruz]

Yeah I'm also thinking my 3950x was a little bit weird with the max temp. Because at first couple of second of CB run, the temp will jump to the max and after 10sec or so, it will stabilize.
Maybe asus x570i handle processor differently than the aorus.

I think the block doesn't matter so much, maybe 1-3° difference, anyway I will test it using other block.

Yeah the plan is to try 2x 12x15mm noctua with 1 arctic fan in push pull
And 1x 12x15mm and 1 arctic fan..

 

[DrHudacris]

BTW, I’ve checked your pics more in detail, and I don’t understand why your coolant temp are so high!!! 50 degrees is a lot!!!

Did you solve it?

It might be the way I was testing it. I ran heaven to stress GPU for 20 minutes, then with heaven still running, I ran Cinebench 5x. It's a lot of heat for the little radiator..250W GPU + 130-140W CPU. I don't like it the coolant that high; pump is rated for 50C so not safe to do that long term, but in real world, 100% CPU and GPU load doesn't happen, especially not for 20 minutes!

I will try the blender test (CPU only) on my system (2x A12x15 + A12x25 TX240) and see what I get! Thanks for sharing how you do it; I actually did not know either!

 

[bynio]

It might be the way I was testing it. I ran heaven to stress GPU for 20 minutes, then with heaven still running, I ran Cinebench 5x. It's a lot of heat for the little radiator..250W GPU + 130-140W CPU. I don't like it the coolant that high; pump is rated for 50C so not safe to do that long term, but in real world, 100% CPU and GPU load doesn't happen, especially not for 20 minutes!

I will try the blender test (CPU only) on my system (2x A12x15 + A12x25 TX240) and see what I get! Thanks for sharing how you do it; I actually did not know either!

It's good to have common test protocol and high load doesn't last else you're running old hardware, but to plan my cooling I'd rather stress it to the limit.
For me prime95 at the end of the day when even the case is warmed up. Maybe will combine with heaven for GPU but for now just random game pushing to 100%.

Edit: I was wrong about timing, it's enough just to give it time to stabilise. So heaven + prime95 for about 30min should be enough. As everything warms up ryzen gets undervolted (for me over 10W, but depends on starting temp) so CPU temps stay but chipset and vrm goes up.

 

[Luke NK]

Yeah, but NOT on slim low FPI radiators.
So what is your point?

That is exactly what I meant.

Push-pull on slim low FPI rads makes absolutely no sense.

I never disagreed with that, I was just making a comment on push/pull having better temp differences at low rpms and not highs