DAN A4-SFX: The smallest gaming case in the world

Hmm. If I replace the motherboard standoffs with the 4mm ones (posted earlier) - would the Cryorig C7 become a viable option once again, since it wouldn't be that close to the side panels and cause that noise?
 
A made a couple changes to my CoolJag setup and got pretty good increases in performance. I was finally able to complete a 10 minute (actually did 15) Prime95 run while keeping my temps lower than the C7.

First, was that I realized that the cooler was basically oriented incorrectly in my existing case (the heat pipes were facing down) so I flipped my case over to reflect the correct orientation and to mirror the way it would be sitting in the A4. This caused a very noticeable drop in temps.

Second, I removed the IO area heatsink on the Strix Z270i and put in a TY-100 FAN. This dropped temps about 3C.

CyNvivs.jpg

UjQmqD1.jpg


First comparison is 5 minute stress test with Intel XTU. Second set is 10 minutes of Prime95 blend (I did 15 minutes with the CoolJag though). Ignore the clock speeds in the screenshots. All tests were done @ 4.2Ghz.

Max temps are the thing to look at.

pDkXaO3.png
aKLqC2H.png

sAqQGZY.png
xmDHpGG.png
 
Why is the Cooljag Screenshot when the cpu isnt stressed anymore? So the cooljag is only 6degrees cooler?
 
A made a couple changes to my CoolJag setup and got pretty good increases in performance. I was finally able to complete a 10 minute (actually did 15) Prime95 run while keeping my temps lower than the C7.

First, was that I realized that the cooler was basically oriented incorrectly in my existing case (the heat pipes were facing down) so I flipped my case over to reflect the correct orientation and to mirror the way it would be sitting in the A4. This caused a very noticeable drop in temps.

Second, I removed the IO area heatsink on the Strix Z270i and put in a TY-100 FAN. This dropped temps about 3C.


UjQmqD1.jpg


First comparison is 5 minute stress test with Intel XTU. Second set is 10 minutes of Prime95 blend (I did 15 minutes with the CoolJag though). Ignore the clock speeds in the screenshots. All tests were done @ 4.2Ghz.

Max temps are the thing to look at.

Have you tried flipping the TY-100 fan to blow the air out of the motherboard?
 
So is Thermal Grizzly the new hotness these days? I remember when Arctic Silver was all the rage. I thought I had some lying around when in put my Lp53 on, but used some old Noctua paste instead.
 
EDIT: if you want heatsinks, only buy the small one, the STRIX Z270I uses a non usual power design!


If somebody thinks he wants to remove the heatsink from his STRIX but still wants something there (smaller heatsink) for looks or slightly improved temperatures.
I would recommend something like this:
Thermal conductive adhesive tape

with:
8mm high black heatsink
or
3mm high silver heatsink
or
small 3.65mm high silver heatsink

How to use:
Under the Asus heatsink there will be two MOSFETs per phase (three for iGPU?) with a package size of roughly 5.9mm x 6.9mm. Usually two phases are arranged very closely together.
You can use on of the big heatsinks to cover four MOSFETs at once or you can use four of the small ones to only cover one MOSFET.

The tape above is very nice for this kind of application. Just cut it to the right size and use it to stick the heatsink onto the MOSFETs and you are good to go.

At this point I would like to ask someone who removed a heatsink from their STRIX to make a picture of the phases, that way I can confirm that they are arranged in the usual way.
Also, I would like to say that even these heatsinks are not really needed and you would add them more for the looks.
 
Last edited:
Just got my LP53, zip-tied the TY-100 fan onto it and ......



Idle temps:

LP53 with ty-100 fan.JPG

Prime 95 temps:

LP53 with ty-100 fan stress.JPG

Bloody hell ..... looks like the LP53 + TY-100 is in the lead so far in my tests! This is on my 6700K @4.4Ghz as usual.

I didn't even see it hit 86°C, must have only been for a split second, 99% of the time it was hovering between 81 and 83°C

This is with the 4mm standoffs and no I/O shield. i have used the same thermal paste on all of these tests btw, the Thermal Grizzly Kryonaut, i applied it with the included spreader as per the instructions.

It's silent at idle and pretty quiet at full speed. It's actually a lower pitch when the side panel is on, but sounds more or less the same with the side panel both on and off, which is good. the 4mm standoffs may be helping with this as the fan is further away from the side panel than it would be without the shorter standoffs, therefore creating much less air turbulence.

well guys, guess I got lucky and will be getting a FULL COPPER ultra limited cooler. more details soon

Man i am jealous, seeing how good this tiny little copper cooler performs has me itching for one now lol

I'm wondering if the fins being copper has anything to do with the lower temps, its barely taller than the AR05 and cools much better, and that has copper direct contact heatpipes, the LP53 doesnt have direct contact heatpipes, doesn't look like it even needs them.

I'm hoping Dondan can do a fully copper cooler, the temps show that its far more efficient than aluminium, at least in this super small cooler range.
 
Last edited:
It shouldn't be that hard to go full copper. I remember almost 15 years back thermalright had a full copper xp-90 heatsink. It was a great cooler too.
 
Just got my LP53, zip-tied the TY-100 fan onto it and ......



Idle temps:

View attachment 17693

Prime 95 temps:

View attachment 17694

Bloody hell ..... looks like the LP53 + TY-100 is in the lead so far in my tests! This is on my 6700K @4.4Ghz as usual.

I didn't even see it hit 86°C, must have only been for a split second, 99% of the time it was hovering between 81 and 83°C

This is with the 4mm standoffs and no I/O shield. i have used the same thermal paste on all of these tests btw, the Thermal Grizzly Kryonaut, i applied it with the included spreader as per the instructions.

It's silent at idle and pretty quiet at full speed. It's actually a lower pitch when the side panel is on, but sounds more or less the same with the side panel both on and off, which is good. the 4mm standoffs may be helping with this as the fan is further away from the side panel than it would be without the shorter standoffs, therefore creating much less air turbulence.



Man i am jealous, seeing how good this tiny little copper cooler performs has me itching for one now lol

I'm wondering if the fins being copper has anything to do with the lower temps, its barely taller than the AR05 and cools much better, and that has copper direct contact heatpipes, the LP53 doesnt have direct contact heatpipes, doesn't look like it even needs them.

I'm hoping Dondan can do a fully copper cooler, the temps show that its far more efficient than aluminium, at least in this super small cooler range.

I think a9x14 is a better option than ty-100.The temperature is hotter 2 to 3 degrees Celsius, but it is much quieter.
 
Meccabolix - Apart from Dans comprehensive testing, your testing so far is the most valid enduser comparison tests by being able to test the largest variety of coolers on a fairly constant build, the most comprehensive by far. Keep up the amazing work!! Much appreciated by All of us!

Would be great to see if you could compare the a9x14 with normal standoffs on the LP53 - to represent the most likely scenario that many might end up with. I think quite a few have gone down that pathway, myself included.

Also would be great to compare the TY-100 fan to the Noctua a9-14, both noise and performance wise.
 
Last edited:
At this point I would like to ask someone who removed a heatsink from their STRIX to make a picture of the phases, that way I can confirm that they are arranged in the usual way.
Also, I would like to say that even these heatsinks are not really needed and you would add them more for the looks.

Heatsinks stripped within 30s of removing the z270i from the box :)

I'll bring the FLIR home again this weekend and take some measurements - based on what I saw on my current board, the PCH was the only component generating any heat of note aside from the CPU.
 

Attachments

  • 2017-02-23 21.03.34.jpg
    2017-02-23 21.03.34.jpg
    320.8 KB · Views: 218
  • 2017-02-23 21.04.23.jpg
    2017-02-23 21.04.23.jpg
    344.8 KB · Views: 208
  • 2017-02-23 21.05.44.jpg
    2017-02-23 21.05.44.jpg
    307.2 KB · Views: 211
I like the look of your mobo without those big solid blocks of alum. Its like concept of how heatsinks work is based on how big they look and how many labels or RGB LEDs you can put on one. There have been way too many heatsinks I have seen on mobos where there are fins hidden under a big sticker to block all airflow.

One thing i have wondered with regards to the PCH heatsink on the Z720i Strix, is how much surface area it actually gets for cooling, and whether its so low that it actually ends up heating up the M2 SSD rather than cooling it. Or it might more likely mean that the M2 has a higher idle temp, and when both are under heavy loads, the SSD only runs slightly cooler than the M2 would have alone.
 
repentsinner thanks for posting the shots under the heat sinks. I plan on removing the one by the IO ports so I can fit the nexus cooler with a 120mm fan. Hopefully I can use my ram without modification.

Hopefully this results in some superior cooling.
 
Speaking of all the cooling, what are the group's consensus on delid/relid? I have a 6700k with Cryorig C7 - despite the noise, temperatures are ok. However, lower is always nicer, especially as it results in lower RPMs.
 
Heatsinks stripped within 30s of removing the z270i from the box :)

I'll bring the FLIR home again this weekend and take some measurements - based on what I saw on my current board, the PCH was the only component generating any heat of note aside from the CPU.

Thanks for the pictures. Don't forget to put a bit of electrical tape onto the PCH to correct the emissivity (read for short explanation and good post by FLIR on how to correct it).


About the power design of this motherboard. It is certainly not a usual layout and I don't yet fully understand how exactly it works (I don't see enough MOSFETs for example). I am however pretty sure it is properly cooled and doesn't need a heatsink as long as there is some airflow and the case temperature isn't too high.

usual layout:
vrm.jpg

Asus STRIX Z270I:
power_1.jpg power_2.jpg
 
Speaking of all the cooling, what are the group's consensus on delid/relid? I have a 6700k with Cryorig C7 - despite the noise, temperatures are ok. However, lower is always nicer, especially as it results in lower RPMs.

As far as I have seen from various people who relided their 6th or 7th gen CPU the temperature drop is all over the place. To me it seems like it depends on if you got lucky with the silicon lottery or not.
 
I haven't seen it anywhere in the thread yet but it looks like the AMD Spire 95W cooler that comes with the Ryzen 1700 is 54mm tall so won't fit in the case. Obviously the 1700X and 1800X 140W cooler is larger @ 85mm tall.
http://digiworthy.com/2017/02/20/amd-ryzen-7-wraith-max-and-spire/

Going to be tough to find an ITX board and a cooler for the Ryzen.

Some will be released with the Wraith Stealth cooler (http://imgur.com/a/nTYOb) which seems to be a lot less tall, and doesn't come with the RGB ring, going to try and swap the RGB fan with the Non-RGB one when i get it :)
 
Thanks for the pictures. Don't forget to put a bit of electrical tape onto the PCH to correct the emissivity (read for short explanation and good post by FLIR on how to correct it).


About the power design of this motherboard. It is certainly not a usual layout and I don't yet fully understand how exactly it works (I don't see enough MOSFETs for example). I am however pretty sure it is properly cooled and doesn't need a heatsink as long as there is some airflow and the case temperature isn't too high.

usual layout:
View attachment 17729

Asus STRIX Z270I:
View attachment 17727 View attachment 17728

Good call on the electrical tape - I ran into a similar issue measuring a motor the other day.

The part you called out in question has a reference designator of PDQ1321 and a label 'R10 647 CBL'. And yeah, not sure where the last mosfet is for that inductor, although one of the inductors at the other end has three mosfets close to it?
 
So I delided i7 7700 (non-k). I am using cryorig c7 and there is a 10°C difference with side panel unmounted. When is side panel mounted, temps are higher, but noise from c7 is far worst.
But after 15 minutes of full load (side panel open) on cpu, temps are maxed to 80°C. On iddle, 36-40°C.

Bf1 fullhd maxed out - cpu temp is not going over 65°C (room temp: 21°C)
 
So I delided i7 7700 (non-k). I am using cryorig c7 and there is a 10°C difference with side panel unmounted. When is side panel mounted, temps are higher, but noise from c7 is far worst.
But after 15 minutes of full load (side panel open) on cpu, temps are maxed to 80°C. On iddle, 36-40°C.

Bf1 fullhd maxed out - cpu temp is not going over 65°C (room temp: 21°C)

What paste did you use?
 
OK so this post is going to be a long but very interesting one.

All tests are done on my 6700K @ 4.4Ghz with the LP53 and Noctua NF a-9x14 PWM fan

i unplugged the titan x pascal because sometimes it makes a little rumble when idle, and the NF a9x14 PWM is so quiet you can slightly hear the graphics card in the background, we don't want to hear that, we just want to hear the CPU fan.

i applied fresh thermal paste when i changed orientation of the cooler, the thermal paste i use is Thermal Grizzly Kryonaut.

The first test i did was with the LP53 heatpipes in a horizontal orientation. i used the 4mm standoffs and the a9x14 fan:



Idle temps:

lp53 with a9x14 4mm horizontal idle.JPG

Prime 95 temps:

lp53 with a9x14 4mm horizontal stress.JPG

Next test was the LP53 heatpipes in a horizontal orientation. i used the 7mm standoffs (the original ones) and the a9x14 fan:



Idle temps:

lp53 with a9x14 7mm horizontal idle.JPG

Prime 95 temps:

lp53 with a9x14 7mm horizontal stress.JPG

Next test was the LP53 heatpipes in a vertical orientation. i used the 4mm standoffs and the a9x14 fan:



Idle temps:

lp53 with a9x14 4mm virtical idle.JPG

Prime 95 temps:

lp53 with a9x14 4mm virtical stress.JPG

Next test was the LP53 heatpipes in a vertical orientation. i used the 7mm (the original ones) standoffs and the a9x14 fan:



Idle temps:

lp53 with a9x14 7mm virtical idle.JPG

Prime 95 Temps:

lp53 with a9x14 7mm virtical stress.JPG

All fan speeds og the above tests were:

Idle speed: 800 RPM

Full speed: 2200 RPM

As you can see, those temperatures are very very low, but i don't know how, surely they would be higher with a smaller fan than the TY-100?????

Anyway, as you can tell, all noise tests at idle are silent, but i think the 4mm standoff tests are slightly quieter with the side panel on with max fan speed, theres hardly any difference in it, but the 4mm tests have a slightly lower pitch sound.

For fun i did one last test as i was feeling rather confident with those low temps.

I tested the LP53 heatpipes in a vertical orientation. i used the 4mm standoffs and the a9x14 fan AND used the low noise adaptor that came with the fan:



Idle temps:

lp53 with a9x14 4mm virtical idle low noise.JPG

Prime 95 temps:

lp53 with a9x14 4mm virtical idle low noise stress.JPG

Idle speed: 618 RPM

Full speed: 1611 RPM

Not bad! as you can tell from the video it's silent even at full fan speed! this setup is the one to go for if you want a completely silent Dan pc.

if anyone has any questions feel free to ask.
 
Last edited:
Meccabolix: I wonder what impact the 4mm standoffs will have on the C7 noise level, but especially pitch?

Not interested in temps as there will be only small changes, but the noise might change a lot. Interested as I, and I guess many others have the C7 and are wondering what our next move should be.

Cheers!

I second hoodedhusky..am VERY interested to see noise levels on the c7 with the 4mm standoffs..would you be able to indulge us meccabolix? (Am sure many others would find this useful as well)

I tried to install the C7 with the 4mm standoffs but the nuts than come with the c7 are super thick, and the bolt that goes through the 4 holes in the motherboard around the cpu and very long. i would say the bolts are probably 6 or 7mm long after they come through the holes. i couldnt screw the motherboard in as i didn't want it to flex. i can remove the bolts however, so i'll try to mod it right now and i'll get back to you with a video if i can manage it.
 
Has anyone tried mounting a Scythe KAZE JYU on top of their C7 cooler? It's 100mm x 12mm with 92mm mounting holes. Couldn't find anything thinner.

The XT140 is 13mm (2mm thinner than the C7 fan), but it is just as loud. I'm skeptical that 1 extra mm less will make it any quieter, but if bringing the fan further from the case holes doesn't make it quieter, I don't know what the problem is.
 
OK so this post is going to be a long but very interesting one.

All tests are done on my 6700K @ 4.4Ghz with the LP53 and Noctua NF a-9x14 PWM fan

i unplugged the titan x pascal because sometimes it makes a little rumble when idle, and the NF a9x14 PWM is so quiet you can slightly hear the graphics card in the background, we don't want to hear that, we just want to hear the CPU fan.

i applied fresh thermal paste when i changed orientation of the cooler, the thermal paste i use is Thermal Grizzly Kryonaut.

The first test i did was with the LP53 heatpipes in a horizontal orientation. i used the 4mm standoffs and the a9x14 fan:



Idle temps:

View attachment 17799

Prime 95 temps:

View attachment 17800

Next test was the LP53 heatpipes in a horizontal orientation. i used the 7mm standoffs (the original ones) and the a9x14 fan:



Idle temps:

View attachment 17802

Prime 95 temps:

View attachment 17803

Next test was the LP53 heatpipes in a vertical orientation. i used the 4mm standoffs and the a9x14 fan:



Idle temps:

View attachment 17804

Prime 95 temps:

View attachment 17805

Next test was the LP53 heatpipes in a vertical orientation. i used the 7mm (the original ones) standoffs and the a9x14 fan:



Idle temps:

View attachment 17806

Prime 95 Temps:

View attachment 17807

All fan speeds og the above tests were:

Idle speed: 800 RPM

Full speed: 2200 RPM

As you can see, those temperatures are very very low, but i don't know how, surely they would be higher with a smaller fan than the TY-100?????

Anyway, as you can tell, all noise tests at idle are silent, but i think the 4mm standoff tests are slightly quieter with the side panel on with max fan speed, theres hardly any difference in it, but the 4mm tests have a slightly lower pitch sound.

For fun i did one last test as i was feeling rather confident with those low temps.

I tested the LP53 heatpipes in a vertical orientation. i used the 4mm standoffs and the a9x14 fan AND used the low noise adaptor that came with the fan:



Idle temps:

View attachment 17809

Prime 95 temps:

View attachment 17810

Not bad! as you can tell from the video it's silent even at full fan speed! this setup is the one to go for if you want a completely silent Dan pc.

if anyone has any questions feel free to ask.


Thanks for such extensive testing.

I have noticed that you did not have any I/O shield in-place during these testing, which may increase the temps a little.

I'm impressed with the results you got, in my scenario (Node 202 Case + 4770K (delid) + LP53 Noctua combo ) i top out at 85 C.

Looking at ways to mod the I/O shield to have vents.
 
OK so this post is going to be a long but very interesting one.

All tests are done on my 6700K @ 4.4Ghz with the LP53 and Noctua NF a-9x14 PWM fan

i unplugged the titan x pascal because sometimes it makes a little rumble when idle, and the NF a9x14 PWM is so quiet you can slightly hear the graphics card in the background, we don't want to hear that, we just want to hear the CPU fan.

i applied fresh thermal paste when i changed orientation of the cooler, the thermal paste i use is Thermal Grizzly Kryonaut.

The first test i did was with the LP53 heatpipes in a horizontal orientation. i used the 4mm standoffs and the a9x14 fan:



Idle temps:

View attachment 17799

Prime 95 temps:

View attachment 17800

Next test was the LP53 heatpipes in a horizontal orientation. i used the 7mm standoffs (the original ones) and the a9x14 fan:



Idle temps:

View attachment 17802

Prime 95 temps:

View attachment 17803

Next test was the LP53 heatpipes in a vertical orientation. i used the 4mm standoffs and the a9x14 fan:



Idle temps:

View attachment 17804

Prime 95 temps:

View attachment 17805

Next test was the LP53 heatpipes in a vertical orientation. i used the 7mm (the original ones) standoffs and the a9x14 fan:



Idle temps:

View attachment 17806

Prime 95 Temps:

View attachment 17807

All fan speeds og the above tests were:

Idle speed: 800 RPM

Full speed: 2200 RPM

As you can see, those temperatures are very very low, but i don't know how, surely they would be higher with a smaller fan than the TY-100?????

Anyway, as you can tell, all noise tests at idle are silent, but i think the 4mm standoff tests are slightly quieter with the side panel on with max fan speed, theres hardly any difference in it, but the 4mm tests have a slightly lower pitch sound.

For fun i did one last test as i was feeling rather confident with those low temps.

I tested the LP53 heatpipes in a vertical orientation. i used the 4mm standoffs and the a9x14 fan AND used the low noise adaptor that came with the fan:



Idle temps:

View attachment 17809

Prime 95 temps:

View attachment 17810

Not bad! as you can tell from the video it's silent even at full fan speed! this setup is the one to go for if you want a completely silent Dan pc.

if anyone has any questions feel free to ask.

You are doing god's work here :)

With the 4mm standoffs you should have room for a 25mm thick fan. Do you happen to have a 92x25mm fan?
It would also be interesting how a slim and regular 120mm fan would cool and sound on the LP53. I think you don't have LP RAM so you would need to temporarily remove one instead to fit the fan in.
 
Last edited:
Hold your hype, AMD may work some magic with Vega, like they just did with Ryzen. Rumors say that signature Vega will have 12 TFLOPS, which is more than Titan X. Le'ts just wait and see.

The RX 480 is rated for 5.1 TFLOPS and the GTX 1060 is rated for 4.4 TFLOPS. Yet, generally the 1060 is a tad bit faster.

https://www.techpowerup.com/reviews/MSI/GTX_1060_Armor/30.html
http://gamegpu.com/test-video-cards/opredelenie-naibolee-bystrykh-i-optimalnykh-reshenij-2016-goda

If TFLOPS performance of Vega is extrapolated from this comparison, then it will take 16% more theoretical TFLOP performance than Titan X to equal Titan X speed (12.76 TFLOPS). The rumors peg the 1080 TI to be 11.8 TFLOPS, which looks like Nvidia is going to do to AMD exactly what they did last time - and that is force AMD to either release a card that is always slower than 1080 TI OR force AMD to push the GPU well beyond it's perf/w sweet spot and leave very little OC headroom to be equal at stock vs. stock speeds.
 
I'm waiting for my case but I was wondering why not remove the standoffs all together Meccabolix? or mount them behind so we get negative space, -5mm standoffs lol?
 
Back
Top