635ml (21.5oz) Ryzen 2400G Mini-STX Build

[chinevo]

I've returned. My previous 0.59L KabyLake-G build failed. But Asrock has released an AM4 Mini-STX Motherboard, and now is the time to get back to the idea of a case with a volume of less than 1 liter. Today I updated config in the first post and ordered Zalman CNPS2X cooler with AM4 Kit. To be continued.

 

[Windfall]

I've returned. My previous 0.59L KabyLake-G build failed. But Asrock has released an AM4 Mini-STX Motherboard, and now is the time to get back to the idea of a case with a volume of less than 1 liter. Today I updated config in the first post and ordered Zalman CNPS2X cooler with AM4 Kit. To be continued.

I'll be waiting!

 

[chinevo]

Looking for AM4 cooler with height 27mm and less. I found these:
1. Zalman CNPS2X + AM4 Kit (27mm)
2. Dynatron A18 and clones (27.43mm with stock 15mm blower and 22.43mm with 70x10 fan)

Thanks in advance

Dynatron A18 with 7010 opens up possibility to build a 0.89L case, cause in this build memory will be the highest component.

 

[ondert]

In any case, I would go with the Zalman. Dynatron will surely work much noisier.

 

[chinevo]

In any case, I would go with the Zalman. Dynatron will surely work much noisier.

Me too, but I will try Dynatron with other fan and share performance/noise comparation.

 

[Valantar]

Me too, but I will try Dynatron with other fan and share performance/noise comparation.

Considering just how small the fin stack in that dynatron cooler is (and the fact that it comes with a 6000rpm fan stock) I'm thinking it definitely needs a powerful fan to keep up with a 95W load. You'll no doubt be able to find a quieter fan (or just slow down the stock one, it goes at least as low as 20%PWM/1100rpm, which shouldn't be too noisy) but you'd need to maintain airflow numberd for it to perform at spec.

Noise:
At Duty Cycle 20%: 19.1 dBA
At Duty Cycle 50%: 34.2 dBA
At Duty Cycle 100%: 53.8 dBA

 

[Vlad502]

Dynatron T318 passive with nf-a4x10 case fan(s)?
Maybe not good idea with fins position and STX board front/back i/o ports

Modded to fit AM4: https://post.smzdm.com/p/a07me43w/p2/

 

[chinevo]

Dynatron T318 passive with nf-a4x10 case fan(s)?
Maybe not good idea with fins position and STX board front/back i/o ports
Modded to fit AM4: https://post.smzdm.com/p/a07me43w/p2/

It’s not an option, cause there is no enough space for this solution. I try to find something less than 24mm.

 

[boingk]

Very cool build, be awesome to see you pull it off!

Check out TechYESCity on youtube, he's tested the Dynatron and found a few issues, mainly with mounting it and also its lack of cooling with the stock fan. The stock fan unit seems to be a server-oriented one, designed to be used for tray systems where there is another unit immediately overhead. Using even a standard low-profile fan would improve cooling immensely.

I'm looking to do a similar build, let us know how you go with yours!

- boingk

 

[chinevo]

Very cool build, be awesome to see you pull it off!

Check out TechYESCity on youtube, he's tested the Dynatron and found a few issues, mainly with mounting it and also its lack of cooling with the stock fan. The stock fan unit seems to be a server-oriented one, designed to be used for tray systems where there is another unit immediately overhead. Using even a standard low-profile fan would improve cooling immensely.

I'm looking to do a similar build, let us know how you go with yours!

- boingk

Many thanks for the link to TechYESCity. Before buying and trying Dynatron, I have to make sure that the Asrock mobo doesn’t have elements higher than 8 mm around the socket. And yes I agree that stock cooler very inefficient and needs to be changed.

 

[Valantar]

Replacing the Dynatron's radial fan with an axial one ought to improve its noise/performance ratio dramatically (the only strength of blower fans is compactness and static pressure, after all), but given the thinness of the heatsink (12.5mm including the cold plate, if I'm reading the spec sheet right) you'd likely need a fan with some decent static pressure still - blowing down into and out the sides of such a small cooler will likely be quite restrictive. Looking forward to seeing your results

 

[chinevo]

Great update for the first post. I decided to develop 2 versions of my build. The first version is crazy, enthusiastic, with world record 0.63L volume. The second version is 0.95L PC for using on a daily basis. You can find details in the first post.

 

[Valantar]

Great update for the first post. I decided to develop 2 versions of my build. The first version is crazy, enthusiastic, with world record 0.63L volume. The second version is 0.95L PC for using on a daily basis. You can find details in the first post.

That's pretty cool! A question: it doesn't look like the cold plate of the GPU cooler will align with the CPU IHS if you align the cooler with the motherboard edges. Any plans to rectify this somehow?

Other than that, looking forward to your builds

 

[Vlad502]

Mockup with two orientation cooler mounting

Spoiler: Mockup

8600GTS 53*53 mounting
9800GT 43*43 mounting

Source images:
https://static.techspot.com/articles-info/1816/images/2019-03-27-image-2.jpg
https://www.ixbt.com/video3/images/g84/msi-8600gts-cooler1.jpg
https://images-na.ssl-images-amazon.com/images/I/81bo07xmROL._SL1500_.jpg

 

[chinevo]

Mockup with two directions cooler mounting

Wow! Thanks a lot man! I added your mockup in the first post if you don’t mind.

 

[chinevo]

Interesting find. Zotac 9800 GT Eco.
65W TDP, 76C max temp
12mm thickness

 

[Vlad502]

Videocard is without 6-pin power connector (so it is under 75W) has power consumption ~65W

Here 120W card: ASL G1603 D5 SS (G1606 D5 SS) probably cooled by vapor chamber (G1603 $237 on 1688.com)

 

[Valantar]

Vlad makes a good point above. Might the Eco designation mean this is a low-power variant?
Also, considering that GPUs are far easier to cool than CPUs (better heat distribution across the die/few-to-no hotspots. and direct die contact for the cooler), this could be problematic. Have you considered delidding the APU and running it bare-die for optimal thermal transfer?

Still, I'm curious as to how difficult it would be to ... "semi-DIY" a cooler for this. Vapor chambers can be bought, as can pre-made fin stacks, fans, and thermal adhesive. So would it be possible to get a suitable vapor chamber (as large as can fit), glue it to a thin heatsink, and rig up a 3D-printed fan mount or similar?

 

[chinevo]

Videocard is without 6-pin power connector (so it is under 75W) has power consumption ~65W

Yes, you are right. 65W, not 120W.

Here 120W card: ASL G1603 D5 SS (G1606 D5 SS) probably cooled by vapor chamber (G1603 $237 on 1688.com)

I can buy Zotac 9800 GT AMP for $5, cooler looks very similar to ASL G1603

Have you considered delidding the APU and running it bare-die for optimal thermal transfer?

Great idea, thanks a lot. After watching this video, I will do this without any doubts.

Still, I'm curious as to how difficult it would be to ... "semi-DIY" a cooler for this. Vapor chambers can be bought, as can pre-made fin stacks, fans, and thermal adhesive. So would it be possible to get a suitable vapor chamber (as large as can fit), glue it to a thin heatsink, and rig up a 3D-printed fan mount or similar?

Could you send some links?

 

[OursonSucre]

It's a bad idea to try cool 2200G\2400G with these coolers. The point is that there are throttling point of GPU at ~62C. I had 2200G with Noctua NH-L12S and it was very difficult to hold temps below 62 during CS:GO casual playing in FullHD. Of course delid probably helps but not sure you will achieve anything "working". How many FPS do you expect with downclocked CPU\GPU?
Anyway I wish you good luck.

p.s. info about 3600G is fake. AMD doens't even start producing of these chips. I bet on 4\8 and 6\12 chips (of course there will be low-end athlons with 2\4) and I pretty doubt they hit 5.0 ghz ^^