12gen N-series Nas motherboard (topton, cwwk, ... )
- Thread starter Steffe
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Just a large goop of thermal paste.
I did try putting a small ball of thermal paste on the die (like you would a desktop cpu) and then put the head speader on and it nearly touched, not enough to spread the paste out.
So you need a lot of thermal paste to make contact between the two. But I have now put a 0.5mm thermal pad between the both and they now make contact and squish the thermal pad slightly. Great!
I will run some thermal tests as soon as I can.
Right now I am contemplating whether to use the included jonsbo HP400S cooler or not as it is VERY close the the capacitors when mounted in a specific orientation. The other 3 orientations actually make contact with components on the board...
If you get the Arctic TP-3 0.5MM Thermal Pad you can stack them according to their docs.
So try 0.5mm, if it's too small add another one to make it 1mm etc. I knew 0.5mm was enough as the dies left an imprint in the pad during a test fit.
also curious about thermal tests. Idle and max if you can. Did not change the paste and I was thinking the temps were a little hight
Last edited:
Good luck finding an ASM2812 that is short enough to fit in this motherboard with a x4 PCIe .
I mean if you find one, please link haha
I changed the thermal paste to a pad at the begining so i don't have thermals with the paste. I build my system in an old N40L case, not best option.
I need to use the stock heatsink without fan because i don't have enough space. The only fan is the 12 cm case fan and the PSU fan.
In idle with thermal pad about 48-50C in peek load it reach the thermal limit.
Power usage with a Samsung 980 nvme boot drive, 2x3.5 4TB purple, 2x3.5 1TB Toshiba and Samsung connected to a 9207 HBA and 3x2.5 500Gb HDD connected to the sata ports.
The N305 board use in idle around ~50W at boot ~70W
I running proxmox with 2 wm's and 8 container.
I don't stink the thermal paste will works in a long term, because when it dries up it's just get worse or not transfer any heat to the heatsink.
I need to use the stock heatsink without fan because i don't have enough space. The only fan is the 12 cm case fan and the PSU fan.
In idle with thermal pad about 48-50C in peek load it reach the thermal limit.
Power usage with a Samsung 980 nvme boot drive, 2x3.5 4TB purple, 2x3.5 1TB Toshiba and Samsung connected to a 9207 HBA and 3x2.5 500Gb HDD connected to the sata ports.
The N305 board use in idle around ~50W at boot ~70W
I running proxmox with 2 wm's and 8 container.
I don't stink the thermal paste will works in a long term, because when it dries up it's just get worse or not transfer any heat to the heatsink.
Attachments
Last edited:
Interesting. Though I doubt about boot time power consumption. 7 hdds spin up should be alone close to 100w, for a few seconds.
Well, that depends if the card **requires** all 8 pcie lanes as looking at the layout of some of these cards the lanes map directly to the m.2 slots. So if you plugged it in a x4 slot, there is potential only 1x m.2 card will be detected.
Don't do this, you have ZERO knowledge on if 0.5mm is the correct amount you need to reduce the gap by, it could be different for every board sold due to manufacturing inconsistency/tolerance.
If 0.5mm is too much, you'll crack the die with the pressure.
Don't do this either, same reason as above. These are not designed to mount directly to the CPU dies and you should only do this if you know the EXACT measurements.
Dies are very fragile. The cost of the motherboard is irrelevant to whether a CPU die will crack under too much pressure.
With either, if you apply to much pressure you risk damaging the die.
We will have to agree to disagree here, but I had to make this point so others do not blindly follow your information without first knowing the potential damaging risks.
Yes you are right. The actual power consumption on boot maybe higher. I use a smart meter which doesn't gave me continous measurements.
Thanks for the info. Later i will check the exact distance between the die and the heatspreader.
If i can replace it with a 0.5 spacer or another cooler i will mod it.
I agree the thermal paste with spacer has better heat transfer then a thermalpad, but it needs accurate measurements.
When i build the system i only want a quick fix, so the thermal pad was a good option.
Excellent. I wondered how they did it with no visible bolts on top.
Do the standoffs only sit on top of the motherboard, or do they set into the mounting holes? If just on the surface then it would be pretty easy to file a small amount off the end. The hardest bit would be keeping them even so the plate stays level.
Pleased to hear that the pad is a good fit.
There are other discussions about the tricky fit of the Jonsbo cooler, though I recall someone finding that a number of alternative coolers were actually tighter still, or wouldn't fit.
The Noctua NH-L9i fits perfectly. The gap is close to the caps still but not as close as the jonsbo cooler was.
I am now having an issue with the motherboard screws that came with the motherboard. They seem to be a different thread to the standoffs in the case.
Furthermore the standard motherboard screws are too big for the holes in the motherboard, but the provided screws fit through the holes.
So now I have to somehow find some compatible standoffs...
Ah, that would make it harder to reshape the fit. I saw your other post and that a simple pad closed the gap. Good news especially if the temperatures improve.
There was another insidence of this, a collision between the two standards. 6-32 vs M3
As I recall the case end of the standoffs is always 6-32 but the motherboard end can be either thread. It's more odd if the MB included M3 but they're too skinny for the MB holes.
Usually they'd be like Computer Screws M3 x 1/4in Long Standoff - Fasteners & Brackets | Computer Parts | StarTech.com United Kingdom which are pan head with an additional flange.
Last edited: