NUC 8I3BEH - power consumption

Support
1

Hello,

I do hope this is the right place to place my question…
to the point: I would like to power my NUC via regulated linear power supply, the Sigma11, and I need to know what is the power consumpion, Volts and Ampere, average and peak

I’ve seen here in the forum measured values but about the previous generation, the 7th, and the 8th declare a TPD roughly double (28W vs. 15W) (https://nucblog.net/2018/10/coffee-lake-nuc8i3beh-review/3/…

the point is that I need to know the real value, not that in the power brick coming with the NUC, to calculate the power that has to be dissipated by the heatsink… quite a crucial point

is there anybody that could help?
thks
Paolo

3

You can find real world figures here:

However, they dont take into account the possibly higher requirements for the boot process, and if you’re powering portable USB drives or other USB equipment etc.
I can tell you that an sBooster 15-16.5v CANNOT power the NUC8i3 when the Ultra adapter is involved. (Without the Ultra seems to work, but can be a bit unstable)

4

yes, on the sBooster website is clearly specified that as NUC8 increase the power consumption the sBooster is NOT MORE compatible having max 36W output…

5

The web page you listed achieved a 52.4W test result.

So you need no less than 3A @ 19V.

6

yep, but that is under stress test… I was wondering if the average power consumption was lower as, in that web site, the 52,4W are under stress test involving probably at 100% all the cores… I don’t know how heavy is Rock when running without upsampling/converting to DSD (but with DSP active)… sBooster declare in their web that their product is not more compatible because of the 8th NUC generation request of more then 36W when turn ON… ok, that would not be a problem with a linear psu able to have a decent peak current output

on the other hand a possible more simple solution probably could be a filter, LC, in between the SMPS supplied with NUC and NUC itself… this could allow to keep the favorable ratio power/dimension of the SMPS and, at the same time, reduce the pollution that from that cheap (very cheap!) unit go into th NUC

ok, inside the NUC thera are a lot of DC-DC converter… imho this is not a reason to allow extra-electrical noise contamination as this not only can influence in some way the NUC operativity but also migrate trough the NUC into → cable (irradiating around if not shielded and at HF) → DAC disturbing the conversion accuracy

7

TDP is 28W according to the Intel Specs

But best bet is to test yourself using a simple plug in power meter or something more advanced.

8

From an engineering perspective the worst case behavior needs to be supported.

If this were a self driving automobile you don’t want the computer to crash.

9

you are right… the point is that I would avoid to buy another tester, I’m full of tester (air temperature, walls humidity, water pressure, and so on) I had to buy to check the most strange things that “professionals” nowadays use to work in quite bad way and to not test anything (more safe/confortable)

in the mean time I’ve ordered almost all I need to build my lpsu and you know what is funny? the connector giving power to the NUC, I mean that one on the smps from factory, should be able to support nominal 2Amp… at least this is the maximum value for similar connector you can buy to assemble your own power cable… so, the 4,74Amp on the smps label are, maybe, peak power but can’t be average value as the connector would not survive