I'm attempting to use a NUC7i7 in conjunction with some sensitive analog circuitry. When both are powered from a common 19V DC supply there is excessive noise being picked up in the analog circuit. To eliminate the possibility that the noise is being conducted back through the power connection, I'd like to insert an LC filter at the NUC's DC input. However, tuning the filter requires some basic knowledge of the NUC's internal SMPS (switch mode power supply). I understand that you don't publish full schematics, but would it be possible to get the input capacitance and the switching frequency of the SMPS so that I can properly design the filter? I assume there may be multiple switching regulators on the NUC, so there may be multiple frequencies.
Any feedback is appreciated.
I asked one of the design engineers and here is his response:
I don't think the basic information the customer is asking for will help. The primary source of noise is likely the current transients from the loads, not the switching frequency of the regulators. The current transients from the loads, CPU mainly, can be very large, many tens of amps, and are not necessarily periodic but more likely software workload dependent. I suggest that the customer analyze the noise from his specific setup and software with an oscilloscope to find the actual noise signature.
Hope this helps,
Thanks for quick reply.
With all due respect to the engineer, I think he overlooked a key part of my question ... "When both are powered from a common 19V DC supply". With a switching power supply system there are 2 distinct power delivery systems (PDS) at work. He is referring to the post-regulated, load side PDS transients. Those load transients, in turn, introduce transients into the supply side PDS. While the load side transients are often uncorrelated, the supply side transients are absolutely related to the switching frequency of the regulator(s).
I certainly intend to instrument the design to determine the specifics of my scenario. However, I work remotely and won't be in the same room with the system until late next week. I was hoping to get a head start and the info I requested would allow me to recommend a 3 component filter that could be assembled with parts they probably have in the lab. Please reconsider.