How Quiet ?

I am redoing all my Kiwis and antennas before the snow flies. I'm looking at grounding, CM etc. So I wonder if you have your Kiwi with no antenna attached, the input terminated with 50 or 75 ohms, what noise level do you see DC-30 MHz on the spectrum display? If you see discrete carriers, you have a problem. If the noise is flat across that whole range, probably OK. But what level do you read?

Comments

  • Jim,
    Using python3 kiwirecorder.py -s 10.0.1.71 -p 8073 -f **** m iq -L -5000 -H 5000 --s-meter 100
    and a bash script that sweeps **** from 10kHz to 30MHz and subtracting 40dB to get to noise level in dBm in 1Hz results in the graphs below for two KiwiSDRs when running on a LiPo 5v output battery pack.
    73
    Gwyn G3ZILimage

    Attachments:
    https://forum.kiwisdr.com/uploads/Uploader/ad/1d2ce3f6206af29ed11b51fc2ff842.png
    https://forum.kiwisdr.com/uploads/Uploader/c4/4b3994f80ffeb2156d265b8798b59f.png
  • what does the [SPEC] display show on a the user webpage?
  • Be aware that if the input is not terminated, a large signal anywhere in the spectrum can exercise the top bit or two of the ADC and will typically increase the numbers shown a couple of dB or more due to ADC imperfections.

    The spectrum display is perhaps not the best way to measure noise since it is an entirely different path with different limitations. I suggest using a known bandwidth, such as 10 kHz wide AM, making sure that it doesn't contain any coherent signals, reading the S-meter and then subtracting 40 dB to account for the bandwidth, as Gwyn describes. I think the spectrum display will give different answers depending upon zoom level and the corresponding bin bandwidth. I'm also not sure that it returns RMS noise as does the communications channel, though it may.
  • I'm right there with Glenn as regards measuring noise - hence the graph I shared. But of course, there are times when the spectrum is a useful, quick look. So just taken the attached screenshot. Note that this is with a USB WiFi dongle and not Ethernet. 50 ohm termination. The change in the waterfall was me adjusting the limits. You'll see the spot S meter at -125dBm in AM mode and bandwidth at 1000Hz, so about -155dBm in 1Hz or 3dB higher than the graph above. But it's a spot value, and I've a number of LED and CFL lights on this evening.image

    Attachments:
    https://forum.kiwisdr.com/uploads/Uploader/97/90ff45778a2c9c5c8de4fe7983ae7b.jpg
  • Yes that's the way to do a quick check which is all what many people with do.
Sign In or Register to comment.