KiwiSDR Calibrated Noise Charts

On 2 of my 3 KiwiSDRs I have been running the wsprdaemon script to decode WSPR 24/7 on 14 bands from 2200-10m. The latest release (2.5a) is able to perform a receiver noise reading while collecting the 2 minute WSPR signal sample.

It produces calibrated noise charts for each band with the parameters of noise level in dBm in 1Hz and Temperature in K. Example below:


As far as the interpretation of the graphs is concerned, it’s a learning curve. Diurnal background noise level changes are a good sign for an effective receiving system. I noted on a previous post that I thought that my 30m WSPR reception was affected by vDSL – a look at the 30m chart shows no diurnal variation and has the highest dBm figure.

There is a website which has some other receivers results online, and this will prove very useful when more are added:

6 Replies to “KiwiSDR Calibrated Noise Charts”

  1. I use your WSPR results as a benchmark for the UK, I expect to get nearly the same results except a couple of bands, one where you have local QRM, and those where I do.
    If my results look slow then I check to see if you are in the normal range, if you are slow too I get on with something else.
    I recently tried adding a CB antenna (solaracon 99) to improve my higher HF results and it worked to some extent (17-10m) but adds back in the electrical field noise I struggle with (Big peak at 25Mhz). Swapped it for newer (Gainmaster) and that is useless except on 10m, even stronger 25MHz noise.

    If you could suppress that 30m noise issue I think it would make a big difference as the rest is very lively.
    I did consider trying a phased noise reduction filter here as my main sources of QRM are from the same two locations so should be able to tune for source and frequency of most benefit, not sure how wide the noise affect would be (mainly could it cover two bands like 21&24MHz).
    I have an older Wellbrook loop but here the extra gain just highlights the serious mains born noise. The current active antenna I have tried running under voltage and that seems to help or at least not be detrimental, lowest voltage on paper is 12V at the antenna, I was running at 9.5V at the feedpoint for a few days then 11.1V until this morning when I’ve gone to 12.1V at the feed.
    The noise plots have helped as I can see when a setting flatlines, before it was mainly comparing waterfalls or results.
    Never thought I’d get so interested in subject line WSPR but then I have given up on HF so many times due to QRM never thought I’d bother with HAM at all.

    1. Stu,

      Thank you for your very interesting and welcome comments. I’m currently on holiday so I will give a full response when I return as I wish to discuss the various issues you have raised in more depth.

      73 Geoff

    2. Hi Stu,

      I had hoped to provide some further insights from the noise charts, but these stopped working after a re-installation and I have not yet found the problem. I am working on this, with a goal of having them available in real-time on the website.

      I was interested to hear of your experience of noise around 25 MHz. When trying to setup a HF receiver at the caravan, I am receiving strong noise around this are which makes reception of the higher HF bands useless. At home, I have no noise in this area at all. Driving past the caravan area in a car with the radio on 198 kHz LW, interference broke through – confirming there is a strong noise source.

      I have a Wimo QRM Eliminator which I have owned for some years. The drawback I find is that when you gain something on one frequency, you lose something somewhere else. However, in the light of your contribution, I will try to add it inline with a 30m only dedicated receiver so I can optimise for that band only. It might be a way forward.

      I have not considered running the loop under voltage, but have tried various power supplies – some are better than others. I will do some further tests and report.



  2. Hello Geoff
    I was very pleased to read that you’ve found the noise extension to wsprdaemon useful. I worked on the two noise analysis methods with Rob, writing the code, which drew on software I’d written for an adaptive noise cancellation system I’ve been working on for a while. It helps to a fair extent, given the incessant problems here in suburban Southampton, at least on 15, 40, and 60 simultaneously.

    Gwyn G3ZIL

    1. Hello Gwyn,

      Thank you for your comment, and the work you’ve done with Rob on the noise analysis.

      I am only beginning to try and use the charts in a practical way but I can already see they will be very useful in assessing how changes made in the local environment, such as eliminating noise sources, affect the noise floor.

  3. Hello Geoff
    It looks as if we need to do something with the y axis scales and or to better explain the use of the user-set offsets in the wsprdaemon conf file to account for gain (or loss) ahead of the KiwiSDR’s SMA socket. We are sure of the calibration back to that point from the digital data, but of course almost every installation will be different in some way ahead of the antenna socket. And without knowing circumstances one can’t judge whether a high reading is due to noise or to pre-Kiwi gain.
    Gwyn G3ZIL

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.