Sunday, February 27, 2022

Appeared as a panelist on ABC Download This Show

This week I was invited to be a guest on Marc Fennell's Download This Show. Marc must be the busiest person in media but he still has time to put together a very professional radio/podcast show every week.

In the old (pre-COVID) days we used to go to ABC studios but these days we're well practiced in recording from home. Even Marc was at his home.

My internet here at Drummond is NBN fixed wireless and unfortunately upload speed isn't reliable enough for my taste so I travelled to Melbourne for the recording.

If you're interested you can get more info and listen here.

Wednesday, February 23, 2022

QJ-PS30SWI low noise switching power supply with "noise offset" feature

In my pursuit of low receive noise I've been using a linear power supply but the one I have makes a loud audio hum sound that is a bit annoying. Searching for low noise switching power supplies I stumbled across the QJ-PS30SWI which I purchased for AU$120 from AliExpress.

The supply is quite low noise, but there are lines visible on the waterfall about every 2kHz if I turn the preamp on.

Unlike switching supplies I've used with HF receivers in the past, where there are broad bands of noise that drift up and down, in this case they are narrow and the "noise offset" control lets me move them away from the frequency I'm listening to.

The supply is very compact and has a fan that doesn't seem to run, presumably it does when it gets hot.

Thursday, February 17, 2022

Command line WSPR receive with weakmon

Work to reduce noise here at Drummond has been going well and,  yesterday during a voice contact with WSPR reception leader, Phil VK7JJ, he alerted me to the fact that I was on top of the leaderboard by distance. By a fair margin too.

My iOS app, WSPR Watch, displays WSPR spots pulled from and features a very simple transmit capability. People often ask if I could make it receive WSPR as well.

The obvious approach would be to port the decoding code from wsjtx but that source base is pretty daunting and includes several languages including Fortran. 

I went looking for other implementations and stumbled across rtmrtmrtmrtm's weakmon which is an almost pure python implementation (there's a few bits of c). The project hasn't been updated for some time and I ran in to a few issues running under python 3 on Ubuntu linux.

Weakmon directly reads audio from an input (including things like the IC-7300 I'm using) and can use serial CAT control to hop bands so it's very functional.

I found the most recent fork by kholia and forked that to create my version here.

My focus is on WSPR and running under python3. Other modes may not be working.

Changes I've made

  • Updated the to explain what my fork is about
  • Running under Python 3.9.7 I struck a number of exceptions during decode all complaining about things that should be integers but instead were floats. I've fixed these as I've seen them but there may be more. (All in
  • Phil, VK7JJ, pointed out that transmit maidenhead locators were incorrect. I've included a fix by Ross, EX0AA.
Since moving out of the city and receiving WSPR I've had some wonderful emails from people astonished that I was able to pick up their 200mW beacons from 16,000km away. They ask what magic I'm using - the answer is a quiet location and hunting down every bit of noise.

Friday, February 04, 2022

Set up OpenWebRX at Drummond

My low noise reception seems worth sharing with a few friends so I set up OpenWebRX. I initially tried on an Ubuntu 21.10 laptop but for some reason couldn't get the AirSpy HF+ to work with it even though SoapySDR said all was well. 

In the end I gave up and tried the easy option with was a pre-made image for Raspberry Pi. This worked well right out of the box.

This is a very low noise location but do you see that big band of interference down the left side? It seems to be from the Raspberry Pi or it's USB power supply. Very distressing. The lowest noise USB power supply I have is a battery pack with USB output. I've ordered a metal case for the Pi and some clip on ferrite chokes for cables and will be interested to see if things can be improved.

Despite my reservations, the 31m shortwave broadcast band comes alive in the evening here:

I first set this up on a Pi 3B which worked but struggled a bit. The Pi 4 is a very capable computer and handles the load with ease even with some background decoding of WSPR and FT8. 

To answer Paul, and other's question - no it's not publicly listed as my IP isn't fixed and it will be frequently off line.

Great work by the OpenWebRX developers!