Further Experiments with Computer Morse Decoders

Having tested XDeMorse with machine simulated Morse Code, I wanted to test it on air. It runs on the small Acer Aspire One, so I wanted to test it with a similarly small radio, the MTR3B. The MTR3B has a fixed internal filter which may not help in crowded band conditions, but works well with a good signal to copy.

The antenna for this test was a 10m wire running up the 12m fiberglass mast. This is an urban location, so noisy. The MTR3B’s volume levelling circuit presents any amount of noise at the same level. What is needed is an incoming signal which is louder than it and quietens it.


XDeMorse copying Morse off air

Different stations are visible in the text here as I tuned around to find them. I am impressed. It took a little fiddling at first to set the threshold. It also found itself stuck at 10wpm so needed a hint to try faster. It soon locked on to the amateur signals around 25 to 30wpm.

We can see from the waterfall display how the signal quietens the noise when received, and how narrow the MTR3B’s filter is. XDeMorse does not provide a sound output, so a Y cable was used to feed a small powered speaker.

I could imagine taking this to next week’s Scout Camp as a demo, especially given the need not to carry too much kit, and the use of PMR446 walkie talkies for our main radio themed activity.

CW Skimmer

CW Skimmer copying from the MTR3B

I’m note sure that CW Skimmer performed any better. It is designed to deal with multiple signals, which I expect could help in a situation with more than one station audible in the MTR3B’s pass band. It seems to be seeing two copies of the same signal here, maybe harmonics due to the diode limiting in the radio? A nice feature of CW Skimmer is if it is unsure of a character it will go back and correct if, for example, it sees the same callsign repeated.

CW Skimmer provides an audio monitor feature, useful given that I’d had to disconnect the small speaker to plug the second laptop in. It sounds very distorted. I don’t know if that is a weakness in the laptop speakers of in CW Skimmer’s digital processing.

2 thoughts on “Further Experiments with Computer Morse Decoders

    1. It’s one for me to try now I’ve got the Raspberry Pi up and working.
      I’ve been working hard on CWOps Level 2 since this post, thinking I’m getting there at manual decoding. A week or so off and I wonder if that’s still the case.

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Your email address will not be published.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.