FT8 dynamic range

Submitted by Thomas Nilsson on

"Interesting" short commings

To be updated!

There are some problems that arrise when running a narrow band mode as FT8 with a bandwidth of 50 Hz with a wider filter, like a normal SSB filter, with a bandwidth of 2500 Hz. The exact behaviour depends on the design of your rig. For me it is an IC-7100.

AGC activated in a "normal" way

As seen on the picture the even 15 s slots are less "noisy". The reason for this is that the transmitted signal are very strong, FT8 reports up to 32 dB in my case. This pushes the AGC to reduce the gain. The signal I want to receive are something like -20 dB. the dynamic here is in other words something like 50 dB. This makes it impossible for me to detect anything interesting when the extremely strong transmitters are on. I see at least the following ways of fixing this:

Running with a more narrow filter

A one way would be to run with a more narrow filter during an ongoing QSO. For this to work some support would be needed in the program. However with a solution like this you would loose the overview of the band. So this I don't consider as a possible way forward.

Or with a notch

A notch on the strongest station

If there is mostly one interfering station also a notch will work great as seen on the picture. In this case the AGC is still used, so the notch is taken into account for the AGC level. If you have more than one interfering station this will not work either. I did also do some tests with the automagic noise cancelling of the IC-7100. With this I received basically the same result.

Running with AGC off

During this test I changed the "Slow AGC" mode of the IC-7100 into being defined as AGC off. This can be done in the AGC setup of the AGC for the IC-7100.

AGC switched off

As can be seen this causes spurious at the right part of the detected AF band. As they are twice the detected signal they are probably caused by clipping in the baseband due to an extremely strong signal. It is important that the levels are carefully adjusted, which they are NOT here. The noise is still disappearing which is also an indication of that the levels are not 100% correct yet. In this case the strong signals are indicated as +30 dB and the lowest detectable signal for me are something like -23 dB. This gives a total dynamic range of 53 dB. The dynamic range is there on my rig, both for short-way and for VHF / UHF.

The RF gain setting is critical and need to be adjusted when I switch band but in general it works very well. The important thing is to put the noise at the correct level and then leave it there for each band.


It should be mentioned that I use a omnidirectional antenna, a halo for 2m, which make the situation much worse for me as I can't turn the antenna in a specific direction to lower the interfering signal. All measurements were done using the SignaLink USB modem (sound card). But the soundcard of the IC-7100 uses almost the same chip, the PCM2901 for the IC-7100 and the PCM2906 for the SignaLink converter.