Enhancement request: Noise squelch for FM
This probably only applies to the CB and 10m bands, but it would be nice if there was a noise squelch option since it gives much better separation between an empty channel and a readable signal for NBFM and is what walkie-talkies and most CB radios use. Often CB radios provide noise and RSSI squelch as options, on the CRT SS 9900 the squelch is turned left for noise squelch and right to switch to RSSI squelch.
A noise squelch works by triggering on the audio level via a high-pass filter that lets through frequencies from slightly above the highest modulated audio frequency, which will probably be about 3KHz and up. The audio is obviously sampled before any de-emphasis filter.
There may well be a cleverer way to discriminate between a quieting signal and noise in DSP but I am only familiar with the analogue way of doing this.
I find that the RSSI squelch is very fiddly to adjust for NBFM reception and FM is where the squelch is needed the most because of the harsh channel noise.
Comments
Our current NBFM squelch code comes from the CuteSDR project: https://github.com/jks-prv/Beagle_SDR_GPS/blob/master/rx/CuteSDR/squelch.cpp
I believe it does so by measuring noise power above the filtered voice frequencies. But I am not an expert. If you know of other open source implementations worth looking at please let me know.
Hi JKS,
Yes, that is the type of squelch I am referring to. It's fantastic if it is already implemented. Does the squelch slider change mode on FM? Is there anything else I need to do to activate it? The squelch slider appears to behave as an RSSI squelch even on FM. Am I missing something?
Update: I did some tests on blank carriers and it does indeed seem to behave as a noise squelch on FM. However listening to actual CB conversations it does seem to chatter a bit. It may be a bit slow to open/close. I will have to investigate further.
Thanks
The noise squelch is enabled automatically for NBFM modes, but as you have noticed the hysteresis is not brilliant on some types of signals.
CB is also bad because you get so many co-channel signals, which sometimes cause heterodynes that 'fool' the Squelch into believing that noise is present, when it is not.
Regards,
Martin
Hi Martin,
I was using it to listen to the UK-only channels just below 28MHz where there is a lot less sporadic-E interference from Europe, which tends to mainly in the "Mid" and "High" CB bands (up to CH 26, UK band).
So the signals were perfectly in the clear with no same-channel or co-channel interference. The conversations on the UK band are also very sparse where I am, it is not a crowded band any more.
I did some test at close range on 10m FM and found similar results. The squelch is very finicky and doesn't distinguish well between noise and a quieting, readable signal without chattering.
I know that in the analogue world there are noise squelches that would perform a lot better than what I was hearing. I should be possible to do digitally, but obviously squelch.cpp isn't quite there yet.