Asymmetrical Filter Use on FM Broadcast

💥 A general forum for topics not directly linked to SDR.
RadioSDR1
Posts: 55
Joined: Sun Jun 26, 2022 4:41 pm

Asymmetrical Filter Use on FM Broadcast

#1

Unread post by RadioSDR1 »

In the US, digital radio is accomplished by broadcasting the data on the upper portion of the lower adjacent channel and also on the lower portion of the higher adjacent channel. This poses a problem where if a station is present on either adjacent channel, there is interference from the digital signal.

With AM Broadcast, if there is interference from an adjacent channel, it's easy to filter out one of the side bands. It's also easy on FM Broadcast but the problem is that it causes heavy distortion.

Is there any technique available to eliminate either the upper or lower portion of an FM signal such that there is no distortion? Or is this just the way FM is and thus impossible to do?

jdow
Posts: 785
Joined: Mon Aug 10, 2020 8:17 pm

Re: Asymmetrical Filter Use on FM Broadcast

#2

Unread post by jdow »

I would suggest that Youssef's SDRSharp might be an easier platform for taking your question to the next level. One trick that comes to mind and has not been shoved through any practicality filters beyond my initial BS detector. Suppose you put a receiver on that adjacent channel, dug out its signal, modulated that onto a clean virtual carrier, filtered and frequency shifted the recreated signal, and use that to perform some level of a nulling process against a delayed version of the signal. To make it work its best you must clean up both receivers at the same time.

Off hand in the US I do not believe this is a meaningful problem if you use 180 kHz to 200 kHz wide "IF" filters and forget about those lower fidelity digital sidebands.* I understand that Europe has gone out of its way to make this a real problem, though.

You MAY be able hack something up using SDRConsole receivers.

* The digital audio uses lossy (spell checker suggests lousy) compression. What comes out of compression resembles what went in to a lesser degree than the main "analog" channel.

{^_^}

RadioSDR1
Posts: 55
Joined: Sun Jun 26, 2022 4:41 pm

Re: Asymmetrical Filter Use on FM Broadcast

#3

Unread post by RadioSDR1 »

The theory that underlies your suggestion and the techniques involved I do not fully understand.

In the US, the trouble is that not every station broadcasts digital side bands. This is actually not trouble but rather a blessing because you can at least dig out weak stations on some frequencies that by chance are not affected by any adjacent digital broadcasting.

Just to be clear about what the issue is in the US, suppose you have a station broadcasting on 90.1 FM and also suppose it is broadcasting digital. That means its digital signal is destroying up the upper side band of 89.9 FM and also the lower side band of 90.3 FM. More specifically, each side band is approximately 75 kHz and in this example that digital signal would occupy spectrum from 89.90 through 89.975 and also 90.225 through 90.3. In other words, none of the 200 kHz on channel spectrum is used for digital. Instead, spectrum gets robbed from the lower and upper adjacent FM channels.

So if you want to try and dig out a station on 89.9 or 90.3 your choice is to either put up with the interference from the digital signal or remove the offending side band. But if the side band is removed, it of course causes intense distortion. The cat's meow would be a way to fix the distortion if technically possible.

jdow
Posts: 785
Joined: Mon Aug 10, 2020 8:17 pm

Re: Asymmetrical Filter Use on FM Broadcast

#4

Unread post by jdow »

It may be possible. I question how practical it would be. Everything around me is on 400 kHz steps so I was not sure if any of the 200 kHz steps between 400 kHz steps were used in the US. Of course, I am located where I could in theory listen to San Diego although I am nominally in the Los Angeles service area. I never went to investigate this. My FM receiver is SDRSharp and seems to be stuck on 91.5 MHz tuning off that frequency only when I fat finger something while fighting the Teratec (sp?) RTLSDR dongle have for it's front end. The AirSpy dongles seem to gravitate to real uses such as receiving signals from the dozen or so temperature and humidity sensors sprinkled around the house. They confirm that a rock and lime mortar wall is a very low impedance but has a very long time constant. (The AirSpy dongles are WONDERFUL. They just work. There is no guesswork.)

Anyway, your testimony confirms 200 kHz steps other than (4*N-1)*100 kHz, i.e. (4*N+1)*100kHz.

As I mentioned investigate what Youssef has done in this regard with sdrsharp. Be sure to download the manual a user wrote for it. He did a GOOD job.

I wonder what would happen if the upper side band was mirrored into the lower sideband to replace it. This thought had me wondering what the performance of different FM demodulators might be. The ancient ratio detector would work differently from the standard FM demodulator. And I've not cogitated long enough on PLL demodulation of varying degrees of loop filtering to form an opinion. It might act more like a ratio detector than conventional FM demod. An arctangent sort of demodulator would behave more like a ratio detector than FM demod. (FM demod is essentially a pair of simple filters one tuned high and one low both feeding amplitude detectors. The difference voltage is the demodulated FM. It's slope detection on a small dose of steroids. It is very amplitude sensitive.)

{^_^}

Post Reply