Effect of rf notch filter on AM sound quality

[Circlotron]

In a normal AM receiver if we increase the selectivity more and more the received audio ends up sounding dull and muffled because the high audio frequencies in the sidebands are attenuated. What about the opposite situation where we tune a very sharp RF notch filter with a bandwidth of say 2kHz to a local strong station. Would this have the effect of attenuating all audio *below* 1kHz.

[TimFox]

For normal AM (two sidebands plus carrier) demodulation (e.g., simple diode rectifier), you need both sidebands and the carrier in the passband sent to the detector.
If you only pass the upper part of the upper sideband, you need a more complicated detector than what is supplied in an AM radio.

[Circlotron]

To restate the question, you are listening to your favourite AM BC band station and someone comes along and inserts a 2kHz wide, 100dB deep bandstop filter centred on your favourite station, in your antenna line. What will that do to the received audio spectrum?

[TimFox]

If the stop-band is centered on the signal, it will kill the carrier.
Without the carrier, your normal AM detector will not properly detect the baseband signal from the resulting double-sideband, suppressed-carrier signal.
You will need to artificially inject a carrier, after which your detector will extract a baseband signal without any bass.

[T3sl4co1l]

And, detecting it without a BFO*, results in distorted audio: the envelope is phase-sensitive, so you end up with something like abs(x(t)) which is a nonlinear function of the signal x(t).  Voice will still be intelligible (if poorly?) but overall fidelity, obviously, isn't great.

*Beat frequency oscillator, basically reinjecting the carrier, and by way of the detector, mixing it with the sidebands.  So, if it's not perfectly centered, you get split tones, e.g. a 1kHz signal shows up as Fc +/- 1kHz; if the BFO is off by say 100Hz from where Fc would fall in the IF chain, you get a new carrier Fbfo with sidebands of -900 and +1100Hz, detected as tones of 900 and 1100Hz.  And, I suppose, tones of 200, 2000 and etc. Hz because the sidebands mix with each other too.

Which also means you can still use a PLL to lock onto the original (suppressed) carrier.  So it's not too bad, but it is more work to do that of course.

SSB of course intentionally omits the carrier, and also a whole-ass sideband, so no double-mixing occurs, and so you don't get pilot tones or beat frequencies, just the warped Mickey Mouse voice when it's not tuned quite on center.

(Heh, I think you know much of this already, so, just putting this out there for anyone who doesn't know, and is curious what the full picture is.)

Tim

[cdev]

In a normal AM receiver if we increase the selectivity more and more the received audio ends up sounding dull and muffled because the high audio frequencies in the sidebands are attenuated. What about the opposite situation where we tune a very sharp RF notch filter with a bandwidth of say 2kHz to a local strong station. Would this have the effect of attenuating all audio *below* 1kHz.

If you have a selective RF notch fiilter and use it on an AM signal you definitely get an effect which has resemblances to the sound of a parametric equalization. But different. Presumably youre trying to eliminate the hetrodyning effect of a strong local carrier. Linrad does this fairly well. Its a pretty good SDR program for receiving very weak AM signals.. Sometimes Ive been able to use it to clean up very weak AM signals quite well.. You can get some very strange sounding effects with AM signals sometimes.. I have heard multiple echos of HFsignals.. Sideband also can do some strange things. Ive heard sounds that I could not explain.. using my SoftRock and Linrad.. Situations where I really had no idea what was going on.  There still are lots of parts of Linrad that I dont get.. Have never understood.. Sometimes I think very few people really understand them.

I think that others might be hard put to explain what was happening. Perhaps multiple simultaneous propagation effects due to selective delay on one path.. does that..Very strange sounds.  I love trying to figure out whats going on there.

[cdev]

How could they insert a bandstop filter into your receiver without direct access to your receive chain. It soulds almost like the mythical and kind of ridiculous "dome of silence" that would be the equivalent of negative RF..which doesnt exist.

To restate the question, you are listening to your favourite AM BC band station and someone comes along and inserts a 2kHz wide, 100dB deep bandstop filter centred on your favourite station, in your antenna line. What will that do to the received audio spectrum?

[Circlotron]

How could they insert a bandstop filter into your receiver without direct access to your receive chain.

It's just a "what if" story that I invented.
It's well known what a narrow RF bandpass filter does to the audio.
I wanted to know what a narrow RF bandstop filter of the same width would do to the audio.

[gf]

In a normal AM receiver if we increase the selectivity more and more the received audio ends up sounding dull and muffled because the high audio frequencies in the sidebands are attenuated.

This happens if you reduce bandwidth. But increasing selectivity does not necessarily imply that you must reduce bandwidth, but it rather means to increase stop band rejection and the steepness of the filter's transition band. Selectivity can be increased without renouncing bandwidth by using a filter with a higher order. I'm not sure whether there is a generally accepted metric for "selectivity". Some spectrum analyzers define selectivy as the ratio between -3dB and -60dB bandwidth - but this may be a proprietary definition.

What about the opposite situation where we tune a very sharp RF notch filter with a bandwidth of say 2kHz to a local strong station. Would this have the effect of attenuating all audio *below* 1kHz.

Basically, a symmetric notch filter centered at the carrier frequency is more or less equivalent to a highpass filter applied to the audio signal. However (as already mentioned in the thread), since it also suppresses the carrier, the signal is no longer a regular AM signal after filtering. If you want to achieve high-pass filtering of the audio signal, better do it directly in the audio domain, either prior to modulation or after demodulation, and not in the RF domain .

[cdev]

This is kind of interesting.. I have experience attempting to inject carrier to perform reception of ssb without the proper kind of demodulation.. Are you aware of any more software (beside linrad) that can appy a notch to a received AM signal? 

I recently got oe of these so called msi.sdrs.. basically a fairly broadband SDR, and have attempted to use that with linrad and its notch filter with a bit (not as much as I'd like) luck.. It works but it keeps having segmentation faults and with such a complicated program its frustrating relaunching it over and over.. I am looking for some way to prevent this..

[TimFox]

In normal multiplex FM, the (L-R) signal is encoded as double-sideband, suppressed-carrier (DSB) amplitude modulation with a suppressed carrier frequency of 38 kHz.  This is done with a balanced modulator, where the output is zero when the input (audio band) signal is zero (suppressing the carrier).
To decode this, a low-level "pilot" tone is included at 19 kHz, coherent with the suppressed carrier at 38 kHz.
Without a pilot, a "Costas Loop" PLL circuit  https://www.tutorialspoint.com/analog_communication/analog_communication_dsbsc_demodulators.htm  can be used to produce a coherent carrier for the demodulator, starting from the modulated DSB signal.

[cdev]

a narrow notch filter at an IF frequency is definitely possible, and the lower your IF is the more selective you can make this filter if you use a traditional analog design. Does that make sense in a communications receiver, would it be useful? This kind of thing and the difficulty of making allthese parts are why traditionally built receivers are expensive.

Unrelated: I am watching the NASA CATV channel today and the AX-1 mission listening to them trumpeting the "expansion of the abilities in private citizens in the development of private spaceflight". What this means is not so much private individuals so much as their supposed equivalent, corporations. This is very complicated in its implications and I don't honestly know yet how I feel about it. I hope that it works out and that they complete their mission smoothly.

[TimFox]

1950-era communications receivers often used a tunable crystal filter in the IF strip, where a notch could be moved around to reject interfering signals.
Some examples are in pp 3-7 of  https://eddystoneusergroup.org.uk/Restoration%20projects/Tech%20Short%205%20-%20Receiver%20Selectivity%20and%20Crystal%20Filters.pdf

[cdev]

Thanks!

[mawyatt]

FM Radio used pre-emphasis to improve SNR at higher audio frequencies. Think the inventor of FM, Armstrong, also invented this technique. The idea is the use a lead network, or highpass before FM modulation and then de-emphasize at the receiver which is a lag network or lowpass. Later audio tape recorders pickup up on this pre-emphasis/de-emphasise technique and was part of the older analog signal processing techniques that mostly have been replaced by DSP.

The reverse technique can also be very useful when the wanted signal is corrupted by noise and interference, here the signal is de-emphasized as close to the input preamp as possible, ideally making the preamp a lag network, low pass or integrator. Later in the analog processing chain a lead, high pass, or derivative is applied. The benefits of this technique is that large interference can be significantly reduced by integration, and selectively removed before signal expansion by the derivative to compensate for the earlier signal integration and help preserve high analog chain dynamic range. We utilized this technique in 70s to process a signal as low as a few tens of nanovolts at very low frequencies (<10Hz) in the presence of massive mains 50 and 60Hz noise and other large noise sources. Edit: Should mention the sensor was a magnetostrictive cable (MILES) which was rate sensitive to tiny pressure changes, and the up front integrator helped normalize the rate sensitivity.

Best,