Prj01 - A Simple Reliable Double Sideband Suppressed Carrier (DSB-SC) Demodulator

[KerimF]

To those who are interested in AM communications, I attached below LTspice files which introduce the simplest reliable analogue AM demodulator. It works for all modulation indexes; from m=0 (no modulating signal) to infinity (no carrier).

It uses one PLL (not two PLLs in quadrature as in Costas Loop). It locks with the suppressed carrier (frequency and phase). So, it has also a frequency lock range much like in FM receivers. Also, unlike the case of the Squaring Method, it doesn’t need any selective filter (passive or active). Being simple and reliable, it could, therefore, be made as a low-cost analog integrated circuit, IC.

On the attached example, the carrier 455 KHz (AM IF) is assumed suppressed (worst case). For simplicity, the audio signal is emulated by the model of a white noise generator.

Kerim

 

[D.A.(Tony)Stewart]

I would expect DSB-SC is not used due to the SNR problem at low signal levels. With a minimum carrier level and known white noise level of the Rx, there is a guaranteed minimum CNR and SNR after filtering. I seem to recall thinking this in the lab at Univ. 50 yrs ago.

 

[KerimF]

I would expect DSB-SC is not used due to the SNR problem at low signal levels. With a minimum carrier level and known white noise level of the Rx, there is a guaranteed minimum CNR and SNR after filtering. I seem to recall thinking this in the lab at Univ. 50 yrs ago.

I guess what you say has to be true in certain applications.

But the DSB-SC system worked fine in my private RF links on the MW band for voice transmission.

I also used it to scramble my FM channels. The voice signal modulated a suppressed carrier of frequency 32768 Hz (yes, of a crystal) instead of the 38 KHz (used in stereo FM). Obviously, the audio baseband (R+L) was empty, and my demodulator didn't need a pilot (as the 19 KHz in stereo FM).

By the way, in year 1979, I was told at the university that it is somehow impossible for a simple reliable DSB-SC demodulator to exist. So, I decided to prove, as an MS thesis, that receiving more information (DSB-SC) should help recovering the modulating signal much easier than in case of receiving less information (SSB-SC). But I ended up proving this to myself only, because I couldn't submit my work (this has its own story). And I took advantage of what I did for my personal use only (taking advantage that no one knows it ).

 

[D.A.(Tony)Stewart]

I didn't say that it doesn't work
This is what OpenAI says
how does simple reliable a DSB-SC AM demodulator work with a pll




A DSB-SC AM (Double Sideband Suppressed Carrier Amplitude Modulation) demodulator with a PLL (Phase-Locked Loop) is a common method used for demodulating a DSB-SC AM signal. The demodulation process involves the following steps:

1. Input signal: The DSB-SC AM signal is the input signal to the demodulator.
2. Bandpass filter: The input signal is first passed through a bandpass filter that allows only the desired frequency range of the modulated signal to pass through, while rejecting unwanted frequencies.
3. Mixer: The filtered signal is then mixed with a local oscillator signal that is set to the carrier frequency of the modulated signal. The output of the mixer is a baseband signal that contains the original modulating information.
4. Phase detector: The output of the mixer is then fed into a phase detector, which compares the phase of the output signal with that of a reference signal from the local oscillator. The phase detector output is a DC voltage that is proportional to the phase difference between the two signals.
5. Low-pass filter: The output of the phase detector is passed through a low-pass filter that removes the high-frequency components and leaves only the DC voltage.
6. Voltage-controlled oscillator (VCO): The DC voltage from the low-pass filter is then used to control the frequency of a voltage-controlled oscillator (VCO). The VCO output is fed back to the mixer as the local oscillator signal.
7. Output signal: The demodulated signal is obtained at the output of the low-pass filter.

The PLL is used to lock the phase of the local oscillator to that of the carrier signal, which helps to maintain the correct phase relationship between the modulated signal and the local oscillator. This ensures that the mixer produces an output signal that contains only the baseband information. The PLL also helps to reduce the effect of noise and interference on the demodulated signal, making it a reliable demodulation method.

 

[KerimF]

I didn't say that it doesn't work
This is what OpenAI says
how does simple reliable a DSB-SC AM demodulator work with a pll




A DSB-SC AM (Double Sideband Suppressed Carrier Amplitude Modulation) demodulator with a PLL (Phase-Locked Loop) is a common method used for demodulating a DSB-SC AM signal. The demodulation process involves the following steps:

1. Input signal: The DSB-SC AM signal is the input signal to the demodulator.
2. Bandpass filter: The input signal is first passed through a bandpass filter that allows only the desired frequency range of the modulated signal to pass through, while rejecting unwanted frequencies.
3. Mixer: The filtered signal is then mixed with a local oscillator signal that is set to the carrier frequency of the modulated signal. The output of the mixer is a baseband signal that contains the original modulating information.
4. Phase detector: The output of the mixer is then fed into a phase detector, which compares the phase of the output signal with that of a reference signal from the local oscillator. The phase detector output is a DC voltage that is proportional to the phase difference between the two signals.
5. Low-pass filter: The output of the phase detector is passed through a low-pass filter that removes the high-frequency components and leaves only the DC voltage.
6. Voltage-controlled oscillator (VCO): The DC voltage from the low-pass filter is then used to control the frequency of a voltage-controlled oscillator (VCO). The VCO output is fed back to the mixer as the local oscillator signal.
7. Output signal: The demodulated signal is obtained at the output of the low-pass filter.

The PLL is used to lock the phase of the local oscillator to that of the carrier signal, which helps to maintain the correct phase relationship between the modulated signal and the local oscillator. This ensures that the mixer produces an output signal that contains only the baseband information. The PLL also helps to reduce the effect of noise and interference on the demodulated signal, making it a reliable demodulation method.

I am afraid that one PLL whose VCO mid-frequency is made to be equal to the frequency of the suppressed carrier doesn't work
Why?
At every zero crossing of the modulating signal, the phase of the suppressed carrier shifts 180 degrees. And this reverses the direction of the phase comparator output.
To avoid this, doubling the VCO mid-frequency and adding a duty shaper to change the duty cycle of the signal squarewave (at the signal input of the phase comparator) from 50% to 25%, the phase comparator won't be affected by the phase reversal of the suppressed carrier.

For instance, after 3 months of failed designs at the university lab, I don't know how I had the idea, just a couple of days before leaving the lab, to use a PLL and set its mid-frequency at 910 KHz instead of 455 KHz which is of the suppressed carrier (AM IF). And I did the test though my study showed me clearly that it couldn't work... another failure.
To my big surprise, it worked (much like it happened to Archimedes ). First, I thought that the musical audio input (from a tape recorder) was connected, by mistake, to the output speaker. But when I varied, a little, the frequency of the carrier, the sound was off.
Truth be said, on that day, I had no clue about how it was possible for this simple PLL to work (Costas Loop uses two PLLs in quadrature). Only when I returned home, I had enough time to examine every node of it. The cause was simply the natural imperfection of the high-gain comparator (as LM339, to convert the analogue DSB-SC signal to squarewave). The duty cycle of the generated squarewave was slightly different from 50%. This small difference was enough to let the PLL locks though in a very narrow band.
A further study let me know that the widest lock range occurs if the duty cycle at the output of the high-gain duty shaper is made to be 25% (or 75%).
Although, on the actual schematic, there are two 25% duty shapers, the demodulators in my private links worked fine with one only.

 

[D.A.(Tony)Stewart]

I never mastered LTSpice but I get a similar response to yours (in my graphic theme) However the output does not match the input audio and appears to have significant 2nd order distortion which is produced by the XOR PC1 mixer.

Reducing the RUN time has the same effect

--- Updated Apr 7, 2023 ---

--- Updated Apr 7, 2023 ---

OK after finding something moved in sub-cct But why no connections to VCO RC inputs for f.

 

[KerimF]

But why no connections to VCO RC inputs for f.

Me too, I asked this question when I used the model of CD4046B for the first time. Unfortunately, the model, as it was, didn't work (it stopped the simulation).
If you open its schematic 'CD4046B_h.asc' you will find that the pins of R1, R2, C1A and C1B have no functions. They were added just to be seen on the CD4046B symbol. And their functions are done by a special LTspice device, 'Modulate', (its label A18 on the schematic) which is good for AM and FM (explained in LTspice help).
You will also notice on its schematic that the functions of PC2, PCP and ZD pins don't exist at all. I had to also remove the circuits of these functions (from the original model) to let the simulation run. After all, these pins are not important in this application. I think, I also did some minor modifications.
I attached its original schematic 'CD4046B_h-org.asc' (I changed its name by appending '-org' to it).

 

[D.A.(Tony)Stewart]

I saw the ideal FM functions. param F_mid=920KHz, F_lock=50KHz and guessed that somewhat. Falstad's Sim. does something similar with ideal FM except uses the actual RC values to modulate the VCO of the 4046.

I like your method of AM IF encoding +/- inputs signals of 2f. Then both open collector comparators drive DSout to SIG with pullup. I see the Audio modulate 2f PM then VCO integrates to FM then the /2 FF samples the phase of the SSB carrier using the S&H 4066 as a mixer to demodulate the input signal V(DSB-SC+)/2+2.5

It looks much like how a PWM is generated and degenerated to baseband but with a synchronous demodulated suppressed carrier.

It works ! Although I didn't stress test the design for a transfer function with a wide dynamic amplitude and frequency range.

 

[KerimF]

My first motive to find out a simple reliable topology to demodulate DSB-SC signal was, about 4 decades ago, to prove (to myself in the least) that receiving more 'good' information (as in DSB) has to make life easier than receiving less (as in SSB). Even in these days, the inverse of this is taught (believed) at all universities!
It happened that I used to be, since I was teen, a person of reason (at school, my favorite was math, leaving literature and history for others). This applies on any matter, material or spiritual, which I 'needed' really to think of to avoid walking in the darkness of ignorance (a weakness with which every intelligent human baby is born, I included). I couldn't (can't) believe something just because some others do, even if they form the greatest majority in the world (this reminds me Galileo ).

It works ! Although I didn't stress test the design for a transfer function with a wide dynamic amplitude and frequency range.

I guess this may be necessary to do if there is a need to apply it in a real application.

I used this demodulator when I 'needed' really an RF voice link between my apartment (in a rather poor neighborhood at that time and it didn't have a phone line) and my workplace (about 3 km apart, air distance). This was necessary because this had to be a private transmission (for many years) and any radio listener had to think it was just a noisy interference (on MW band) or a blank/silent channel (on FM band).

 

[D.A.(Tony)Stewart]

. This was necessary because this had to be a private transmission (for many years) and any radio listener had to think it was just a noisy interference (on MW band) or a blank/silent channel (on FM band).

I would think one could understand your voice by tuning AM radio to one side of the DSB rather than the worst case being centred. So it might not be as secure as you say. In 1982, I recall designing a Moog Like synthesizer for TV video baseband using a dozen different encoding methods for encoder/decoder so that the client could see the security of introducing Pay-TV over the air and on cable. It included every known method at the time. The end-choice was the cheapest which was suppressed Sync. which we joked would sync up on dark "bedposts on adult films" In your case, it was security from political rivals.

 

[KerimF]

I would think one could understand your voice by tuning AM radio to one side of the DSB rather than the worst case being centred. So it might not be as secure as you say.

It seems you missed that the frequency of my suppressed carrier on AM band varied (+/- 30 KHz) at a rate of 6 Hz. Your remark is right if its frequency was fixed (as in a conventional DSB-SC system).
There was no special reason to scramble my transmission other than being private while we talk as we do on an ordinary telephone.

The irony was that when I met some local intelligence agents (related to communications) at their offices to show them, by a test, how my DSB-SC demodulator could be also used in a scrambling system, they couldn't believe me. Yes, they (and a so-called professor in electronics at the capital university) couldn't believe, no matter what I did/said, that a local engineer can present something not known yet in the world (mainly in USA or Russia). For instance, what happened is much more surprising than this, mainly at the national research center of the capital then of my city. So, I ended up being free to scramble my own links, for many years, till I got a phone line

 

[D.A.(Tony)Stewart]

It seems you missed that the frequency of my suppressed carrier on AM band varied (+/- 30 KHz) at a rate of 6 Hz. Your remark is right if its frequency was fixed (as in a conventional DSB-SC system).

where?
It was not listed or shown, there was only 1 PLL a well.

 

[KerimF]

where?
It was not listed or shown, there was only 1 PLL a well.

https://www.edaboard.com/threads/q01-%E2%80%93-voltage-to-phase-shift.406471/page-2

post #26

"But truth be said, this is not important, speaking practically. This gain variation wasn't noticed in a voice transmission when I used the DSB-SC system in the 80's (for many years) in my private short-range RF links (about 3 km). I deliberately varied the frequency of the suppressed carrier over many channels on MW band (about +/- 30 KHz, at a rate of 6 Hz) so that the MW listeners hear sort of noise interference when a conversation was in progress (no RF signal during silent period which is one of the advantages of the DSB-SC system)."

 

[D.A.(Tony)Stewart]

No RF in a silent period can be an advantage, to save power, but still not silent enough for co-channel FDX, but certainly OK for offset Tx-Rx carriers.

 

[KerimF]

but still not silent enough for co-channel FDX,

I wonder what the connection between DSB-SC and co-channel FDX could be.
I am not good in abbreviations/acronyms. I assumed that FDX could mean Full Duplex.