Talk:Double-sideband suppressed-carrier transmission

Spectrum picture is wrong, and the examples are not suitable
It is a lousy article, in general, it looks like a brainstorm on DSB-SC made by undergraduates. The formulation is not consistent, nor clear. The example is with a spectrum, but the formulation is done with a tone "Am cos (wmt)", which is a case incapable of showing the generic operation. It would be necessary to make the explanation using the Fourier transform. It makes it seem more complex than it actually is.

Merge proposal
The merge proposal first said merge to Double sideband suppressed carrier, but there was no discussion started and no matching tag there, so I did the logical thing and switched it to propose merging that more primitive orphaned article with wrong title punctuation into this more mature one. There was nothing there not already here, so I went ahead and just redirected it. Any objections? Dicklyon (talk) 06:02, 4 December 2008 (UTC)
 * I figured the shorter name was the better, but what the heck. Thanks; now it's just a matter of finding double redirects and making them longer direct links.  Jim.henderson (talk) 01:37, 9 December 2008 (UTC)


 * There's a bot that fixes double redirects, so I wouldn't sweat that. As for which name is preferred, I don't have any objection to using the shorter one, but the proper merge was into the one with a substantial history.  Now that it's done, you can do a move request (the link to show you how comes up if you try a simple move and hit a conflict). Dicklyon (talk) 16:12, 9 December 2008 (UTC)

Demodulation
"DSB-SC can be demodulated if modulation index is less than unity."

According to the AM article, ideal DSB-SC has a modulation index M/A = 2 (200%); assuming the modulating signal varies on (-1,+1) with a carrier amplitude of unity. Am I missing something? — Preceding unsigned comment added by WikiSavesLives (talk • contribs) 06:11, 23 May 2014 (UTC)


 * There are two equivalent formulas for the AM modulation index [m]:
 * [1] m = Vm / Vc


 * [2] m = (Vmax-Vmin) / (Vmax+Vmin)
 * where:


 * Vm is the peak voltage of the modulating signal.


 * Vc is the peak voltage of the sinusoidal carrier.


 * Vmax is the algebraic maximum amplitude of the AM signal (it is positive, being in phase with the carrier always).


 * Vmin is the algebraic minimum amplitude of the AM signal (it is positive if in phase with the carrier, otherwise it is negative).


 * Also:


 * Vmax = Vc + Vm


 * Vmin = Vc - Vm


 * Therefore,


 * If Vm = 0, m = 0


 * If 0 < Vm < Vc, 0 < m < 1


 * If Vm = Vc, m = 1


 * If Vm > Vc > 0, 1 < m < ∞


 * If Vc = 0, m = ∞


 * The modulation index of a double side band suppressed carrier [DSB-SC] signal is infinity [ m = ∞ ].
 * KerimF (talk) 13:43, 29 July 2021 (UTC)

A simple low-cost linear demodulator for DSBSC
About 37 years ago, I built, as a test (started as an MS thesis project), a simple linear demodulator for a double-sideband suppressed-carrier (DSB-SC) signal. By this test, I liked to prove the opposite of the statement that says: Recovering the modulating signal of a DSB-SC signal is much harder than of a single-sideband suppressed-carrier one (SSB-SC). After all, it would be illogical that receiving two side-bands (symmetrical) is worse than receiving just one of them.

Its topology could be simplified as:

AM-DSB-SC -->|Limiter|--> |Duty shaper 25%| --> |PLL at 2Fc| --> |Div. by 2| |                                                                |     |                                                                 |      > |switch| ---> |LPF| ---> signal

For fc=455 KHz (AM IF), I used: LM339 for the blocks of |Limiter| and of |Duty shaper 25%|,

CD4046 for |PLL at 2Fc|,

CD4013 for |Div. by 2|,

CD4066 for |switch|.

Obviously, this demodulator works for any AM index; that is from m=0 (no modulation) to infinity (DSBSC). It is also suitable for hi-fi audio signals as well (as long the receiver bandwidth is wide enough). And its circuit could be integrated in one low cost IC, if necessary.

But I was surprised that, after about 3 decades, this simple topology for DSB-SC demodulation is not known universally yet (mainly at any faculty of electronics in the world). No one among the professors/experts in AM communications, on these days, seems being interested in it, perhaps because it is too late to make money of it. Meanwhile, all graduates in communications are made to believe that a simple linear versatile AM demodulator, for any AM index (including m=infinity), doesn’t exist (not known yet).

It will be my pleasure to provide and discuss its actual analysis, done on LTspice, to/with anyone who may be interested in. Otherwise, since I am already 67 years old (living in Aleppo city, Syria), my work will, sooner or later, have to die with me ;)

To the old/expert editors of wikipedia:

Since you removed my note in the "Demodulation" section on the Article page, I wonder for which reason you didn't remove this post too. And, if you insist/believe that what I say is not true, I guess I should help you by also erasing what I wrote here with my best wishes to wikipedia.

Edited:

It seems, after about 3 months, no professional people in electronics and/or RF communications (or brilliant undergraduates) visited this page on Wikipedia. I usually talk to serious individuals only who may have the same interest in a certain subject (as the one here). So till a miracle happens, the undergraduates in communications will have to believe the fallacy that says: "Demodulating SSB-SC signal is much easier than demodulating DSB-SC one". In reality, the inverse is true... but who cares! And I don't intend to convince the high class as Galileo did ;)

— Preceding unsigned comment added by KerimF (talk • contribs) 15:35, 28 March 2016 (UTC)


 * What stops the PLL going slightly off frequency or even just out of phase during silent periods in the transmission? I suspect that it will and when transmission starts again you will get a brief period of distortion while it locks on again. BrianDGregory (talk) 17:50, 5 December 2018 (UTC)


 * How is adding the limiter and PLL making it simpler than an SSB demodulator which only needs the oscillator to be manually tuned within a few Hz of the original carrier frequency? You seem to be claiming that removing the limiter and PLL so it becomes a normal BFO type SSB demodulator makes it more complicated?! BrianDGregory (talk) 17:58, 5 December 2018 (UTC)


 * Yes, the PLL will go slightly off frequency and out of phase during silent periods in the transmission. But we will not get, after a brief period of silence, a noticeable distortion. Its lock in is relatively fast. I personally used this technique for many years (in the 80’s) in my private short-range RF voice transmissions. Even with a suppressed carrier of a low frequency as 32,768 Hz, we didn’t notice a noticeable distortion during normal conversations with silent periods. For instance, this 32KHz DSB-SC was used instead of the 38KHz (used in stereo FM); obviously I didn’t need to add a pilot and I let the 16 KHz base-band be empty (my FM channels sounded silent on conventional FM receivers).


 * On the other hand, my technique is simpler because its circuit is made by elements which could be integrated all on one low cost package. It is also simpler because it functions like AFC (Automatic Frequency Control). And being a synchronous demodulator, it can be used for Hi-Fi music (if the channel bandwidth is made wide enough). But being much simpler and more practical than SSB-SC demodulators has obviously a price; a DSB signal covers twice the bandwidth of the SSB one.


 * Anyway, who on earth can listen to Kerim? Therefore, the world's Elite of Science will keep teaching the new generations that a simple reliable analogue demodulator for a DSB AM signal having any modulation index (including m=infinity, that is DSB-SC) CANNOT exist. KerimF (talk) 01:28, 26 March 2021 (UTC)

Why don't they consider this modulation method for the current improvement of the AM broadcast band?

It has been stated in many articles concerning the state of the AM broadcast band (540-1650 kHz) that due to high operating costs, and spotty coverage areas, that the usefulness of "Antique Modulation" is no longer cost effective. This modulation method with reduced carrier, would still be able to be demodulated with a diode based envelope detector, and use less power (50% vs 33.33%) than the current methods. And although it's been a long time since I reviewed modulation indexes, I believe the resultant signal would be a vast improvement over the current AM wasting so much power on the carrier.KD6UAO (talk) 20:29, 23 October 2018 (UTC)


 * The reason is that FCC has been made to believe that a reliable DSB-SC demodulator cannot be made cost-effective. About 40 years ago, during my MS studies, I liked to prove, once for all, that by using twice the SSB-SC bandwidth helps the DSC-SC demodulator be much simpler and more practical than of SSB-SC. And I designed and built the first version of my low-cost DSB-SC analogue demodulator as an MS thesis. You may also like reading my recent reply to Brian above. KerimF (talk) 02:07, 26 March 2021 (UTC)

Added: The one-PLL demodulator (introduced above a few years ago) can be tweaked to recover two signals modulated in quadrature as shown below:

|-< Stereo-AM-DSB-SC [Fc] >-| |                        |                         |        |                         |                         |        |                      Limiter                      | |                        |                         |        |                         |                         |        |                 Duty shaper 12.5%                 | |                        |                         |        |                         |                         |        |                    PLL at 4*Fc                    | |                        |                         |        |                         |                         |        |                     Div. by 2                    | |                        |                         |        |            _____________|_____________            |        |           |                          |            |        |           |                          |            |        |     Div by 2, 0 dg            Div by 2, 90 dg     | |          |                          |            |        |           |                          |            |        |--> switch1                    switch2 <---| |                         |                    |                          |                   LPF1                       LPF2 |                         |                    |                          |               Left + Right            k*(Left - Right) k < 0.5

I also can’t elaborate this topology on this page because Wikipedia is not a place for novel solutions. But it will be my pleasure to provide its LTspice files to the first one who may be interested (perhaps for his MS or PHD thesis) to implement it practically. I assume he is also familiar, speaking scientifically, to such type of transmission so that we can discuss any part of it when necessary. KerimF (talk) 22:05, 30 May 2021 (UTC)