AM SSB

[Jarfaut]

Hello, comrades!
Okay, at first I was too shy to ask publicly, so I submitted my questions to Spogg's personal account. But Spogg, after giving me an answer, advised me to write about the problem in a public topic. And so now I have gained some courage.

The question seems simple, but in fact requires both technical training and an understanding of the environment in which the experiment is to be conducted (i.e. FS).
I want to build the following scheme in FS:

fsd.jpg (25.01 KiB) Viewed 9342 times

Many if not most of you know what amplitude modulation is.
The diagram above depicts a method to mimic the effect observed by radio amateurs in the Short Wave band. This circuit simulates the reception of signals transmitted by SSB modulation.

In my first thread I asked about the upper possible frequency of the oscillators in FS, meaning close to the radio frequency.
And now I want to correct right away that for the above scheme to work it is enough to generate the carrier at frequencies of the order of 20 kHz.

I tried to make an imitation of this Shortwave reception mode myself. But I am very poorly familiar with the principle of the elements and modules in FS.
Nevertheless, I attach my example at the end. Connect it to a DAW, and listen through it any audio file with speech. It's certainly not quite what you need, but a bit like real radio reception.

So, anyone who can help implement the above scheme in FS will be in for a surprise and my thanks!

Regards!

My-aM1.rar

(2.24 MiB) Downloaded 243 times

[Jarfaut]

I asked this question on the Russian-language resource, and on the Spanish-language resource. But they could not answer me because they do not work in this program (Flowstone).

If this scheme cannot be made in Flowstone, then it will be made in other programs.

In short, this circuit imitates the reception of AM SSB radio signal.
No radio frequency is required for this circuit. A 20 kilohertz oscillator is sufficient.

[Tepeix]

They are certainly people here that know SSB, but maybe not so commonly, as it's more for transmission than for FX sound.

It remind me frequency shifter that seams really close, they are some on the forum.

But if you have a receptor, don't you need to have a transmitter, which is probably a frequency shifter ?

Here's a topic about it : viewtopic.php?f=3&t=3175&hilit
I also have frequency shifter from Trogz but i could not find from where i get it.

[martinvicanek]

Oh, that looks familiar! That’s a frequency shifter, if I’m not mistaken. Right, there is nothing that would restrict it to RF, it works perfectly well for audio, too. I built one some time ago, you might still find it on the forum.

[Jarfaut]

Oh, that looks familiar! That’s a frequency shifter, if I’m not mistaken. Right, there is nothing that would restrict it to RF, it works perfectly well for audio, too. I built one some time ago, you might still find it on the forum.

martinvicanek, in the audio materials, as I have seen, it is really called that.
Did you do it according to this scheme?

I'm curious to understand - how to do it on the elements in FS.
For example, the first mixers act as modulators, and the second mixers act as demodulators. But how to make these demodulators in FS?
This would be a great course for me on working in this program.
Respectfully!

[Jarfaut]

Greetings, comrades!
Many thanks to Tepeix and Martinvicanek for their great work (viewtopic.php?f=3&t=3175&hilit) !!!
And now I will add my comment on that work.

You made the most complex variants of the SSB converter, which shows your high level of technical expertise. (I, on the contrary, am weak in math and had to compare data from technical sources).
So. Your thread discusses two basic methods of getting SSB. Both are difficult to implement.
Graphically, these two methods can be depicted as follows:

two meth sb 1.png (74.09 KiB) Viewed 9247 times

The first of these is the filter method.
As the technicians write, its problem is with signal convolution at frequency boundaries, the consequence of which is a decrease in the useful bandwidth. Therefore, in this method significant frequency shifts are possible only with a multistage procedure.

The second method is called phase-shifting, and uses a phase-shifting all-pass filter called the Hilbert Transformer.
The difficulty of such a method is the complexity of such a phase-conversion filter itself. Such a filter gives the desired phase shift only in a limited range of frequencies. That is, for an expressive effect, you will have to apply a group of bandpass filters, each of which must be followed by a Hilbert transformer.

The third method is shown in my diagram in the first post.
Unlike the first two methods, this method does not require any phase-shifting filters, and requires only two quadrature oscillators.
It works clearly and without errors. The low-pass filters can be any kind of the same, and their phase characteristics do not matter.

The frequency shift is performed by adjusting either of the two oscillators. And it is effective over a wide range of frequencies.

Therefore, I ask you to participate, and make this circuit in FS.

Respectfully!

[martinvicanek]

Here is a demo for you. I thought I had posted something similar before, but I can't find it any more. Any way, this is not optimised, but it is fully functional. Have fun with it!

[Jarfaut]

martinvicanek, great!!!

Writing especially for you and Tepeix!
In that thread of yours, you did the hardest job with the hardest method of getting SSB.
The basic methods are two: filter and phase (with a Hilbert transducer).
Their graphic sketch:

two meth sb 1.png (74.09 KiB) Viewed 9247 times

Both are difficult to implement.
The first is the filter method has the problem of signal convolution at frequency boundaries, which reduces the useful bandwidth.
The second, with Hilbert, allows a shift in a very limited bandwidth, and therefore requires several "bandpass filter-phase-shifting Hilbert" stages.

You have done the most difficult implementation.

But my picture in the first post shows a third less known method, which does not require phase-shifting filters, and it needs only two oscillators. It works clearly, and over a wide frequency range. It is the best solution, and easy to implement! The shift frequency can be adjusted by either of these two oscillators.

Thank you for your work!
Regards!

[Tepeix]

I understand more now too.
I remember trying to make your schematic, but failing.
Not knowing witch frequency i must use for the first quadrature oscillator.

So to simulate the transmission and reception, we must shift 2 time with the same amount but inversed.
But in this case, i don't hear the difference, maybe there's an effect but it's very subtle, because the shifter are very precise.

So here's my bad frequency shifter, we hear a not subtle effect in case of transmission, it's not very beautiful here i think there's to much imprecision..
(I tried to make an Hilbert approximation with allpass filter in parallel but it's very not precise
Also i use an imperfect oscillator..)

(The schematic also include the frequency shifter of Trogz to compare)

[Jarfaut]

Tepeix, hi!

You know, Martinvicanek made it work! He made a demo using this scheme, and this simulation works great.
So no more phasing filters!