EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: man_marius on October 17, 2018, 09:34:23 pm

Title: Superimposed differential signal
Post by: man_marius on October 17, 2018, 09:34:23 pm

I have a system that communicates by a differential signal, that is a superimposed signal of 4 different frequencies, say  f1,f2,f3,f4. These frequencies are amplitude modulated, How can I design a system to translate this differential signal into voltages f1->v1, f2->v2 ... f4->v4 simultaneously. The input differential signal can sometimes contain some of the frequencies, but not all.

My theory is to have four narrow band filter in parallel each filtering one signal, can that work?

Title: Re: Superimposed differential signal
Post by: moffy on October 17, 2018, 11:48:50 pm

Yes, if followed by a peak detector each. Or have 4 multipliers each fed with only one of the frequencies then low pass filter the output. Be aware of aliasing in your design. Look up amplitude modulation demodulation, should help.

Title: Re: Superimposed differential signal
Post by: man_marius on October 18, 2018, 05:56:53 am

What is easier to implement, the analog filters I have an ideal on how to make them, but it is a bit complicated because of the narrow band (or high Q factor).
On the other hand, the demodulator seems easy but I am not knowledgeable in that field

Title: Re: Superimposed differential signal
Post by: macboy on October 19, 2018, 02:31:54 pm

Is each frequency actually fully AM modulated or just OOK (On/Off Keyed)? The latter case is still technically AM (at 100%) but is much easier to decode.

Title: Re: Superimposed differential signal
Post by: man_marius on October 19, 2018, 03:32:32 pm

Its is fully modurated not only OFF/ON

Title: Re: Superimposed differential signal
Post by: moffy on October 21, 2018, 10:46:39 pm

The MC1496 multiplier is cheap, but requires a few passive components. The BW of the low pass filter is effectively your Q and easy to set.