Hello
I am demodulating a signal. In which Low Pass Filter (LPF) is used to attenuate high frequency modulated signal. The input and outup of LPF is shown. The problem is the edges of LPF output signal is not square [as shown in figure] which result in poorly selection of lower and upper switching voltage of schmitt trigger [which retrieve orignal message signal]. So,i need the output of LPF to be exactelly square edge , which facilitate me to accurately select lower and upper switching voltage of schmitt trigger.
I keep on changing the cutoff frequency of LPF to obtain my required result but failed. If any one would suggest me any solution that result in obtaining square edged LPF output.

Thanks.

 

So,i need the output of LPF to be exactelly square edge...

Low-pass filters do not produce outputs with square edges.

 

Low-pass filters do not produce outputs with square edges.

Only someone who had no idea of what a filter is supposed to do would even contemplate such a request. You just cannot make this stuff up.

 

There is not enough information in your posting to allow me to try to give you an answer.
Obviously, the output from the low pass filter will not contain the rejected high frequency components of the signal so the waveform will have slow rise and fall times.
Why are you not using the unfiltered signal to trigger the schmitt trigger?
Regards,
Keith

 

Hello
I am demodulating a signal. In which Low Pass Filter (LPF) is used to attenuate high frequency modulated signal. The input and outup of LPF is shown. The problem is the edges of LPF output signal is not square [as shown in figure] which result in poorly selection of lower and upper switching voltage of schmitt trigger [which retrieve orignal message signal]. So,i need the output of LPF to be exactelly square edge , which facilitate me to accurately select lower and upper switching voltage of schmitt trigger.
I keep on changing the cutoff frequency of LPF to obtain my required result but failed. If any one would suggest me any solution that result in obtaining square edged LPF output.

Thanks.

Thanks for your response!
I know that LPF do not give exact square wave output ,however i would like to know what extra design at output of LPF should i add to get exact square edge output.

 

Low-pass filters do not produce outputs with square edges.

Hmm , well thanks for your appreciation
i know that i cannot get square edge from LPF output. I would like to know, extra circuit that i need at LPF's output to get square edge.

Looking forward!

 

Hmm , well thanks for your appreciation
i know that i cannot get square edge from LPF output. I would like to know, extra circuit that i need at LPF's output to get square edge.

Looking forward!

There isn't any. Square edges and LPF are requirements in CONFLICT!

 

i know that i cannot get square edge from LPF output. I would like to know, extra circuit that i need at LPF's output to get square edge.

There's your problem: you're focusing on the wrong thing. Somehow you've got the idea that the solution to your problem is to somehow get a "square edge" out of your LPF or some circuit connected to it, whereas the REAL problem is that your LPF doesn't adequately attenuate your carrier frequency relative to your modulation frequency. A 2- or 3-pole filter, instead of the single-pole filter you show, would do a lot better job and produce a much more usable input to your Schmitt trigger circuit.

I would suggest you do some reading about single- versus multi-pole filters, and about different filter characteristics (Butterworth, Bessel, Chebychev, Ellliptic, etc.) before going further.

 

There is not enough information in your posting to allow me to try to give you an answer.
Obviously, the output from the low pass filter will not contain the rejected high frequency components of the signal so the waveform will have slow rise and fall times.
Why are you not using the unfiltered signal to trigger the schmitt trigger?
Regards,
Keith

Well thanks for your response and explaination.
Moreover, i have attached block diagram of my system , which i am following. I hope this will answer that why i am using filter before schmitt trigger .
I can retrieve message signal by chosing variables of schmitt trigger , but there is slight difference between orignal message signal and retrieve signal at demodulator side[ attached].

 

I can retrieve message signal by chosing variables of schmitt trigger , but there is slight difference between orignal message signal and retrieve signal at demodulator side[ attached].

There will ALWAYS be a difference; it's unavoidable. You need to accept that, and deal with it.

 

There will ALWAYS be a difference; it's unavoidable. You need to accept that, and deal with it.

Well thanks sir for sharing your valuable knowledge and suggesions. Your comments always motivate and show me right direction.

I will try to study and design high order filters [as you suggested] and will see if it work !

Stay safe & blessed !

 

I will try to study and design high order filters [as you suggested] and will see if it work !

There is a tool available by Analog Devices you can simulate a filter response there and see how close you can get to the output you desire.

https://tools.analog.com/en/precisionstudio/

Check this out

 

Well thanks sir for sharing your valuable knowledge and suggesions. Your comments always motivate and show me right direction.

I will try to study and design high order filters [as you suggested] and will see if it work !

Stay safe & blessed !

see.

 

There is a tool available by Analog Devices you can simulate a filter response there and see how close you can get to the output you desire.

https://tools.analog.com/en/precisionstudio/

Check this out

Thank you so much ! I will definitly try it.

 

please watch this video on opto isolator you might get some hint. as a new engineer you should solve a problem by trail and error.

 

no idea

the filter can be anything-filter not just the frequency-band one . . . say -- you can "filter" out a "random" range of ser. comm. "bytes"

 

A square wave has harmonics of the fundamental, up to "infinity" ..
If say you have a 1 MHz square wave, the harmonics are 3,5, 7 etc MHz.

If you LPF a square wave, in the above example say at 2 MHz, ( with lots of attenuation ) then your only left with the 1 MHz fundamental , a sine wave.

How to get back to a square wave ?

there are many techniques of converting a sine wave to a square wave,
https://electronics.stackexchange.com/questions/163253/sine-wave-to-square-wave-schmitt-trigger
https://forum.allaboutcircuits.com/threads/sine-wave-to-square-wave-converter.85771/
https://www.engineersgarage.com/contributions/waveform-converter-circuits/

BUT, note,
you can get nice fast edge square wave, but the mark space ratio tends to not be to good,
the problem tends to be "noise" around the reference point, and or the variability in the window