Question about wireless digital signaling

[Bararontok]

Is it possible to send a digital square-wave signal directly to a wireless transmitter without converting the signal into a continuous format such as the phase shift keying used in WiFi?

 

[kended]

Yes it is possible, but you'll have much trouble with your local broadcasting regulators as square waves contains many different frequencies (see Fourier analysis). Those wouldn't all be transmitted at high power due to the antenna's limited bandwidth, but still, you would be transmitting across multiple bands which is no good. Modulation serves that purpose so as to use/restrain to allocated frequency bands. Hope this helps.

 

[berkeman]

Is it possible to send a digital square-wave signal directly to a wireless transmitter without converting the signal into a continuous format such as the phase shift keying used in WiFi?

Good reply by kended. Bararontok -- do you have some application in mind? Or are you just trying to understand the limitations of wireless communication?

 

[Integral]

Not clear to me how a digital signal will contain the harmonics required to form a square wave. 1111122222111112222211111 is a digital square wave, there will be no harmonics introduced until you attempt to recreate the analog representation.

 

[berkeman]

Not clear to me how a digital signal will contain the harmonics required to form a square wave. 1111122222111112222211111 is a digital square wave, there will be no harmonics introduced until you attempt to recreate the analog representation.

But if you extend your sequence out a bit, window it and do a FFT on it, you will get components at DC, at your period, and at odd harmonics of your period, no?

 

[nsaspook]

I would expect the data transmitted to be self-clocked (to easily recover long 1 or 0 sequences in any practical system) so there would be edge transitions at every bit. http://en.wikipedia.org/wiki/Manchester_code
It's a small cheat but still completely digital and not a RF encoding method.

Each edge transition has a rise time. http://en.wikipedia.org/wiki/Rise_timehttp://upload.wikimedia.org/wikipedia/commons/f/f8/SquareWave.gif

 

[Integral]

But if you extend your sequence out a bit, window it and do a FFT on it, you will get components at DC, at your period, and at odd harmonics of your period, no?

Are you trying to get the analog representation?
:)
Looks like there may be a whole field here of which I am pretty ignorant. Sorry for interupting.

 

[Bararontok]

The understanding of square waves is that there are only discrete values with respect to time unlike the infinitely variable sawtooth, triangular and sinusoidal waveforms. Since inductors are used to convert electrical energy into electromagnetic fields and sending a fluctuating DC or AC current will cause the inductor to emit radiation, will square waves not just generate constant magnetic fields and not emit radiation that can be transmitted at a long distance and converted back to an electrical signal by the receiver? Because with a square wave, the power of the signal remains constant over a given period of time until the next component of the signal which can be a negative or 0 which will also remain constant for a given period of time. Does there not need to be a continuously fluctuating waveform electrical signal sent to the wireless transmitter for the transmitter to emit radiation?

Waveforms:

 

[Bararontok]

The thread originator is requesting an answer as to whether or not it is possible to send a square wave signal directly to a wireless transmitter inductor to convert the signal into radiation and then receive it at the end with a receiver inductor and covert it back into an electrical signal which is also a square wave.

 

[Bararontok]

Perhaps this other question might get an answer. Is the square wave digital signal first converted into a time-varying sinusoidal, triangular or sawtooth analog waveform using a digital-to-analog converter before it can be sent to the wireless transmitter? Because the inductor at the base of the antenna may need a time-varying oscillating electric current and not a discreet one to generate the oscillating electromagnetic field or radiation that transmits information.

An image of an antenna complete with circuitry, transmission inductor and the radiating element:

 

[berkeman]

I believe that is a receiving antenna, BTW. Ferrite rod antennas are not used for transmission, AFAIK (at least not efficient transmission).

The issue was mentioned earlier in the thread -- antennas are resonant structures (at least efficient antennas are), so you would have a situation where the current on the antenna would be resonant at the fundamental frequency of the square wave, so the current on the antenna would be mostly sinusoidal. To the extent that there are some harmonics present, the resonances would not generally be very efficient, so those harmonic currents would be attenuated compared to a square wave. They would still be large enough to cause you to emit electromagnetic interference at those harmonics, which would attract the attention of the relevant government agency (the FCC here in the US).

 

[Bararontok]

So it is the harmonic distortion of the square wave that can be emitted as radiation since there cannot realistically be a perfect square wave with a flat power level throughout the period of the wave. So are the time-varying sinusoidal, triangular and sawtooth waveforms more efficient for wireless transmission?

The issue was mentioned earlier in the thread -- antennas are resonant structures (at least efficient antennas are), so you would have a situation where the current on the antenna would be resonant at the fundamental frequency of the square wave, so the current on the antenna would be mostly sinusoidal.

So if a square wave is sent to the inductor, the antenna element at its core will have an induced electromagnetic flux that is sinusoidal, effectively converting a discreet signal into a continuous signal.

 

[berkeman]

So it is the harmonic distortion of the square wave that can be emitted as radiation since there cannot realistically be a perfect square wave with a flat power level throughout the period of the wave. So are the time-varying sinusoidal, triangular and sawtooth waveforms more efficient for wireless transmission?

Time varying sinusoids are used for RF transmission. It is very important to control the nature of the sinusoids, in order to share the RF spectrum efficiently.

So if a square wave is sent to the inductor, the antenna element at its core will have an induced electromagnetic flux that is sinusoidal, effectively converting a discreet signal into a continuous signal.

Correct.

 

[Bararontok]

The antenna's conversion of the discreet signal at the inductor to a continuous signal may be because when the power is cut off from the inductor abruptly due to the discreet signal that is running through it, there is still residual energy, analogous to inertia in a moving object, in the core element so the electrons in the core will not just come to an abrupt stop but will have gradually weakening currents resulting in the time-varying fluctuations.

Additionally, if the radiation received by a receiver antenna will once again be converted into a sinusoidal electromagnetic flux which will in turn be converted into a sinusoidal current by the inductor, does the receiver need an analog-to-digital converter to convert the signal into a discreet signal if a discreet signal is needed?

 

[berkeman]

The antenna's conversion of the discreet signal at the inductor to a continuous signal may be because when the power is cut off from the inductor abruptly due to the discreet signal that is running through it, there is still residual energy, analogous to inertia in a moving object, in the core element so the electrons in the core will not just come to an abrupt stop but will have gradually weakening currents resulting in the time-varying fluctuations.

The square wave is not a discrete signal -- it is a continuous time signal. But you are starting to see the "resonant circuit" part of the situation. Just think of the antenna as an LC resonant circuit, and you are driving it from a square wave source through some source impedance.

 

[Bararontok]

The square wave is not a discrete signal -- it is a continuous time signal. But you are starting to see the "resonant circuit" part of the situation. Just think of the antenna as an LC resonant circuit, and you are driving it from a square wave source through some source impedance.

Yes, the square wave is never perfectly discreet because it is impossible for a circuit to make completely instantaneous transitions between assigned values because of the residual energy that causes the power levels of the electrical signal to increase and decrease gradually. This time-varying decrease and increase of the power levels are referred to as the rise and fall times.

Source:

http://en.wikipedia.org/wiki/Square_wave#Characteristics_of_imperfect_square_waves

But what about the previous question:

Additionally, if the radiation received by a receiver antenna will once again be converted into a sinusoidal electromagnetic flux which will in turn be converted into a sinusoidal current by the inductor, does the receiver need an analog-to-digital converter to convert the signal into a discreet signal if a discreet signal is needed?

 

[Bararontok]

The thread originator is requesting an answer to the question in post #16.

 

[nsaspook]

The thread originator is requesting an answer to the question in post #16.

Yes. The received RF will need to be converted from a analog signal back to digital. This can be something as simple as a zero-cross detector after an amplification stage to recreate the data pulses from the RF signal zero crossing.

http://sound.westhost.com/appnotes/an005.htm