Quadrature Signals
Two signals are in quadrature if:
- They have the same frequency
- They differ in phase by exactly 90 degrees
Example:
These are quadrature signals.
The signals are orthogonal, i.e., they do not interfere with eachother which, for example, allows two independent signals to share the same carrier frequency - very useful!
Sine and cosine are othogonal because:
I think of the orthogonality of the signals in the same way I was taught orthogonality for vectors (see linear algrebra notes on orthogonality and indpendence).
In linear algebra world, assuming the normal cartesian basis vectors, if I have a vector, , I can decompose it into basis vectors like this - - and I can only decompose it this way. There is no other combination of basis vectors that
can yield the vector . It's like I could mix two "signals", the vectors and in various quantities into a third signal, and then recover the original two signals and their "amplitudes" from that third, mixed up signal.
Same with the signal . If I just have the signal , in the same way as I did above,
I can decompose it into the basis functions and . I.e. is some mixed up jumble of two sinusoidal signals and I can recover the exact mix of the two sinusoids because those sinusoids are orthogonal or quadrature signals.
How?
To get , do the following...
We have:
Multiplying both sides by gives:
Thus, we have
Integrate that result over one period.
Which means that
Thus we have a forumla that results in . Multiple by to remove the half-period to ge just . This gives us the result
This completely an unambiguously extracts the "amount", of the basis vector contained in our signal .
To find the same method is repeated, but with sine, to get:
IQ (Quadrature) Demodulator
This All About Circuits article is a good ref.
A signal composed of two quadrature waveforms can be "decoded" by an IQ demodulator to extract the in phase component and the quad component.
If I have a singal I