Loading...

Postulate is the best way to take and share notes for classes, research, and other learning.

More info

Lecture 24: modulation

Profile picture of Samson ZhangSamson Zhang
Dec 4, 20232 min read

Signals can be encoded in amplitude or phase. A subset of phase modulation is frequency modulation.



Modulation

Amplitude modulation can be accomplished with a mixer (how?).



Frequency modulation can be accomplished with a VCO (voltage controlled oscillator). Describe FM with:

  • carrier frequency
    f_c fc f_c
    , the frequency of the underlying sinusoid
  • frequency deviation
    \Delta f Δf \Delta f
    , the difference between highest and lowest frequencies being modulated between
  • modulation frequency
    f_m fm f_m
    , how fast the change between different frequencies

The input to the VCO is

V_{f0} + V_{\Delta f}(t) Vf0+VΔf(t) V_{f0} + V_{\Delta f}(t)
.

The bandwidth is

BW = 2 \big( \frac{\Delta f}{f_m} + 1 \big) f_c BW=2(Δffm+1)fc BW = 2 \big( \frac{\Delta f}{f_m} + 1 \big) f_c
(why?).





Phase modulation is complicated. A simple example is to flip the phase 180 degrees between 0 and 1 bits. This can be accomplished by switching between sin and cos signals. The bandwidth is

BW = 2 \big(\frac{\Delta \phi}{2 \pi} + 1\big) f_c BW=2(Δϕ2π+1)fc BW = 2 \big(\frac{\Delta \phi}{2 \pi} + 1\big) f_c




Lastly, there's something called an IQ modulator. I don't understand it



Demodulation

Mixers are hard to use for demodulation because they need to be synchronous, i.e. in-phase, to work. Otherwise signals will cancel and the IQ plot will rotate.



AM demodulation is easy, it can be done with an envelope detector or HD2 self-mixing or homodyne demodulator.



FM demodulation can be done with a slope detector (different gains from different frequencies with linear dB dropoff) or other ways. Frequency counters and differentiators are difficult to use for demodulation directly because it is hard to build high-frequency versions.



There's also something called a phase lock loop to make a reference block for PM demodulation



Receivers

Receivers come before demodulators (?)

Direct sampling is difficult because

f_s > 2 (f_c + f_{sig}) fs>2(fc+fsig) f_s > 2 (f_c + f_{sig})
to prevent aliasing

In undersampling, a signal is deliberately aliased and then picked out with a sharp anti-aliasing bandpass filter. This allows the sampling frequency to be a lot lower



Direct downconversion is very sensitive to IIP2. Not sure what that means



Superheterodyne receivers are multi-stage, using an intermediary frequency. Don't entirely understand how they work either



Here's a diagram of IQ modulation/demodulation. confused



Comments (loading...)

Sign in to comment

e157: radio frequency circuit design

fall 2023 class w prof. spencer