EE 351 · Communication Systems · Exam preparation

Major Exam No. 1 — Guided Practice: AM and FM

Recognize the signal. Choose the model. Solve with units. Explain what the result means.

This practice covers skills that may appear in Major Exam No. 1. It uses only material from EE 351, Weeks 2–3 and 5–6. The exam may not repeat these questions. It may change the waveform, the numbers, the diagram or the situation. Learn the method, not the answers.

8 guided exercisesAM → FM → engineering decisions18-minute self-test

Build a method you can reuse

Each card has a short review, a guided example, a problem for you to try, and a worked solution. Try the problem first. Open the hint only if you are stuck. After you check your answer, read the exam box and ask yourself: how would I start if the given and unknown quantities were swapped?

Your working pattern: Given → Find → Solution → What it means. Keep \(f_c\), \(f_m\) and \(\Delta f\) separate. A big number in a signal expression is not always the quantity the question asks for.

What you must be able to do

For Major Exam No. 1, make sure you can:

  • Recognize a conventional AM signal and its envelope.
  • Identify the carrier and both sidebands.
  • Connect the time-domain and frequency-domain views taught in class.
  • Find the AM bandwidth from the highest message frequency.
  • Explain instantaneous frequency in FM.
  • Find the peak frequency deviation \(\Delta f\).
  • Find the maximum and minimum instantaneous frequency.
  • Find the FM modulation index \(\beta\) for a single tone.
  • Tell NBFM from WBFM when \(\beta\) is clearly small or clearly large.
  • Use Carson's rule, as covered in Weeks 5–6.
  • Compare AM and FM using physical reasons and numbers.

Part 1 · Amplitude modulation

Read the amplitude, then find the sidebands

In conventional AM, a slowly changing amplitude multiplies a fast carrier. The envelope and the spectrum describe the same signal in two different ways.

A1 · Expression → physical meaning

Read an AM signal before calculating

What you should recognize

The conventional single-tone AM signal is \[ s_{\mathrm{AM}}(t)=A_c\left[1+\mu\cos(2\pi f_mt)\right]\cos(2\pi f_ct). \] The cosine outside the bracket is the fast carrier. The bracket is the amplitude that changes with the message. If the amplitude factor does not start with 1, factor it first. Then you can read \(\mu\).

Guided example

The amplitude factor \(4+\cos(2\pi f_mt)\) can be written as \[ 4\left[1+0.25\cos(2\pi f_mt)\right]. \] The coefficient \(1\) is a swing in volts. The ratio \(1/4=0.25\) has no unit. It is the modulation index \(\mu\).

Problem

An AM signal is

\[ s(t)=\left[6+1.5\cos(2\pi\cdot1200\,t)\right]\cos(2\pi\cdot240000\,t)~\text{V}, \]

where \(t\) is in seconds.

  1. Find \(f_c\) and \(f_m\).
  2. Find \(A_c\) and \(\mu\).
  3. Write the envelope. What changes physically in this signal?
  4. A student says “\(\mu=1.5\) and \(f_c=1200~\text{Hz}\).” Find the two mistakes.

Try it first

Underline the two cosine arguments. Circle the average envelope level. Rewrite the amplitude factor with a leading 1 inside the bracket before you name \(\mu\).

Hint / discussion

Which cosine is slow, and which is fast? In AM, the message changes the carrier's amplitude. It does not move the carrier frequency. Compare each argument with \(2\pi f\). Then divide the envelope swing by the average envelope level.

Show solution — A1
  1. Given The amplitude factor \(6+1.5\cos(2\pi\cdot1200\,t)~\text{V}\) and carrier \(\cos(2\pi\cdot240000\,t)\)
  2. Find \(f_c\), \(f_m\), \(A_c\), \(\mu\) and the envelope
  3. Solution Factor out the average level:
    \[ 6+1.5\cos(2\pi\cdot1200\,t)=6\left[1+\frac{1.5}{6}\cos(2\pi\cdot1200\,t)\right] \]
    Compare with \(A_c\left[1+\mu\cos(2\pi f_mt)\right]\cos(2\pi f_ct)\): \[ A_c=6~\text{V} \] \[ \begin{aligned} \mu &=\frac{1.5~\text{V}}{6~\text{V}} \\ &=0.25 \end{aligned} \] \[ \begin{aligned} f_m &=1200~\text{Hz} \\ &=1.2~\text{kHz} \end{aligned} \] \[ \begin{aligned} f_c &=240000~\text{Hz} \\ &=240~\text{kHz} \end{aligned} \]
  4. Result The envelope is \[ E(t)=\left[6+1.5\cos(2\pi\cdot1200\,t)\right]~\text{V}. \] It varies between \[ 6-1.5=4.5~\text{V} \] \[ 6+1.5=7.5~\text{V} \]
  5. What it means The amplitude changes. The carrier stays a \(240~\text{kHz}\) cosine. The envelope stays positive, so it follows the message. The student's \(1.5\) is a swing in volts, not \(\mu\). The \(1200~\text{Hz}\) is the message frequency, not the carrier frequency.
A2 · Spectrum → transmitter information

Work backward from three spectral lines

What you should recognize

A single-tone conventional AM signal has a carrier line and one pair of sidebands. Multiplying two cosines gives a sum and a difference frequency: \[ \cos A\cos B=\tfrac{1}{2}\left[\cos(A-B)+\cos(A+B)\right]. \] So the sidebands are at \(f_c\pm f_m\).

Guided example

Once you know where the carrier is, measure the distance from the carrier to one sideband. That distance is \(f_m\). The distance between the two sidebands is \(2f_m\).

Problem

The figure shows the positive-frequency spectrum of a conventional single-tone AM signal.

Three-line AM spectrum for reverse interpretation Positive-frequency stems at 448, 450 and 452 kilohertz. The center stem is taller. Heights need not be measured. 448450452f (kHz)
Line heights are qualitative. All three lines are present.
  1. Which line is the carrier, which is the LSB, and which is the USB?
  2. Find \(f_c\) and \(f_m\).
  3. Write the three line positions in terms of \(f_c\) and \(f_m\).
  4. Sketch the spectrum with the correct labels.
  5. A draft labels \(448~\text{kHz}\) as “USB” and \(452~\text{kHz}\) as “LSB.” Fix it.

Try it first

Mark the middle line and two equal distances from it. Then explain why the \(4~\text{kHz}\) gap between the sidebands is not \(f_m\).

Hint / discussion

The carrier sits exactly halfway between the two sidebands. “Lower” and “upper” describe the position relative to the carrier, not the height of the line. Expanding the AM product shows where the two offsets come from.

Show solution — A2
  1. Given spectral lines at \(448~\text{kHz}\), \(450~\text{kHz}\) and \(452~\text{kHz}\)
  2. Find the line names, \(f_c\) and \(f_m\)
  3. Solution The carrier is halfway between the sidebands, and the message frequency is half of their separation: \[ \begin{aligned} f_c &= \frac{f_{\mathrm{USB}}+f_{\mathrm{LSB}}}{2} \\ &= \frac{452+448}{2}~\text{kHz} \\ &= 450~\text{kHz} \end{aligned} \] \[ \begin{aligned} f_m &= \frac{f_{\mathrm{USB}}-f_{\mathrm{LSB}}}{2} \\ &= \frac{452-448}{2}~\text{kHz} \\ &= 2~\text{kHz}. \end{aligned} \] Check: \(452-450=2~\text{kHz}\).
  4. Result \[ \text{LSB}=f_c-f_m=448~\text{kHz} \] \[ \text{carrier}=f_c=450~\text{kHz} \] \[ \text{USB}=f_c+f_m=452~\text{kHz} \] Put these labels under the matching stems. The draft had the two sideband labels reversed.
  5. What it means The \(2~\text{kHz}\) message multiplies the \(450~\text{kHz}\) carrier and creates the difference and sum frequencies. The carrier line stays in the middle. Both sidebands carry the message. The carrier carries no new information. The span from the lower sideband to the upper sideband is \(452-448=4~\text{kHz}\).
A3 · Message support → channel fit

Can another audio tone fit?

What you should recognize

A message with highest frequency \(B_m\) gives sidebands out to \(f_c-B_m\) and \(f_c+B_m\). So conventional AM needs \[ B_{\mathrm{AM}}=2B_m. \] The modulation index changes the sizes of the components. It does not change where they are.

Guided example

A message that reaches \(3~\text{kHz}\) needs \(2\times3=6~\text{kHz}\) of channel width. This is true even if the message has many tones. The carrier line is in the center, so it does not make the signal wider.

Problem

A test channel passes \(732\) to \(740~\text{kHz}\). A conventional AM transmitter is centered at \(736~\text{kHz}\). It carries audio from \(0\) to \(3.4~\text{kHz}\).

  1. Does this signal fit in the channel?
  2. A technician adds a \(4.6~\text{kHz}\) tone. Does the signal still fit?
  3. The technician then cuts the modulation index \(\mu\) in half (still not zero). Will this make the signal fit again? Explain.
  4. What is the highest message frequency that fits at this carrier frequency?

Try it first

Draw the channel. Mark the two outer edges of the signal before and after the new tone. Keep amplitude and bandwidth separate in your mind.

Hint / discussion

How far can each sideband go from \(736~\text{kHz}\) before it reaches a channel edge? If you reduce \(\mu\) but keep the tone, the sideband gets smaller. It does not move closer to the carrier.

Show solution — A3
  1. Given channel \(732\text{–}740~\text{kHz}\); \(f_c=736~\text{kHz}\); \(B_m=3.4~\text{kHz}\), later \(4.6~\text{kHz}\)
  2. Find whether the signal fits before and after the new tone, the effect of changing \(\mu\), and the largest \(B_m\)
  3. Solution The channel is \(740-732=8~\text{kHz}\) wide. The AM bandwidth is \(B_{\mathrm{AM}}=2B_m\), and the outer edges are at \(f_c\pm B_m\). For \(B_m=3.4~\text{kHz}\): \[ \begin{aligned} B_{\mathrm{AM}} &=2B_m \\ &=2(3.4) \\ &=6.8~\text{kHz} \end{aligned} \] \[ \begin{aligned} 736-3.4 &=732.6~\text{kHz} \\ 736+3.4 &=739.4~\text{kHz} \end{aligned} \] For \(B_m=4.6~\text{kHz}\): \[ \begin{aligned} B_{\mathrm{AM}} &=2B_m \\ &=2(4.6) \\ &=9.2~\text{kHz} \end{aligned} \] \[ \begin{aligned} 736-4.6 &=731.4~\text{kHz} \\ 736+4.6 &=740.6~\text{kHz} \end{aligned} \] \[ \begin{aligned} B_{m,\max} &=\min(736-732,\ 740-736) \\ &=4~\text{kHz} \end{aligned} \]
  4. Result The original message fits (\(6.8<8~\text{kHz}\)). With the added tone it does not fit (\(9.2>8~\text{kHz}\)). The highest message frequency that fits is \(4~\text{kHz}\).
  5. What it means Cutting \(\mu\) in half does not help. The \(4.6~\text{kHz}\) offsets from the carrier stay the same. To fit, you must remove audio above \(4~\text{kHz}\) or use a wider channel. Removing audio loses part of the message, so reducing amplitude is not a fix for a bandwidth problem.

Part 2 · Frequency modulation

Understand the frequency swing before using the formula

AM changes the carrier amplitude. FM moves the instantaneous frequency above and below the center frequency. The deviation \(\Delta f\) tells you how far it moves. The message frequency \(f_m\) tells you how fast it repeats.

F1 · Message sign → instantaneous frequency

Follow the message above and below the center

What you should recognize

The carrier frequency \(f_c\) is the center. The instantaneous frequency \(f_i(t)\) is the frequency at one moment in time. With a positive sensitivity \(k_f\), \[ f_i(t)=f_c+k_fm(t). \] When \(m(t)\) is positive, \(f_i(t)\) is above \(f_c\). When \(m(t)\) is negative, it is below \(f_c\).

Guided example

Let \(k_f=4~\text{kHz/V}\). If \(m(t)=+0.5~\text{V}\), the frequency is \(+2~\text{kHz}\) above the center. If \(m(t)=-0.5~\text{V}\), it is \(2~\text{kHz}\) below the center. The sign tells the direction. The peak deviation \(\Delta f\) is still a positive number.

Problem

An FM signal has

\[ f_i(t)=90\times10^{6}~\text{Hz}+\left(8000~\text{Hz/V}\right)m(t). \]

The message is a cosine with peaks of \(\pm1.5~\text{V}\). The figure shows one period of the message.

One cycle of the message, not the RF waveform Cosine message starts at P, positive 1.5 volts; crosses zero at Q after a quarter period; reaches R, negative 1.5 volts, after a half period; returns to its peak after one period. +1.5 V0 V−1.5 VPQR0T/4T/23T/4Tt
m(t), in volts, over one period T. This is a message plot. It is not a spectrum or an FM envelope plot.
  1. What do \(f_c\), \(f_i(t)\) and \(\Delta f\) mean in this signal? Find \(f_c\) and \(\Delta f\).
  2. At points P, Q and R, is the instantaneous frequency above, at, or below the center? Find its value at each point.
  3. Why does the amplitude of the FM signal not follow this plot?

Try it first

Make three rows: message value → frequency offset → instantaneous frequency. Keep the center separate from the offset.

Hint / discussion

Multiply volts by Hz/V. Then add the result to the center frequency. The deviation \(\Delta f\) is the largest offset from the center. It is not the center frequency, and it is not the full swing from lowest to highest. A negative message does not give a negative RF frequency.

Show solution — F1
  1. Given \[ f_c=90~\text{MHz}, \] \[ k_f=8~\text{kHz/V}, \] \[ A_m=1.5~\text{V} \]
  2. Find \(f_i(t)\) at P, Q and R, and \(\Delta f\)
  3. Solution \[ f_i(t)-f_c=k_fm(t) \] \[ \begin{aligned} \Delta f &=k_fA_m \\ &=8\times1.5 \\ &=12~\text{kHz} \end{aligned} \] The offsets at P, Q and R are \[ \begin{aligned} \text{P: }(8~\text{kHz/V})(+1.5~\text{V}) &=+12~\text{kHz} \\ \text{Q: }(8~\text{kHz/V})(0~\text{V}) &=0 \\ \text{R: }(8~\text{kHz/V})(-1.5~\text{V}) &=-12~\text{kHz} \end{aligned} \]
  4. Result \[ \begin{aligned} \text{P: } f_i&=90.012~\text{MHz}\\ \text{Q: } f_i&=90.000~\text{MHz}\\ \text{R: } f_i&=89.988~\text{MHz} \end{aligned} \] \[ \Delta f=12~\text{kHz}. \]
  5. What it means The center is \(90~\text{MHz}\). The instantaneous frequency moves up and down around it, and \(\Delta f\) is the largest distance from the center. A positive message gives a frequency above the center, zero gives the center, and a negative message gives a frequency below it. In ideal FM the amplitude stays constant. The message changes how close together the cycles are, not how tall they are. The full swing from lowest to highest is \(2\Delta f=24~\text{kHz}\).
F2 · Measured extremes → center and excursion

Recover the settings, then predict an amplitude change

What you should recognize

For a zero-mean single-tone message, the highest and lowest frequencies are the same distance from the center: \[ \begin{aligned} f_{\max} &=f_c+\Delta f \\ f_{\min} &=f_c-\Delta f \end{aligned} \] The midpoint of the two gives \(f_c\). Half of the distance between them gives \(\Delta f\).

Guided example

Let \(f_c=102~\text{MHz}\) and \(\Delta f=10~\text{kHz}=0.010~\text{MHz}\). Then \[ \begin{aligned} f_{\max} &=102.010~\text{MHz} \\ f_{\min} &=101.990~\text{MHz} \end{aligned} \] Convert to the same unit before you add. This avoids a mistake by a factor of 1000.

Problem

A calibration log for a zero-mean cosine message shows \[ f_{\min}=83.992~\text{MHz},\qquad f_{\max}=84.008~\text{MHz} \]

  1. Find the center frequency \(f_c\).
  2. Find the peak deviation \(\Delta f\).
  3. The message amplitude is increased by \(50\%\). The values of \(k_f\), \(f_m\) and \(f_c\) stay the same. Find the new \(f_{\max}\) and \(f_{\min}\).
  4. What changes and what stays the same?

Try it first

Find the midpoint and half the range before you apply the amplitude increase. A \(50\%\) increase means multiply by \(1.5\), not by \(0.5\).

Hint / discussion

Use \(\Delta f=k_fA_m\). If \(A_m\) grows by a factor, \(\Delta f\) grows by the same factor. The center frequency does not change. The numbers are instantaneous frequencies, not exact edges of the spectrum.

Show solution — F2
  1. Given \[ f_{\min}=83.992~\text{MHz}, \] \[ f_{\max}=84.008~\text{MHz}, \] \[ A_m'=1.5A_m \]
  2. Find \(f_c\), \(\Delta f\), and the new \(f_{\max}'\) and \(f_{\min}'\)
  3. Solution \[ \begin{aligned} f_c &= \frac{f_{\max}+f_{\min}}{2} \\ &= \frac{84.008+83.992}{2} \\ &= 84.000~\text{MHz} \end{aligned} \] \[ \begin{aligned} \Delta f &= \frac{f_{\max}-f_{\min}}{2} \\ &= \frac{84.008-83.992}{2} \\ &= 0.008~\text{MHz} \\ &= 8~\text{kHz} \end{aligned} \] \[ \begin{aligned} \Delta f' &= 1.5\,\Delta f \\ &= 1.5(8) \\ &= 12~\text{kHz} \\ &= 0.012~\text{MHz} \end{aligned} \]
  4. Result \[ \begin{aligned} f_{\min}' &=84.000-0.012 \\ &=83.988~\text{MHz} \end{aligned} \] \[ \begin{aligned} f_{\max}' &=84.000+0.012 \\ &=84.012~\text{MHz} \end{aligned} \]
  5. What it means A larger message amplitude gives a wider frequency swing. The center stays at \(84~\text{MHz}\), and the swing repeats at the same rate. In ideal FM the amplitude of the RF signal stays constant. A bigger message does not make a bigger RF signal.
F3 · Same deviation, different index

Does a larger β always mean a larger frequency swing?

What you should recognize

For a single tone, the FM modulation index is \[ \beta=\frac{\Delta f}{f_m}. \] Both \(\Delta f\) and \(f_m\) are frequencies, so \(\beta\) has no unit. In the clear cases here, \(\beta\ll1\) is NBFM and \(\beta>1\) is WBFM. \(\beta\) can become larger just because \(f_m\) becomes smaller, even if the frequency swing stays the same.

Guided example

Let \(\Delta f=10~\text{Hz}\) and \(f_m=1000~\text{Hz}\). Then \[ \begin{aligned} \beta &=\frac{10}{1000} \\ &=0.01\ll1 \end{aligned} \] Compare \(\Delta f\) with the message frequency, not with the carrier frequency.

Problem

Two FM tests use the same carrier and the same peak deviation, \(\Delta f=0.40~\text{kHz}\). Test P has \(f_m=8~\text{kHz}\). Test Q has \(f_m=100~\text{Hz}\).

  1. Find \(\beta\) for each test.
  2. Is each test NBFM or WBFM? Explain your answer.
  3. Does test Q have a larger frequency swing than test P?
  4. If you only double the carrier frequency, does either answer in part b change?

Try it first

Write both message frequencies in kHz. Then compare the two values of \(\beta\) and the two values of \(\Delta f\) separately. They answer different questions.

Hint / discussion

\(100~\text{Hz}=0.10~\text{kHz}\). \(\Delta f\) is the same in both tests. Only \(f_m\) changes. Does \(f_c\) appear in \(\beta=\Delta f/f_m\)?

Show solution — F3
  1. Given \[ \Delta f=0.40~\text{kHz}, \] \[ f_{m,\mathrm{P}}=8~\text{kHz}, \] \[ f_{m,\mathrm{Q}}=0.10~\text{kHz} \]
  2. Find \[ \beta_{\mathrm{P}},\qquad \beta_{\mathrm{Q}} \]
  3. Solution \[ \begin{aligned} \beta_{\mathrm{P}} &=\frac{\Delta f}{f_{m,\mathrm{P}}} \\ &=\frac{0.40}{8} \\ &=0.05 \end{aligned} \] \[ \begin{aligned} \beta_{\mathrm{Q}} &=\frac{\Delta f}{f_{m,\mathrm{Q}}} \\ &=\frac{0.40}{0.10} \\ &=4 \end{aligned} \]
  4. Result Test P: \(\beta=0.05\ll1\), so it is NBFM. Test Q: \(\beta=4>1\), so it is WBFM. Both have the same peak deviation, \(400~\text{Hz}\), and the same full swing, \(800~\text{Hz}\).
  5. What it means \(\beta\) is larger in test Q because its message changes more slowly, not because its frequency moves farther. Doubling only \(f_c\) changes neither \(\Delta f\) nor \(f_m\), so both values of \(\beta\) and both answers stay the same. Also, \(\beta>1\) is not “overmodulation” as in AM.
F4 · Bandwidth estimate → decision

Find the missing term in the bandwidth estimate

What you should recognize

Carson's rule estimates the practical bandwidth of an FM signal: \[ B_{\mathrm{FM}}\approx2\left(\Delta f+B_m\right). \] For a single tone, \(B_m=f_m\). The estimate uses both the deviation and the message frequency. The frequency swing \(2\Delta f\) alone is not the full estimate.

Guided example

Let \(\Delta f=7~\text{kHz}\) and \(f_m=1~\text{kHz}\). Then \[ \begin{aligned} B_{\mathrm{FM}} &\approx2(\Delta f+f_m) \\ &=2(7+1) \\ &=16~\text{kHz} \end{aligned} \] The frequency swing is only \(2\Delta f=14~\text{kHz}\). These two numbers mean different things.

Problem

An engineer says: “Both FM tests have \(\Delta f=18~\text{kHz}\), so both need only \(36~\text{kHz}\).” Test A uses a \(2~\text{kHz}\) tone. Test B uses a \(6~\text{kHz}\) tone.

  1. Which term is missing from the engineer's estimate?
  2. Find the corrected bandwidth for each test.
  3. The available RF bandwidth is \(46~\text{kHz}\). Which test fits?
  4. For the \(6~\text{kHz}\) tone, find the largest \(\Delta f\) that fits in \(46~\text{kHz}\).

Try it first

Write one line for each test: same \(\Delta f\), different \(f_m\). Then rearrange Carson's rule to solve for \(\Delta f\) before you put in the available bandwidth.

Hint / discussion

The missing term is the message frequency. Start from \[ 2\left(\Delta f+f_m\right)\le B_{\text{available}}. \] Divide by 2, then subtract \(f_m\). This is a made-up planning exercise, not a regulatory limit.

Show solution — F4
  1. Given \[ \Delta f=18~\text{kHz}, \] \[ f_{m,\mathrm{A}}=2~\text{kHz}, \] \[ f_{m,\mathrm{B}}=6~\text{kHz}, \] \[ B_{\text{available}}=46~\text{kHz} \]
  2. Find \(B_{\mathrm{A}}\), \(B_{\mathrm{B}}\), and the largest \(\Delta f\) for the \(6~\text{kHz}\) tone
  3. Solution \[ \begin{aligned} B_{\mathrm{A}} &\approx2(\Delta f+f_{m,\mathrm{A}}) \\ &=2(18+2) \\ &=40~\text{kHz} \end{aligned} \] \[ \begin{aligned} B_{\mathrm{B}} &\approx2(\Delta f+f_{m,\mathrm{B}}) \\ &=2(18+6) \\ &=48~\text{kHz} \end{aligned} \] \[ \begin{aligned} \Delta f_{\max} &\le\frac{B_{\text{available}}}{2}-f_m \\ &=\frac{46}{2}-6 \\ &=17~\text{kHz} \end{aligned} \]
  4. Result Test A fits, with \(6~\text{kHz}\) to spare. Test B is \(2~\text{kHz}\) over the available bandwidth. For the \(6~\text{kHz}\) tone, the largest deviation that fits is \(17~\text{kHz}\), because \(2(17+6)=46~\text{kHz}\).
  5. What it means The same deviation does not give the same bandwidth. Carson's rule gives a practical band that holds about 98% of the signal power, as taught in class. It does not say there is zero energy outside the band. Lowering \(\Delta f\) to make the signal fit changes the FM design, and this calculation does not tell us what happens to noise performance.

Part 3 · Engineering synthesis

Choose using the requirements, not a slogan

Use this order: Claim → engineering evidence → numbers → tradeoff. First check that an option carries the audio you need and fits in the available bandwidth. Then explain its advantages and limits.

C1 · Two services, two decisions

Announcements in the morning, music in the evening

What you should recognize

Audio bandwidth and RF bandwidth are two different things. A receiver cannot bring back audio that the transmitter's filter already removed. In conventional AM, amplitude disturbances change the envelope, and the envelope carries the message. An FM limiter can remove amplitude variations before the frequency detector, if the signal is strong enough.

Guided example

A good claim has a condition. For example: “This option works because it passes the audio we need and its RF bandwidth is inside the allocation.” Only after that, talk about noise. Saying “FM is better” does not show that either condition is met.

Problem

A museum runs a local broadcast service. Both transmitter configurations below give good coverage and strong reception. Their settings are fixed. You cannot change the deviation or widen the audio filters. There are moderate amplitude disturbances nearby.

Available options — hypothetical engineering data
ConfigurationAudio supportOther information
Conventional AM\(0\text{–}3~\text{kHz}\)Normal modulation; envelope receiver.
FM\(0\text{–}9~\text{kHz}\)Peak deviation \(\Delta f=24~\text{kHz}\); limiter and frequency detector.

Morning: announcements need audio up to \(3~\text{kHz}\). The maximum allowed RF bandwidth is \(12~\text{kHz}\).

Evening: music needs audio up to \(9~\text{kHz}\). The maximum allowed RF bandwidth is \(80~\text{kHz}\). Resistance to amplitude disturbances is a priority.

Use the full audio support of each configuration.

  1. Find the RF bandwidth of each configuration.
  2. Which configuration do you recommend for the morning session?
  3. Which configuration do you recommend for the evening session?
  4. Give one tradeoff for each decision.

Try it first

Make a two-row table: Is the audio requirement met? Does the RF bandwidth fit the maximum allowed? How does it handle noise? Your two recommendations do not have to be the same.

Hint / discussion

Use \[ B_{\mathrm{AM}}=2B_m \] for AM and Carson's rule for FM. The two audio bandwidths are different on purpose. Compare each one with what the service needs. Do not claim a “fair” bandwidth ratio between AM and FM, or say that FM always has better audio quality.

Show solution — C1
  1. Given AM: \(B_m=3~\text{kHz}\). FM: \(B_m=9~\text{kHz}\) and \(\Delta f=24~\text{kHz}\). Maximum allowed RF bandwidth: morning \(12~\text{kHz}\), evening \(80~\text{kHz}\).
  2. Find \(B_{\mathrm{AM}}\), \(B_{\mathrm{FM}}\), and one choice for each session
  3. Solution \[ \begin{aligned} B_{\mathrm{AM}} &=2B_m \\ &=2(3) \\ &=6~\text{kHz} \end{aligned} \] \[ \begin{aligned} B_{\mathrm{FM}} &\approx2(\Delta f+B_m) \\ &=2(24+9) \\ &=66~\text{kHz} \end{aligned} \] An option must meet both the audio requirement and the maximum allowed RF bandwidth.
  4. Result Morning: choose AM. Its \(3~\text{kHz}\) audio is enough for announcements, and \(6<12~\text{kHz}\). FM needs \(66>12~\text{kHz}\), so it does not fit. Evening: choose FM. Its \(9~\text{kHz}\) audio is enough for music, and \(66<80~\text{kHz}\), with \(14~\text{kHz}\) to spare. The AM filter stops at \(3~\text{kHz}\), so it cannot carry audio up to \(9~\text{kHz}\).
  5. What it means Morning AM saves spectrum and needs a simpler receiver, but amplitude disturbances still affect it. Evening FM uses much more RF bandwidth and a more complex receiver. In return, the limiter removes amplitude variations, and the message stays in the instantaneous frequency. FM does not remove all noise, and we do not claim a numerical noise improvement here.

A complete evening claim: “Choose the fixed FM configuration. It passes audio up to \(9~\text{kHz}\), and its \(66~\text{kHz}\) Carson estimate fits within the \(80~\text{kHz}\) maximum allowed RF bandwidth. Its limiter helps against the moderate amplitude disturbances. The tradeoff is a wider RF channel and a more complex receiver.”

Independent check · 18 minutes · Calculator permitted

Ready for Major Exam No. 1?

Close the guided solutions and work without hints. Use seconds in signal models and show units. Spend about 3 minutes on T1, 5 on T2, 6 on T3 and 4 on T4. These four questions use new numbers. Open the answers only when you finish.

T1 — Interpretation

An ideal single-tone AM signal has nonzero spectral lines at \[ 1.598~\text{MHz},\qquad 1.600~\text{MHz},\qquad 1.602~\text{MHz}. \] A report calls it “DSB-SC” and calls the \(1.600~\text{MHz}\) line the “message.”

  1. Correct the report.
  2. What is the role of the \(1.600~\text{MHz}\) line?
  3. Find the message frequency.

T2 — Modeling from calibration data

An FM device has a linear sensitivity. The table shows its calibration.

Message voltageInstantaneous frequency
\(-0.8~\text{V}\)\(93.994~\text{MHz}\)
\(0~\text{V}\)\(94.000~\text{MHz}\)
\(+0.8~\text{V}\)\(94.006~\text{MHz}\)
  1. Find \(f_c\) and \(k_f\).
  2. The message is \[ m(t)=(0.8~\text{V})\sin(2\pi\cdot1500\,t). \] Write \(f_i(t)\) in Hz.

T3 — Quantitative AM and FM

AM: The message goes from zero up to an unknown highest frequency. The lower sideband covers \(620\text{–}624~\text{kHz}\) and the upper sideband covers \(624\text{–}628~\text{kHz}\).

  1. Find the RF bandwidth.
  2. Find the highest message frequency.

FM: A separate FM signal uses a \(3~\text{kHz}\) cosine message. Its extreme instantaneous frequencies are \[ f_{\min}=104.991~\text{MHz},\qquad f_{\max}=105.009~\text{MHz}. \]

  1. Find \(\Delta f\) and \(\beta\).
  2. Is it NBFM or WBFM?
  3. Find the Carson bandwidth.
  4. Why is the swing \(f_{\max}-f_{\min}\) not the Carson bandwidth?

T4 — Engineering reasoning

A local announcement service needs audio up to \(3~\text{kHz}\). The maximum allowed RF bandwidth is \(20~\text{kHz}\). Choose between the AM and FM configurations from T3. Their audio and deviation settings are fixed. Both give good coverage. Moderate amplitude disturbances nearby favor FM, if it fits.

  1. A colleague suggests doubling the FM carrier frequency so that it fits. Will this work? Explain.
  2. Which configuration do you recommend? Use your T3 results.
  3. Give one tradeoff of your choice.
Show model answers — self-test T1–T4

T1 — A transmitted carrier, not a baseband message

  1. Given lines at \(1.598\), \(1.600\) and \(1.602~\text{MHz}\); ideal single-tone AM
  2. Find the correct label, the role of the middle line, and \(f_m\)
  3. Solution Conventional AM (DSB-TC) keeps a line at \(f_c\). Ideal DSB-SC has no carrier line. The sidebands are at \(f_c\pm f_m\), so \[ \begin{aligned} f_m &= 1.602-1.600 \\ &= 0.002~\text{MHz} \\ &= 2~\text{kHz}. \end{aligned} \]
  4. Result The signal is DSB-TC. The \(1.600~\text{MHz}\) line is the carrier. The message frequency is \(2~\text{kHz}\).
  5. What it means The \(2~\text{kHz}\) message shows up as sidebands around the carrier, not as the middle RF line. A nonzero carrier line does not match the “suppressed carrier” label.

T2 — Find the sensitivity before building the model

  1. Given A linear calibration: \(0~\text{V}\) gives \(94~\text{MHz}\), and \(+0.8~\text{V}\) gives \(94.006~\text{MHz}\). The message amplitude is \(0.8~\text{V}\).
  2. Find \(f_c\), \(k_f\) and \(f_i(t)\)
  3. Solution Since \(f_i=f_c+k_fm\), \(k_f\) is the change in frequency divided by the change in voltage: \[ \begin{aligned} k_f &= \frac{(94.006-94.000)~\text{MHz}}{0.8~\text{V}} \\ &= \frac{6000~\text{Hz}}{0.8~\text{V}} \\ &= 7500~\text{Hz/V}. \end{aligned} \] Check with the two ends of the table: \[ \begin{aligned} k_f &=\frac{(94.006-93.994)~\text{MHz}}{0.8-(-0.8)~\text{V}} \\ &=\frac{12000~\text{Hz}}{1.6~\text{V}} \\ &=7500~\text{Hz/V} \end{aligned} \]
  4. Result \[ f_c=94\times10^{6}~\text{Hz} \] \[ f_i(t)=\left[94\times10^{6}+6000\sin(2\pi\cdot1500\,t)\right]~\text{Hz}. \]
  5. What it means At \(t=0\) the sine message is zero, so the frequency is at the center. The positive and negative peaks match the calibration extremes. The phase of the message changes when the swing happens, not the sensitivity.

T3 — Separate the RF span from the frequency swing

  1. Given AM: lower sideband \(620\text{–}624~\text{kHz}\) and upper sideband \(624\text{–}628~\text{kHz}\). FM: \[ f_{\min}=104.991~\text{MHz}, \] \[ f_{\max}=105.009~\text{MHz}, \] \[ f_m=3~\text{kHz} \]
  2. Find \(B_{\mathrm{AM}}\), \(B_m\), \(\Delta f\), \(\beta\) and \(B_{\mathrm{FM}}\)
  3. Solution \[ \begin{aligned} B_{\mathrm{AM}} &=628-620 \\ &=8~\text{kHz} \end{aligned} \] \[ \begin{aligned} B_m &=\frac{B_{\mathrm{AM}}}{2} \\ &=\frac{8}{2} \\ &=4~\text{kHz} \end{aligned} \] \[ \begin{aligned} \Delta f &= \frac{f_{\max}-f_{\min}}{2} \\ &= \frac{105.009-104.991}{2} \\ &= 0.009~\text{MHz} \\ &= 9~\text{kHz} \end{aligned} \] \[ \begin{aligned} \beta &=\frac{\Delta f}{f_m} \\ &=\frac{9}{3} \\ &=3 \end{aligned} \] \[ \begin{aligned} B_{\mathrm{FM}}\approx2\left(\Delta f+f_m\right) &= 2(9+3) \\ &= 24~\text{kHz}. \end{aligned} \]
  4. Result AM: RF bandwidth \(8~\text{kHz}\), highest message frequency \(4~\text{kHz}\). FM: \(\Delta f=9~\text{kHz}\), \(\beta=3\), so it is WBFM, with bandwidth about \(24~\text{kHz}\).
  5. What it means The full frequency swing is \(2\Delta f=18~\text{kHz}\). Carson's \(24~\text{kHz}\) is larger because it also includes the message frequency. The extreme frequencies are not sharp edges of the spectrum.

T4 — A different carrier does not fix the bandwidth

  1. Given audio up to \(3~\text{kHz}\); maximum allowed RF bandwidth \(20~\text{kHz}\); the configurations from T3
  2. Find Whether changing only the FM carrier helps, and which configuration is allowed.
  3. Solution Neither bandwidth formula contains \(f_c\): \[ B_{\mathrm{AM}}=2B_m \] \[ B_{\mathrm{FM}}\approx2\left(\Delta f+f_m\right). \] AM: \(8\le20~\text{kHz}\), and its message goes up to \(4~\text{kHz}\ge3~\text{kHz}\). FM: \(24>20~\text{kHz}\). Doubling \(f_c\) leaves \[ \begin{aligned} B_{\mathrm{FM}} &\approx2(9+3) \\ &=24~\text{kHz} \end{aligned} \] unchanged.
  4. Result Doubling the carrier does not help. Choose the fixed AM configuration. If FM's noise behavior were essential, the team would need a larger allocation or a new FM setting.
  5. What it means AM meets both the audio and bandwidth requirements, but its envelope is still sensitive to amplitude disturbances. The FM option has a limiter but does not fit the maximum allowed RF bandwidth. A noise advantage does not make up for an option that does not fit.