Skip to main content

Single tone FM in Bessel coefficients; practical bandwidth

Find the time domain expression for single tone FM modulated wave, in terms of Bessel coefficients. Derive the expression for estimating practical bandwidth of a FM signal.

Single-tone FM in terms of Bessel coefficients

For m(t)=Amcos⁡2πfmtm(t) = A_m\cos 2\pi f_m t:

s(t)=Accos⁡[2πfct+βsin⁡2πfmt]=Re{Ac ej2πfct ejβsin⁡2πfmt}s(t) = A_c\cos\left[2\pi f_c t + \beta\sin 2\pi f_m t\right] = \text{Re}\left\{A_c\, e^{j2\pi f_c t}\, e^{j\beta\sin 2\pi f_m t}\right\}

The term ejβsin⁡2πfmte^{j\beta\sin 2\pi f_m t} is periodic with period 1/fm1/f_m, so it can be written as a complex Fourier series:

ejβsin⁡2πfmt=∑n=−∞∞cn ej2πnfmte^{j\beta\sin 2\pi f_m t} = \sum_{n=-\infty}^{\infty} c_n\, e^{j2\pi n f_m t}

with coefficients (put x=2πfmtx = 2\pi f_m t)

cn=12π∫−ππej(βsin⁡x−nx) dx=Jn(β)c_n = \frac{1}{2\pi}\int_{-\pi}^{\pi} e^{j(\beta\sin x - nx)}\,dx = J_n(\beta)

which is the Bessel function of the first kind of order nn. Substituting:

s(t)=Ac∑n=−∞∞Jn(β)cos⁡[2π(fc+nfm)t]s(t) = A_c\sum_{n=-\infty}^{\infty} J_n(\beta)\cos\left[2\pi(f_c + n f_m)t\right]

So the FM wave has a carrier component AcJ0(β)A_cJ_0(\beta) and infinitely many side frequencies at fc±nfmf_c \pm nf_m with amplitudes AcJn(β)A_cJ_n(\beta), using J−n(β)=(−1)nJn(β)J_{-n}(\beta) = (-1)^n J_n(\beta).

Properties: ∑nJn2(β)=1\sum_n J_n^2(\beta) = 1, so total power =Ac2/2= A_c^2/2, independent of β\beta. The carrier amplitude J0(β)J_0(\beta) falls as β\beta increases and is zero at β=2.405,5.52,…\beta = 2.405, 5.52, \ldots

Practical bandwidth of FM

Although the spectrum is infinite, Jn(β)J_n(\beta) becomes very small when n>βn > \beta. For a fixed β\beta, Jn(β)J_n(\beta) decreases rapidly once nn exceeds about β+1\beta + 1. Keeping the significant sidebands up to nmaxn_{max}:

B=2 nmax fmB = 2\, n_{max}\, f_m

Carson's rule: taking nmax=β+1n_{max} = \beta + 1 (which keeps about 98% of the power):

BT=2(β+1)fm=2(Δf+fm)B_T = 2(\beta + 1)f_m = 2(\Delta f + f_m)

Two limiting cases support this:

  • β≫1\beta \gg 1 (WBFM): BT≈2βfm=2ΔfB_T \approx 2\beta f_m = 2\Delta f (the carrier sweeps over ±Δf\pm\Delta f).
  • β≪1\beta \ll 1 (NBFM): only J0J_0 and J1J_1 are significant, so BT≈2fmB_T \approx 2f_m.

1% rule (universal curve): keep all sidebands with ∣Jn(β)∣>0.01|J_n(\beta)| > 0.01; this gives a slightly larger bandwidth, e.g. for β=5\beta = 5, nmax=8n_{max} = 8 and B=16fmB = 16 f_m versus Carson's 12fm12 f_m.

For a general message with highest frequency WW, use the deviation ratio D=Δf/WD = \Delta f/W: BT=2(Δf+W)=2W(D+1)B_T = 2(\Delta f + W) = 2W(D+1).

✓ Answer checked by Claude Opus 5.5 (high effort)

All Communication Systems chapter 3 questions →