LC Oscillations, Quantitatively — Question 5

Section 31-3 · Physics pages 5020–5021

Question 5 Component tolerances combine through the square root of the LC product A nominal LC oscillator uses positive L0 and C0. Actual values are L = L0(1 + ϵL) and C = C0(1 + ϵC), with independent bounds |ϵL| ≤.02 and |ϵC| ≤.03. LC ω Schematic; labels, not drawing lengths, specify the data. (a) Derive the first-order fractional frequency error. (b) Find exact worst-case high and low frequency ratios.
Show solutionHide solution
Question 5 - Solution Component tolerances combine through the square root of the LC product LC ω (a) Method and calculation ω/ω0 = [(1 + ϵL)(1 + ϵC)]−1/2. ∆ω/ω0 ≃−1 2(ϵL + ϵC) . (b) Interpretation and further calculation ωmax/ω0 = (.98 · .97)−1/2, ωmin/ω0 = (1.02 · 1.03)−1/2. ωmax/ω0 = 1.02565, ωmin/ω0 = .975621 . Physical check Worst-case bounded errors add; a root-sum-square rule would require a statistical error model.

Question 5

Component tolerances combine through the square root of the LC product

A nominal LC oscillator uses positive L0 and C0. Actual values are L=L0(1+ϵL)L=L_0(1+\epsilon_L) and C=C0(1+ϵC)C=C_0(1+\epsilon_C), with independent bounds |ϵL|≤.02|\epsilon_L|\le.02 and |ϵC|≤.03|\epsilon_C|\le.03.

Diagram shown in the original SVG worksheet.

  1. Derive the first-order fractional frequency error.

  2. Find exact worst-case high and low frequency ratios.

Question 5 - Solution

Component tolerances combine through the square root of the LC product

Diagram shown in the original SVG worksheet.

(a) Method and calculation

ω/ω0=[(1+ϵL)(1+ϵC)]−1/2.\omega/\omega_0=[(1+\epsilon_L)(1+\epsilon_C)]^{-1/2}.Δω/ω0≃−12(ϵL+ϵC).\boxed{\Delta\omega/\omega_0\simeq-\tfrac12(\epsilon_L+\epsilon_C)}.

(b) Interpretation and further calculation

ωmax/ω0=(.98⋅.97)−1/2,ωmin/ω0=(1.02⋅1.03)−1/2.\omega_{\max}/\omega_0=(.98\cdot.97)^{-1/2},\quad\omega_{\min}/\omega_0=(1.02\cdot1.03)^{-1/2}.ωmax/ω0=1.02565,ωmin/ω0=.975621.\boxed{\omega_{\max}/\omega_0=1.02565,\quad\omega_{\min}/\omega_0=.975621}.

Physical check

Worst-case bounded errors add; a root-sum-square rule would require a statistical error model.