Skip to main content

Realize 3rd order Butterworth TF with R2 = 2 Ω

Short answer

With unequal terminations (R₁=1 Ω, R₂=2 Ω), dc transmission |t(0)|²=4R₁R₂/(R₁+R₂)²=8/9. From T(s)=1/(s³+2s²+2s+1) compute ρ(s) and Z₁=R₁(1+ρ)/(1−ρ); continued-fraction expansion yields: series L₁=3.261 H, shunt C₂=0.7789 F, series L₃=1.181 H, load R₂=2 Ω (ωc=1 rad/s).

Realize the third order Butterworth lowpass transfer function T(s) = 1/(s³+2s²+2s+1) in the form of resistively terminated LC ladder with R1 = 1Ω and R2 = 2Ω.

Method. Find t(s)t(s), then ∣ρ∣2=1−∣t∣2|\rho|^2 = 1-|t|^2, then Z1=R11+ρ1−ρZ_1 = R_1\dfrac{1+\rho}{1-\rho}, then expand Z1Z_1 by continued fraction. Here t(s)=2R1/R2 V2/V1t(s) = 2\sqrt{R_1/R_2}\,V_2/V_1, and the given T(s)=1/(s3+2s2+2s+1)T(s) = 1/(s^3+2s^2+2s+1) fixes the shape; its dc level is set by the terminations.

Step 1: Transmission coefficient. Third order Butterworth (Table 1): B(s)=(s+1)(s2+s+1)=s3+2s2+2s+1B(s) = (s+1)(s^2+s+1) = s^3+2s^2+2s+1. At dc the inductors are shorts and capacitors open, so the ladder reduces to a divider: V2/V1=R2R1+R2=23=0.6667V_2/V_1 = \frac{R_2}{R_1+R_2} = \frac{2}{3} = 0.6667. Hence the largest possible dc transmission is

∣t(0)∣2=4R1R2(R1+R2)2=4(1)(2)(1+2)2=89=0.8889|t(0)|^2 = \frac{4R_1R_2}{(R_1+R_2)^2} = \frac{4(1)(2)}{(1+2)^2} = \frac{8}{9} = 0.8889

so ∣t(jω)∣2=0.88891+ω6|t(j\omega)|^2 = \dfrac{0.8889}{1+\omega^6}, i.e. t(s)=0.9428s3+2s2+2s+1t(s) = \dfrac{0.9428}{s^3+2s^2+2s+1}.

Step 2: Reflection coefficient.

∣ρ(jω)∣2=1−0.88891+ω6=0.1111+ω61+ω6  ⇒  ρ(s)ρ(−s)=0.1111−s61−s6|\rho(j\omega)|^2 = 1 - \frac{0.8889}{1+\omega^6} = \frac{0.1111 + \omega^6}{1+\omega^6} \;\Rightarrow\; \rho(s)\rho(-s) = \frac{0.1111 - s^6}{1 - s^6}

The zeros satisfy s6=0.1111s^6 = 0.1111, i.e. they lie on a circle of radius a=(0.1111)1/6=0.6934a = (0.1111)^{1/6} = 0.6934 at the Butterworth angles. Choosing the left-half-plane zeros (like the Butterworth poles, scaled by aa):

ρ(s)=(s+a)(s2+as+a2)B(s)=s3+1.3867s2+0.9615s+0.3333s3+2s2+2s+1\rho(s) = \frac{(s+a)(s^2+as+a^2)}{B(s)} = \frac{s^3 + 1.3867 s^2 + 0.9615 s + 0.3333}{s^3+2s^2+2s+1}

Step 3: Input impedance (R1=1R_1 = 1):

Z1(s)=R11+ρ(s)1−ρ(s)=2s3+3.3867s2+2.9615s+1.33330.6133s2+1.0385s+0.6667Z_1(s) = R_1\frac{1+\rho(s)}{1-\rho(s)} = \frac{2s^3+3.3867s^2+2.9615s+1.3333}{0.6133s^2+1.0385s+0.6667}

Step 4: Continued fraction expansion (computed by repeated long division):

Z1(s)=3.261 s+10.7789 s+11.181 s+2Z_1(s) = 3.261\,s + \cfrac{1}{0.7789\,s + \cfrac{1}{1.181\,s + 2}}

So series L1=3.261L_1 = 3.261 H, shunt C2=0.7789C_2 = 0.7789 F, series L3=1.181L_3 = 1.181 H, and the remainder is the load R2=2 ΩR_2 = 2\ \Omega, as required (this choice of ρ\rho sign gives the correct R2R_2; the other sign ends in 1/2 Ω1/2\ \Omega and is rejected).

Step 5: Circuit.

  R1=1  L1=3.261 H     L3=1.181 H
 o-/\/\---UUU----+----UUU---+
 |               |          |
(~) V1     C2=0.7789 F   R2=2
 |               |          |
 o---------------+----------+

Check: analysing this ladder gives V2/V1=0.6667s3+2s2+2s+1V_2/V_1 = \dfrac{0.6667}{s^3+2s^2+2s+1}, the required Butterworth response.

Answer: R1=1 ΩR_1 = 1\ \Omega, L1=3.261L_1 = 3.261 H, C2=0.7789C_2 = 0.7789 F, L3=1.181L_3 = 1.181 H, R2=2 ΩR_2 = 2\ \Omega (normalized to ωc=1\omega_c = 1 rad/s).

✓ Answer checked by Claude Sonnet 5.5

All Filter Design chapter 5 questions →