Reference: https://en.wikipedia.org/wiki/Propagation_constant#Phase_constant

Let's say I have a propagation constant ($\gamma$), where $jk_z = \gamma = \alpha +j \beta$. Why does the attenuation constant have no role in determining the wavelength, especially in a waveguide where TE/TM modes exists peacefully?


  1. According to the reference we can write wavelength as $\lambda = \frac{2 \pi}{Im\{ \gamma \}}$ (see reference)
  2. Constant equation in a waveguide is $\gamma ^2 = h_{cut}^2 - k^2$

No Problem in TEM Mode because $h_{cut} = 0$, hence, $\gamma = jk$ and $im \{ \gamma \} = im\{jk\} = k$. Conclusion: $\lambda = \frac{2 \pi}{k}$

Problem in TE/TM Mode because $h_{cut} \ne 0$, hence, $\gamma = \sqrt{ h_{cut}^2 - k^2}$ and $im\{\gamma \} = Im\{ \sqrt{ h_{cut}^2 - k^2} \}$. Conclusion: $\lambda = \frac{2 \pi}{ Im\{ \sqrt{ h_{cut}^2 - k^2} \} }$ and sources I am finding define the phase constant ($\beta$) as $\beta = Im\{ \sqrt{ h_{cut}^2 - k^2} \}$

So why is the attenuation constant being ignored, even in wave guides?


1 Answer 1


I think your question should be asked in a different way, and the answer to that simpler question will also answer your question implicitly, namely: what is the frequency of a damped oscillation, such as $f(t)=e^{-\alpha t } sin(\omega_0 t)$?

First of all, this $f(t)$ is an exponentially unbounded signal and has no Fourier transform, so one cannot really analyze it over the full time axis but it can be analyzed for $\alpha >0 $ as a transient from $t=0$ to $t\to \infty$,

$$F(\omega) = \mathcal F [f] = \int_0^{\infty}e^{-\mathfrak j \omega t}e^{-\alpha t } sin(\omega_0 t)dt\\ =\frac{\omega_0}{(\alpha +\mathfrak j \omega)^2+\omega_0^2}$$ This $|F(\omega)|$ has frequency content everywhere but where is its peak? Obviously the peak is where the denominator's modulus $|(\alpha +\mathfrak j \omega)^2+\omega_0^2|$ is the smallest, that is we have to find the minimum of $(\alpha^2 + \omega_0^2 - \omega^2)^2+2\alpha^2\omega^2$. Simple differentiation shows that happens at the frequency $\omega = \hat \omega$ where $$\hat \omega = \sqrt{\omega_0^2 - \alpha^2}.$$ If $\omega_0 <\alpha$ then the oscillation is so heavily damped that there is no clear frequency at which we can meaningfully say that it oscillates. If $\omega_0 >> \alpha$ then $\hat \omega \approx \omega _0\left(1-\frac{\alpha}{2\omega_0}\right)$, and you see that indeed the peak is dependent on the attenuation, but its relative importance decreases with their ratio and in a waveguide with metallic walls this is the normal propagating situation.

  • $\begingroup$ Thank you. English is not my first language and this is a difficult subject $\endgroup$ Jan 17 at 7:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.