Short answer: usually no for a classic cantenna, but yes in specific variants.
Why standard cantennas use parallel sides
A conventional cantenna is a cylindrical waveguide designed to support the dominant TE₁₁ mode cleanly. Parallel sides give you:
- Predictable cutoff frequency
- Stable impedance
- Clean mode structure (minimal higher-order modes)
- Simple, repeatable construction and tuning
For narrowband work (Wi-Fi, 1420 MHz experiments, etc.), this is exactly what you want.
What happens if the sides are not parallel?
1. Tapered / conical can (flared walls)
This is the main case where non-parallel sides can help.
Potential advantages
- Smoother impedance transition → slightly wider bandwidth
- Reduced reflections compared to an abrupt open end
- Can act as a crude horn, improving illumination of free space
Trade-offs
- Harder to design accurately
- Cutoff frequency varies along the length
- Can excite unwanted modes if the taper is too steep
- Gains are usually modest unless carefully engineered
This starts to become a horn antenna, not really a cantenna anymore.
2. Irregular or asymmetric walls
Generally disadvantageous:
- Mode distortion
- Increased sidelobes
- Poor polarization purity
- Unpredictable impedance and radiation pattern
Unless you are deliberately experimenting, this is almost always worse.
Practical guidance
- If your goal is maximum simplicity, efficiency, and predictability → keep the sides parallel.
- If your goal is slightly more bandwidth or smoother matching, a gentle conical flare at the open end can help—but at that point, a proper horn is usually a better choice.
- For radio astronomy at 1420 MHz, the gains from non-parallel sides are typically not worth the added complexity unless you’re doing controlled experiments.