Printed dipole vs ordinary (3D) dipole

Question

In order to make a comparison between an ordinary (3D) half - wavelength dipole and its printed (planar version), I've done a simulation in both cases. Both were designed to work at 3GHz (100mm wavelength). Input port is a voltage source with 75Ohm of impedance. To avoid dielectric effects, I've not considered any dielectric substrate for the printed dipole.

1. The 3D dipole has been built with:

• Length 45mm (a bit lower than half-wavelength. That's a choice the result of the simulator optimization).

• Radius 1mm (much lower than the Length)

• Gap1 1mm (much lower than the Length)

The results are:

S11 (reflection coefficient). It's minimum at 3GHz. So it's fine.

Radiation pattern cuts (at 3GHz)

2. The printed dipole has the same size values (Length 45mm, Width 1mm, Gap 1mm) of the 3D one, except for the fact it's a strip.

The results are:

S11 (reflection coefficient).

Radiation pattern cuts (at 3GHz).

---

In conclusion:

• same resonance frequency (for a fixed length)
• same matching at the resonance frequence
• same pattern shape and same directivity values

(Only very very very small differences are present in the previous graphs)

but a printed dipole has lots of advantages, being lightweight, easily embeddable with active devices in PCB, and flexible to make flexible pcb.

I can't believe there isn't at least one disadvantage in choosing the printed version. I would have expected a big variation in radiation pattern, since the 3d dipole has the same shape if seen from each direction on the transverse plane, while a printed dipole is a strip if observed frontally, and a wire if observed laterally. But the simulations say I'm wrong: the printed dipole is as omnidirectional (and with same directivity value) as the 3D dipole.

Answer 1

The diameter and cross-section of the dipole element do have some effect on bandwidth (the thicker the element, the broader the bandwidth) but very little effect on directivity. This is because the cross-sectional dimensions (1 mm) are still small compared to a wavelength at 3 GHz (100 mm).

However, the printed version isn't going to be floating in space, but supported by a PCB dielectric, which will have a significant effect. The dipole length for resonance will be shorter, depending on the dielectric constant of the material, and there will be some decrease in efficiency as well.