3
\$\begingroup\$

I want my board to have the ability to opt for either an onboard antenna or an external one, with a 0R chip resistor, naturally.

This is my initial setup:

enter image description here

But then I am told that this is flawed and better go with this one:

enter image description here

You see, 2 & 3 should connect together and with 2 and if I had to go for onboard antenna, I need to solder the 0R at 0R in the picture, and at the same time, a capacitor in the CAP footprint shown in the pic, in order to avoid interference. You see if I keep the seating soldered on the antenna track at all time, like picture 1, the antenna is actually emitting, in other words, I will have two competing feeds (I hope the terminologies are correct), and I have to say that I agree.

Problem is, is the 2nd pic overthinking? Isn't two "de facto" antennas better than one?

\$\endgroup\$
  • \$\begingroup\$ What's wrong with using a physical switch, ro PIN diodes to switch to the ext. antenna whenever present, then default to the PCB antenna at all other times? \$\endgroup\$ – Robherc KV5ROB Apr 19 '17 at 2:52
  • \$\begingroup\$ Two concurrently radiating antennas will likely create nulls in the radiation pattern, and may create high VSWR feeds for the Transmitter. High VSWR means low power available to be radiated. \$\endgroup\$ – analogsystemsrf Apr 19 '17 at 2:53
  • \$\begingroup\$ @RobhercKV5ROB Nothing more than cost. \$\endgroup\$ – Matt Cox Apr 19 '17 at 3:14
  • \$\begingroup\$ @analogsystemsrf ha, you will be surprised to know that some of the most prestigious companies gave demo board with pic1 design on it, I kid you not! \$\endgroup\$ – Matt Cox Apr 19 '17 at 3:15
  • \$\begingroup\$ @analogsystemsrf not to mention the possibility of damage to the transmitter if it wasn't designed to handle the reflected energy from all the reflections...could result in the acrid smell of overheated silicon....... \$\endgroup\$ – Robherc KV5ROB Apr 19 '17 at 3:30
3
\$\begingroup\$

(1) is flawed, you would be better to go with (2).

Isn't two "de facto" antennas better than one?

Depending on your power amplifier specifications, and the details of antenna match and line lengths, in the best case the total power emitted by both antennae will be less than from a single one due to the impedance mismatch, with nulls occurring at various frequencies and directions due to mutual interference, and in the worst case your power amplifier will be damaged.

A setup designed to use two antennae will usually be better than one. However they key here is designed to, not just another antenna slapped in parallel.

... is the 2nd pic overthinking?

No. The problem is that RF is not DC. (1) would be fine at DC. The problem is that even when you use the 0ohm jumper and socket to connect only one antenna, there is a stub of line hanging off the transmission line to the connected antenna.

In the case of the socket, while the stub of extra track from the socket to the 0ohm resistor pad is fairly short, it is not zero. How significant it is will depend on the frequency of operation. A \$\frac{\lambda}{20}\$ stub will likely have a small effect, as it behaves like a small capacitance. It may still make your antenna deviate from specifications, and in a more frequency dependant way than expected. A \$\frac{\lambda}{4}\$ stub will be catastrophic, behaving like a short circuit at the socket.

In the case of the on board antenna, the residual length of the socket will behave like a stub of line.

Even if the effect of either stub is minimal, is the correct design of (2) so much more bother?

\$\endgroup\$
0
\$\begingroup\$

Design #1 is better if you can disconnect the external antenna - because you only need to solder or desolder one link.

Design #2 is better if you can not disconnect the external antenna - but you always have to make one link and break the other link.

You have more options! For example, you can deliberately mismatch the on-board (PCB) antenna by making it smaller than needed for a proper match or by driving your matched (50Ω) PCB antenna through a 220Ω resistor instead of the 0Ω link.

Assuming that your RF IC can handle the 270Ω mismatch, you will see reduced range due to the mismatch when using the on-board antenna. (Roughly 3x reduction in range for a 10dB mismatch loss)

However, if you now plug in your 50Ω external antenna in parallel with the 270Ω, over 80% of the power will flow to the external antenna. This essentially implements low-cost "automatic" antenna switching without needing to solder/desolder any links!

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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