I intend for this to be a general/conceptual question that could apply to any wireless communication hardware, however I will use a specific example to talk about.

Let's consider the nRF24L01. (Datasheet).

nRF24L01 module
(Image from makerlab-electronics.com)

As shown above, the standard module uses a PCB antenna.

There is also a version that uses a screw-on omnidirecitonal antenna (AKA "whip"):

nRF24l01 with antenna
(Image from ebay.com)

Let's say that a pair of PCB-antenna transceivers give a usable range of 10 meters.

Let's also assume a pair of whip antenna transceivers exhibit better range. (The common consensus I've read is that there is a noticeable improvement. Let's say it results in a range of 20 meters).

My question is about having a mixed pair of antennas for half-duplex (single direction) communication. What if one were the PCB antenna, and the other were the whip antenna... Which transceiver (transmitter vs receiver) should use which antenna? Would it even matter?

I'm interested from both a theoretical and practical standpoint, as they are often different. One extra variable I can think of is the orientation of the antennas.

I'm also open to alternate title suggestions from someone who knows better terminology.

  • \$\begingroup\$ BTW, the stick antenna is usually referenced as a whip antenna (although if you say stick, they'll know what you're talking about). \$\endgroup\$ – isdi Oct 18 '18 at 17:55

If these were the only two devices in the world, and if the transmitter were ideal it would not matter.

However, neither is the case.

  • Using the "better" antenna on the transmitter would help make the transmitter signal dominate over other sources at the receiver; wheras using the better antenna on the receiver would increase the level of both desired and also undesired sources.

  • Using a "bad" antenna is more likely to negatively impact the operation of a practical transmitter than it is a practical receiver. This is likely a small issue - power levels and durations are low, so you're not really going to overheat the amplified due to a mismatched antenna. And the oscillator here is probably sufficiently decoupled from the antenna output that it is not going to be pulled off frequency by the antenna.

But even that is not the end of the story

  • Legal limitations on transmitter power typically include not only the source power, but also the effective output from the antenna. So using a better antenna may mean you need to turn down the transmitter power level

  • It's not absolutely clear which antenna is "better" - we might guess the plastic-encased whip is better than the PCB trace antennna, but that might not be the case. You might for example have been given an antenna made for a different band, or one that is mismanufactured.

  • Since you're actually talking about swapping the entire modules and not just the antennas, there could be implementation issues on either module that made it underperform

| improve this answer | |
  • \$\begingroup\$ Actually thinking a bit more about it, I'd optimize the PA matching, FCC doesn't care if you receive, but they do care about radiated harmonics. \$\endgroup\$ – isdi Oct 18 '18 at 17:48

Under ideal conditions it doesn't matter if the matching from receive/transmitter is equal. You may find that the matching circuit plays better with one or the other, it depends how the impedance matching network was optimized.

Sheet 33 of that datasheet is probably -not- what's on the module you have (at least in terms of parts values, if not circuit topology). You'll likely find that the radiation/orientation/antenna efficiency plays a bigger role in signal strength.

If you have an RSSI indicator on the chip you could setup up some tests, comparing transmit and receive characteristics for each antenna, but its likely to be a toss-up.

| improve this answer | |
  • \$\begingroup\$ Unfortunately these infamously don't have RSSI, only a "something there" bit \$\endgroup\$ – Chris Stratton Oct 18 '18 at 17:41
  • \$\begingroup\$ Yes, I started reviewing that datasheet some more, 0dBm max output level, I suppose you could hook up a Yagi or helical to it to get longer distance/directionality. My biggest worry with these types of designs are the harmonics, not so much the power levels. \$\endgroup\$ – isdi Oct 18 '18 at 17:46

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.