Need a simple way of generating/transmitting roughly 2.4 GHz (unmodulated), ideally at over 20 dBm (0.1 W)

Question

Purpose / Problem

The aim is to track (large) insects, similar-ish to the approach in this paper. In my design I have a series of high-gain transmitters that rotate. The insect carries a device that records when these signals reach it^*. Later the device is recovered and read, and the path is reconstructed.

The problem is what to use to generate the (GHz)^** signal.

Details

I tested a prototype 5-6 years ago:

1. The device on the bee consists of: a tiny battery (e.g. this one), a simple receiver/detector and a microprocessor (e.g. an atTiny).
2. A handful of ~GHz transmitters are placed around the landscape, each with a high-gain antenna (satellite dish) on a rotating platform (rotating at about 50°/s). Each one is sending, encoded as simple on-off keying^***, about 8 times per second:
   • the angle it is facing (7 bits)
   • the time to the nearest 3 seconds, modulo 96 seconds (5 bits)
   • the transmitter ID (3 bits)
   • needs about 120 bits/s
3. The bee-device regularly wakes to see if there's a signal. If there is, it stores the time, angle, and ID.

Work done so far

For the experiment I just put a WiFi dongle in the focus of a (parabolic) satellite dish (but I didn't try modulating it on/off). The approach seemed to work, in that I could detect from 10s of metres when the dish was pointed at the detector.

The dongle was great in that it needed no RF expertise. I'm looking for a solution that requires as little RF electronic construction as possible. One ludicrous approach could be to use the ESP12 (datasheet, power output = 20.5 dBm) and hope I can somehow toggle on/off the WiFi signal, but this doesn't sound very promising. I just wonder what else there might be?

I wonder if I need to just hook up a GHz oscillator (e.g. this) to an RF amplifier (e.g. this) and feed that into a dipole (placed at the dish's focus). I suspect that this is not at all simple, so I'm wondering what simple module exists that makes ~2.4 Ghz (at 20 or 30 dBm+) and has an 'enable' pin that I can toggle on and off, with as little electronic work as possible (ideally I don't want to have to design PCBs etc.).

A note about interference etc.

(To answer these points before they're raised.) At the moment I'm just looking at what is practical, with the aim of having a device and system that could be built without too much RF expertise by ecologists, etc. Obviously, even just putting a WiFi-dongle in the focus of a satellite dish is technically against the rules. I'll be finding out how to do this work without causing interference etc. to others. More info is on the ofcom page.

^* I'm mainly interested in bumblebees which forage within a few hundred metres of the nest, to give an idea of the range needed.

^** If it's in the GHz range, it means the antenna on the device on the bee is small enough to carry.

^*** This makes the bee-device RF electronics simple (an inductor, a capacitor, a large loop antenna, and a diode).

Edit: What are you using to generate the OOK signal? This will be some simple microcontroller (maybe another arduino/ATtiny). That also controls the stepper motor that rotates the antenna, etc.

Answer 1

One option that could be interesting would be to use a bluetooth transceiver (like nrf52840) and put it in test mode. This will allow you to transmit unmodulated 2.4Ghz signals that can be turned on and off, probably up to the 100hz you need. A module like this will allow you to vary the power from -4dbm to +20dbm with no electronic modification.

If you wanted to do it yourself, the approach you describe would work (although that part only has 16db of gain at 2.4ghz) but you would need rf matching between your antenna and your PA.

Answer 2

There are transceivers out there that can transmit and receive ASK modulated signal, though I don't think they go to such a low symbol rate. Don't expect long range with such a setup, the 2.4GHz band is crowded in most places. A simpler solution would be to pick a wireless MCU. You can't beat the level of integration of these devices.