Basics in radio and directional finding using RPi, antennas and arduinos.
This information is for educational and informational purposes only. You have the final decision on modifing your hardware, software, and use of it. And yours is the responsability of any consequences of the improper use of it.
This is only a prototype, intensionally limited in speed and efficiency to aid in learning radio, any modifications for commercial or malicious intentions implies you know what you are doing, the risks and consequences.
Everything started with a comment during a twitch live on zerialkiller channel, someone suggested of ways to find WiFi devices and here I am a couple of months later.
The first part is how to keep your hardware safe, electronics after reaching
their limits, there is no UNDO button. And as described in the disclaimer,
this is a slow (but working) prototype with the sole objective of learning,
to serve as a minimal base to start. Check the video, I'm sure you will
find many ideas on how to continue.
This play have four acts:
- RF hardware connections and not frying electronics.
- Simple arduino firmware that moves the antenna.
- Slow and inefficient scripts capturing and plotting data.
- Creating plots.
This is a bad option, the diameter is off by several millimeters. In the other and, it is easy to get one, easy to drill and install the monopole. Make a 6mm hole 3-cm from the back of the can to insert the monopole later.
SMA and RP-SMA connector There are two types of SMA connectors, SMA and RP-SMA, the RP for reverse-polarity, RF-SMA was designed to comply with FCC regulations and avoid what we are going to do, connect Wi-Fi equipment to radio equipment not designed for that. Check the connector and buy the adapter you need.
The total length is ~31-mm including any extra length from the connector itself. The first time I made one of these antennas I used calipers and sand paper to be sure the length was right down to 0.1-mm now that I have the equipment to see how close it needs to be, you can save time, +/-1-mm is close enough, in fact I tried 29-mm to 34-mm there are differences on performance and they will work too.
I make a clamp to hold the cantenna, hold the USB dongle on top and plug into
the stepper motor axis directly. I'm using a 28BYJ-48 motor, in case you use
the same check the OpenSCAD just change
axis_d and axis_t to make the square hole match the axis of your motor.
You need this? No. Duct-tape and hot-glue are much faster and simple. My inner child wants to persuade my adult me that it wasn't an impulse buy.
NOTE: Please check the video with zerialkiller there are more details there that I can add here. Hardware can make or break the project, modifying 90-cm parabolic antennas to use bi-quad feeds will increase coverage area and detect devices with lower power output. The issue is you loose accuracy, instead of 0.5-m indoors you will get several meters on proportionally larger area.
Here things start to get inefficient (even more). The arduino expect commands on serial port, three, no four, no five. Three 😁
- R to reset the counter this defines the current heading of the antenna to zero.
- cNNNN get the current position of the antenna it is the step counter. NNNN is a four digits number.
- gNNNN go-to
NNNNposition, it needs to be a 4-digit number, leading empty spaces for 999 and lower must be zeros.
Why this is slow? After a go-to the arduino doesn't reply to any messages. If you thought on requesting a full circle, capture data and get the position during the rotation to increase readout speed, you are in the right path. Fix this issue and you can reduce time.