HT of the Future – New Design

A few years back, Chris Testa and I tried to make a mobile SDR called “Whitebox” or “HT of the Future”. That didn’t work out because the design was too noisy, and we made some mistakes, but we learned a lot.

A few years later, I’m ready to do a new design. It’s a smartphone-controlled 40 mW single-band transceiver based on the AX5043 from On Semiconductor. The hardware design and software are 100% Open Source. You connect to it using a web browser over WiFi, and it takes over your phone’s microphone and speaker, runs a codec in webassembly, and provides a completely-portable UI that runs on Android, Mac, Windows, and anything that runs an un-handicapped version of Chrome or Firefox (which may still not be Apple iPhone). It will be dirt-cheap, and should be able to do FM, packet (1200 and 9600), digital voice using Codec2 (with the codec running in the smartphone), and many experimental modes with data from 0.1 to 125 kilobits per second. It has WiFi (client and standalone access point simultaneously) to connect to your phone and the internet. and Bluetooth Low-Energy.

There is also a cryptographic coprocessor on the board. It cost less than $2. Under FCC rules, we can use digital signature to authenticate hams over the air, as long as the message content is not encrypted. So, there will be lots of interesting applications. But my main reason is that I don’t want to support clones. So, your radio can authenticate itself over the internet. It also provides support for SSL/TLS with a hardware-sequestered key, and lots of interesting general-purpose cryptographic power.

The overall bill of materials is under $25. So, I might be able to create the Baofeng of digital modes.

Here’s my first pass at a schematic. This is my first time using KiCad, and I’m not an electrical engineer or an RF head at all, so expect mistakes.

This simple design can obviously be enhanced, but I like how simple it is because it will hopefully be easy to get it working.  For example, the radio chip is so cheap (under $2 in quantity) that it works best to do multiband with one chip per band – and thus one separate working radio per band.

 

Want to chat? Email bruce at perens dot com