RDS For Part 15 FM Transmitters

A while ago, I became interested in RDS (RDBS in North America), and wondered if it would be possible to do it with Part 15 FM transmitters.  For those not familiar with this technology, it uses a high frequency subcarrier (57 Khz) to transmit digital data.  You can use this data to transmit the name of your station, the genre of your programming, information about the artist and song currently playing, etc.  Receivers capable of decoding this information will then display it.

RDS/RDBS is common in modern car radios, and can also be found in some portables.

There are various techniques to generate this data.  Hardware is perhaps the most common one, but this hardware is expensive, and you need support from the transmitter as well.

I turned to software as a potential solution, and found JMPX, an open source program that can generate the data, as well as a stereo signal (from mono).  It also contains a composite clipper.

The software as released was somewhat finicky with its settings, and required you to manually enter in the RDS/RDBS data.  I wanted it take the artist/song data from the Zara (automation software) current song file.  So I modified the source, and it's become one of my production applications.

I use it with my mono Decade MS-100.  I don't use the stereo feature (a mono signal gets longer noise-free range) but the RDBS data pops up onto my car radio display when I'm in range.  The data transmission is slow, and the Decade's range is small, so you pretty much have to be parked wtihin the target listening area to eventually see it.  Not as useful as I would have liked, but it certainly provides a proof of concept.

It should be noted that not all transmitters will pass along the higher frequencies required for RDS/RDBS.  The Decade will.  The Whole House 3 will not, nor will the Broadcastvision AXS-FMT.

On the computer side, to generate those high frequencies you require a DAC capable of handling 192Khz sampling (44/48 or 96 Khz just won't cut it).  One simple solution there is a Moshi USB-C to headphone jack audio cable that at the time was about US$15.  You just route output to that cable at the required sampling rate from your automation system.

Unfortunately, I lost the source code modifications at one point, so I cannot make any further changes without recreating the development environment again (which was a pain).  But if anyone is interested, I can certainly point them in the right direction for those changes and/or provide the runtime version of what I have.