Jump to content
Light-O-Rama Forums

Recommended Posts

Posted

I want to add RDS to my FM transmitter this year (assuming support will be added to LOR2!).

First, before I get into the details, one may ask why go to the trouble of RDS on a Christmas light display? After all, I want my visitors to look at my lights, not at their dashboard to see the name of the song. Not only that, but many of my songs are popular Christmas tunes, and everyone already knows the name!

The answer is that this is just another toy for me ;). And who knows, maybe I'll learn something!

I ordered a Pira32 RDS encoder kit, and had my first experience with soldering surface mount devices. It's not pretty, but I'll put some pictures up here later. I studied the schematic and found that the encoder has two power voltages on the board, 5 volts for the digital portion, and 8-20 volts for the analog (signal) section. I could have build the board as specified, but then I'd have to find yet another wall-wart power supply. Instead, I realized that I have a perfectly good 5 and 12-volt power supply in my computer, which of course will be on all the time I'm transmitting, so I didn't solder the 5-volt regulator, but hooked it up to a "disk drive" style power connector.

The good news is the RDS encoder seems to work. I hooked it up to the transmitter by finding where the multiplex-encoded audio came out of one of the chips, and cut into the line at that point.

(Actually, I bought a second transmitter to do this. I didn't want to risk not having a working transmitter if something went wrong. I also know that the Ramsey transmitters are susceptible to electrostatic discharge damage from their antenna lead, and I wanted to have a backup just in case.)

The result was somewhat disappointing. The encoder introduced noise, and it had trouble locking into the 19kHz pilot tone. I couldn't even see the waveform on my oscilloscope (which I rarely have a chance to use).

I came to the conclusion that the signal applied to the varactor diode on the transmitter is at too low a level to work well with the RDS encoder. Oh well. Now I will search for a transmitter that someone has already used with that encoder. Any ideas?

  • 2 weeks later...
Posted

I got it to work! :D For a full description, including pictures of the Pira32 kit, read bluebonnetlights.com/RDS.

In order to get it work, I made the following changes:

  1. I used the power circuit specified by the engineers who designed the board. Using the power from computer was a bad idea.
  2. I switched the way the RDS encoder connects to the transmitter. I removed jumper JP2 and changed the wiring in the transmitter (see below).
  3. I used the software from Pira instead of HyperTerminal to connect to the encoder.

A clean power supply is a requirement for audio equipment such as an FM transmitter. I believe the objectional noise was coming from the computer's power supply, or was introduced over the long (~20 inches) power leads. Instead, I used a spare wall wart power supply with the 5-volt regulator and filter capacitors built into the Pira32. Much better! Hopefully this will teach me to trust the engineers who actually designed the product, rather than trying to second-guess them.

The Pira32 can pass the stereo MPX signal from its input to its output, as it adds the RDS signal. I believe this was another source of the excess noise, because the signal applied directly to the varactor diode is much smaller than a typical line-level audio signal. This was the way I first modified the FM25B:
ramsey1.JPG


The second method to connect the Pira32 is to tap the stereo MPX signal in order to get the 19 kHz pilot signal, but only have the RDS signal at the output, and let the transmitter mix the signals. In order to do this, I removed the changes I made in the picture above, and made these changes instead:
ramsey2.JPG

I now hear a fairly clean stereo audio signal in my car, and can see the song name scroll on the display of my Toyota. I still need to test it with my neighbor's Acura, to make sure it works with a variety of RDS radios.

  • 4 weeks later...
Posted

As requested by Doug, here are some pictures of my modifications to my transmitter. This is a really ugly modification, but it works, so I'm not going to mess with it anymore. I also did a few simple noise tests, and I came to the conclusion that it is good enough.

FM25B-back.jpg
FM25B-inside.jpg

Posted

Nice work Steven ... and thanks for sharing with us !

  • 2 weeks later...
Posted

Here I am again to report on more changes! I finally got another chance to use some of what I learned in school getting my EE degree 25 years ago. (I never used it on the job as I stayed in the software industry.)

I studied the VCO circuit in the transmitter, which is composed of a module in the chip, as well as D6, D7, and a bunch of support components. The VCO circuit has three different frequencies. One is the transmit frequency (88-103Mhz). The second is the audio modulation (30Hz - 57kHz). The third is the PLL correction signal (less than 1Hz).

The correction signal is applied to D6 from Q2, through R19 and R20, and filtered by C14. The audio modulation is applied to D7 through C20 and R22.

My latest mistake was applying the RDS signal to D6. This caused the 57kHz RDS signal to be filtered through L1 and the .001µF capacitor to ground. Both the .001µF capacitor and the .10µH coil present a very low impedance to 57kHz (the RDS signal), while not blocking the <1Hz correction signal. The result is that I had to turn the RDS input signal higher, which introduced low frequency noise.

The solution was simply to move the injection point from D6 to D7, where the audio signal is. C11, the 10pF capacitor, keeps the 57kHz signal from being passed to ground. I also plan to increase the 47k resistor to 100k, to help isolate the RDS input a little more, as I believe the RDS signal may be a bit too strong.

But it now works more reliably with less noise! This is what I'm going to use this Christmas!

Here's the new circuit:
ramsey3.JPG

Guest
This topic is now closed to further replies.
×
×
  • Create New...