Posted: 10/20/2014 3:01:11 PM EDT
|
I have a project I've been poking at the last year or so. It involves modifying a Motorola Syntor X9000 29-54 MHz lowband radio. These are old late 80s-early 90s radios. They are EPROM based. The software basically loads an image in an EPROM, the microprocessor selects addresses from the memory and loads the data up. It's pretty basic in its design but the RF hardware is amazing. It will pull out signals nothing else will, especially with the built in noise blanker.
I took an Arduino Mega 2560 and wrote a program that emulates the data to the synthesizer divider. I found I can drive it at incredible scan speeds. This allowed me to tune to any frequency in the usable range of the radio. I then added a keypad and OLED display. The result was direct VFO and keypad entry. It works great. What I don't understand is how the handled the PL and DCS tone signalling. The radio taps the discriminator, cleans it up a bit and sends it through a comparator circuit where it's converted to a nice 0-5 volt logic equivalent of the tone information. It's then sent to the ancient Hitachi microprocessor on pin 24. What I don't understand is how they managed to get an ancient chip like that to detect a tone code so fast and accurately. I tried using numerous different frequency counter library files with the Arduino and the Teensy boards. They worked fairly well but don't seem to be as fast and accurate, especially at low tones like 67-94.8. I know it's a long shot but does anyone have any experience or ideas how to reliably and quickly do the same thing with an Arduino or Teensy? If I could get the signalling down I could put it all together and have myself a nice VFO capable radio. Here's a video of the setup being tested while listening to 10 Meters. 
|
|
If they're doing it after a comparator, it's looking at the pulse width coming out of it, rather than looking at the frequency. it would only take a few
pulses providing a consistent pulse width timing to lock on the tone. If it were done using analog, it's probably be a goertzel algorithm (ie DSP resonator). But from what you described it sounds like they're just computing the pulse width and going from there. |
|
Quoted:
If they're doing it after a comparator, it's looking at the pulse width coming out of it, rather than looking at the frequency. it would only take a few pulses providing a consistent pulse width timing to lock on the tone. If it were done using analog, it's probably be a goertzel algorithm (ie DSP resonator). But from what you described it sounds like they're just computing the pulse width and going from there. Hmm pulse width...ok. I will try to figure that out. Thanks. |
|
PWM is how I'd replicate it…that's how a few friends and I designed a 1000 Hz tone for a MSP430 based controller.
You have my attention with this project…especially because I just purchased one of these radios yesterday. The diplexing article Moto put out with them is quite interesting… |