APRS TINY – From Ron Graham.

I was a beta tester for Byron, N6BG’s “tinytrak”. This unit uses the PIC to
generate the APRS packet transmit tones, thus doing away with the requirement
of a modem IC. It cannot receive APRS data, but handles the transmitt side in
a small, and economical manner, together with low power consumption .. so it
is, I think, the most economical way to build an APRS tracker.

In a similar manner as the “APRS INTERFACE”, it is fed NMEA data from the
GPS receiver and generates the APRS packet (AX-25) tones together with the PTT
(transmitt keying) which is fed to the transmitter. If a transceiver is used,
the muted receive audio may be fed to the APRS TINY, which is used to indicate
that the packet channel is busy .. thus inhibiting the transmitter.

LED’s are fitted to indicate “valid GPS data”, “channel busy” and “PTT”.

The unit is programmed in 2 steps. The initial step requires the PIC to be
programmed in a normal PIC programmer with Byron’s tinytrax.hex. For the
second step, the PIC is fitted in the APRS TINY, the APRS TINY is powered up
and plugged (with a dedicated programming cable) to a serial port of a
computer running WINDOWS 95, or better. Running tinytrax.exe enables one to
programme in your call sign, transmitt interval, beacon text etc.

I did a PCB design and fitted 2 sockets. One is dedicated to the radio
interface. The other has dual functions .. initially for the above mentioned
programming cable .. then once the unit is programmed, to interface to the GPS

Another facility added was the option to power the APRS TINY with 12 volts
from the radio, and the APRS TINY in turn power the GPS receiver .. IF the GPS
is suitable for 12 volt operation. Thus the mess of heaps of power cables is

The PC board measures 44 X 48 mm and was designed to fit into a specially built
solid aluminium box 50 mm square and 30 mm high. Further to making the power supply
arrangments as flexible as possible, a “Japanese” type concentric power socket (centre
positive) is also fitted to the board. I have found this handy for powering the unit
during testing and programming. So another option, if power is NOT supplied from the
radio, is to power the unit via that socket .. the unit, in turn, may power the GPS


Some mods and software updates have been incorporated to enhance the performance
of this device.

  1. Software. The latest software adds some extra LED functions.
    1. “Valid” LED now shows 3 states .. OFF for no GPS data, FLASHING for un-locked GPS
      data, ON for locked GPS data.

    2. “Carrier detect” LED also now shows 3 states .. OFF for no audio heard for “quittime”
      seconds and ready to transmit, FLASHING for for no audio heard but still in “quiettime”,
      ON for audio currently heard.

    Also, 2 timing parameters for the tranmitting period are now incorporated. This is quite
    handy for, say example, 30 seconds transmit period for when mobile, 5 minutes when stationary.
    The 2 parameters are selected by the jumper on pin 12 of the PIC. This could also be wired
    to an external switch.

  2. Hardware mods. Some users reported difficulty with PTT switching probleme with certain brands/models
    of radios.

    1. Reducing the value of the base drive series resistor of the 2N2222 from 10K to 4K7 helps
      in some cases.

    2. using a mosfet device in lieu of the 2N2222 also solved some problems as the “on”
      resistance of these devices is much less than the 2N2222. The inexpensive 2N7000
      is recomended. Note that this device is wired with the “flat” in the opposite
      direction to the 2N2222.

    3. some users have experienced a problem with the PIC losing memory contents after a
      period of not using the unit for a month or so. This makes it necessary to re-program
      the PIC with the”tinytrax.exe” program. One theory is that with the unit running, and
      not plugged into the GPS receiver, the input of the PIC is open circuited and subsectible
      to noise appearing at it input .. and possibly causing the loss (or corruption) of its
      EEPROM section. So an extra 4K7 resistor from the input socket to ground is suggested
      as a cure. Time, plus feedback from users, is now needed to see if this solves
      the problem.

    4. a change, by my supplier, of the type of pre-set pot. has meant that a revision of the PC
      board is necessary. The new pots have metal, connected to the moving arm, on the
      outer edges .. which means that if the bodies of the pots touch, there is a chance
      of a short circuit occuring. So the new PCB spaces them slightly further apart, but
      care will be needed during assembly to keep them seperated.
      The new PCB also incorporates the mods above (2N7000, change 1 resistor, add another)
      plus a couple of cosmetic changes.

Ron Graham can be contacted at http://users.mackay.net.au/~ron/