Hello

I know there are some DDS users/programmers out there!

We have probably not seen the final variant of "sequential single tone
modulation" yet. So, in order to avoid unnecessary (and painful) PIC
software changes in the future, I'd suggest that we establish a standard.

It would be very nice if we could agree on a simple serial protocol
that would allow any PC program, be it DOS or Windoze, to take total
control over the DDS (I know that Andy G4JNT has done something
in this direction).

I am in the process of rewriting the PIC software for G0MRF's DDS
board, a serial input port is on the "todo" list.

If the DDS clock frequency, and possibly also a division ratio and offset
frequency, is known to the PC software (via a setup menu or whatever),
the math can be done in the PC, only the DDS phase increment has to be
sent to the DDS box.


Proposed serial protocol:

  9600 baud, 8N1.

  Only printable ASCII characters are used.

  '=' is used as "start-of-message" character.

  The '=' is followed by eight hexadecimal digits representing the 32-bit
     phase increment, MSB first! Example: "=0245FD4C3"

  The DDS box changes the output frequency if it receives 8 consecutive
     valid hex digits after the leading '='. In case of error, it will do 
nothing
     but continue putting out the old frequency and wait for a new
     '=' start character.

  A simple checksum byte (two hex characters) could be added at the
  end to prevent the TX from exploding if the frequency data should get
  corrupted from RF interference... Either "vertical parity" (just XOR the
  bytes) or a modulo 256 sum. CRC seems to be overkill...

(Alberto has indicated that he might add direct DDS TX control to Jason
if all DDS owners/programmers can agree on a serial protocol)


Please comment!

73
Johan SM6LKM