[App_rpt-users] Ardunio PLL Programming of GE MVS

Benjamin Naber Benjamin at Project23D.com
Tue Jan 9 23:56:00 EST 2018


For those of you who are knowledgeable in the code of Arduino:

I've begun a project that I jumped into the the *very* deep end, while
knowing just barely how to swim.

I've managed to get the radio to tune to a frequency that I want, but
very flaky at best when it comes to tuning well within the limits of
thr VCO; the arduino freezes every now and then, and quite frankly I am
not working with (cough, not understanding), the existing code.

In a nutshell, what I am doing is breaking the MC145159 control lines
from the GE MVS CPU, and soldering in a few wires, plus a ground, to
the base pins of the transistors that ulitamtely lead to the CLK,ENB,
and DATA pins of the PLL.

The reason why I am doing this is to enable free tuning of the radio
without the need to program the radio using normal computer programming
methods, ie, a cable, and a special program that is hard to come by, as
well as must run this program on older hardware

Given that Arduinos can be programmed from just about anything, this
enabled on the fly on at least, on site programming of the radio. I am
working towards enabling the use of these radios, and a couple others
to be controlled by repeater controllers, and allstarlink nodes using
kenwood data command set. Or whatever it is.

Boiled down, this will enable good, inexpensive priced commercial
radios to continue life; anyone who knows how to solder and program an
Arduino can now use a commercial for a repeater VHF/UHF remote base,
Allstarlink, IRLP, Echostink node radio, scanner, whatever.

With the significantly better RX fronts ends on these radios, cleaner
TX, these will significantly outperform any chinese radio attached to
any RoIP node.


---------
Here is where I am stuck:

The MC145159 PLL wants the tuning data sent as follows:
14 bits for R counter
10 bits for N counter
7 bits for A counter
1 high bit for control.

Totaling 32 bits. The current code, posted immediately below does not
seem to allow this:


   emit_byte (0x14);
   emit_byte (0x01);
/* Send "(n << 8)|(a << 1)", 24 bits, MSB first, LSB always zero */
  //emit_byte (n >> 2);
  emit_byte ((n >> 8 ) & 0xFF);  /* N high byte */
  emit_byte (n & 0xFF);   /* N low byte */
  //emit_byte (a & 0b1);   // send only seven bits?
  emit_byte (a << 1);   /* A and LSB 0 */
  //emit_byte (0x01 << 7);        // send only one HIGH bit?
  
  pulse_le();     /* Latch it */

the "//" are my futile attempts to work around this, and leave
commented out so I know what I *think* did not work.

I've also the following to form the four bytes being sent, MSB of each
counter first, control bit last:

emit_byte ((r) & 0xFF);  /* send first eight bits of r counter*/
emit_byte (((r & 0xFF) & n) & 0xFF);	/* send last 6 bits of r
counter, then send first two bits of n counter */
  
emit_byte ((n & 0xFF) & 0xFF);	/* send last eight bits of n
counter */
emit_byte (((a) & 0xFF) & 0x01 << 7); 	/* send seven bits for a
counter, and one HIGH bit for control bit */


What does work now to test to make sure this project is even feasible:

in the void setup() section

....
  emit_byte (0b00101000);  // first eight bits of R counter
  emit_byte (0b00000001);  // last 6 bits of R counter, first two bits
of N counter 
  
  emit_byte (0b00110011);  // last eight bits of N counter
  emit_byte (0b10001001);  // 7 bits for A counter, last control bit
  pulse_le();

.......

the above test code does work to successfully set the radio to 5 KHz
step, and program the N + A counters to 151.820MHz + 45MHz




Sorry it was long, However the more info you have to see what is going
on, the less guessing you have to do.

~Benjamin, KB9LFZ


More information about the App_rpt-users mailing list