*** AM-7910 and 2400 Bd ***

After sending messages with modification for 2400 baud on the TCM-3105 and
a DCD circuit,Due of the BIG Respons here is the modification for the
AM-7910 by ON5ZS and ON7PC


Send your reply's to Walter ON1AOT @ ON6AR


   2400 BAUD                                          December 25, 1991
   ---------

   The 2400 Bd we are using is very simple, it was succesfully experimented
   by Daniel ON5ZS. One of the advantages is that you DO NOT NEED TO
   MODIFY YOUR TRANSCEIVER ! If you have a TNC with AM7910 chip, you
   can be qrv and enjoy this speed after minor modifications to your TNC.

   The first modification is the clock frequency of the AM7910 :
   The master timing of the AM7910 is normally provided by a 2.4576 MHz
   crystal, in order to run 2400 Bd the AM7910 now must be controlled by
   a 3.6864 MHz. You can either use the internal oscillator of the AM7910
   with a 3.6864 MHz crystal connected between  pin 23 and 24 and two 22 pF
   capacitors (beter one of 18 pF, another of 22 pF and a 1 Mohm resistor
   in parallel with the crystal !), or as alternative solution you can built
   an external oscillator (with 2 gates of a 74HC04 for example ... see your
   TTL cookbook) and apply this signal to pin 24 of the AM7910. All other
   connections are the same as in the previous configuration for 1200 Bd
   (Bell 202). With these modifications, the tones are now 1800/3300 Hz.

   Note for Eisch TNC2 model 4 : if you use a 9.8304 MHz oscillator, all
   you need is to put the 3.6864 MHz crystal at the place of the 2.4576 MHz
   crystal and you will find the place for the two 22 pF (or beter 39 and 47
   pF) and the 1 Mohm resistor ... The 2.4576 MHz signal goes from the
   divider (74HC74) first to pin 24 of the AM7910, and from there to the baud
   rate generator (4040), and you have to cut the pcb tracks on TWO points and
   place a piece of wire directly from the 2.4576 MHz point to pin 10 of the
   4040. Finally place a second piece of wire between pin 6 of 4069 (3.6864 MHz
   signal) to pin 24 of AM7910.

   The second modification is to provide the 2400 Bd rate :
   The NRZI coder (or state machine) is fed with 16 x the baud rate, so you
   must also apply now a 16 x 2400 = 38400 Hz clock signal.

   Of course you can wire a switch to have both speeds : 1200 and 2400 Bd !

   Before complaining ... check with a frequencymeter, you must have
   3.6864 MHz at pin 24 of the AM7910 and a clock signal of 38400 Hz !

   WE HAVE NO EXPERIENCE FOR OTHER MODEM CHIPS (so don't ask for it!) but
   we are pleased if you drop a message in the PBBS for other chips, giving
   100% compatibility with the above modification !

   In order to get full advantage of the speed increase, please reduce your
   TXDELAY. Lets compute the time for a  AX.25 frame and lets compare it with
   the usual 300 ms TXDELAY :
					bytes   bits  time at     300 msec
						       2400Bd   represents
   short frame (just "RR" for example)     19    152    63 ms        476 %
   long frame  (256 bytes info frame)     276   2208   920 ms         32 %

   You see that the TXDELAY is really a TIME WASTE, but it is necessary to be
   sure that the PLL of the transmitter is correctly locked and it is
   necessary to be sure that the squelch circuit of the receiver is open.
   (note that both time constants have been designed for "phone transmissions",
   but not for data !).

     *********************************************************************
     *  We beg you to reduce your TXDELAY at the lowest possible value.  *
     *  For example if you loose the connection with a TXDELAY of 11     *
     *  and got a good copy for 12, then use a TXDELAY of 15             *
     *  (you have then a 30 % safety margin). Please do it !             *
     *********************************************************************


73's Walter ON1AOT @ ON6AR


PS: the TCM-3105 modification and the AM-7910 Modification are 100% compatibel.