Buenas de nuevo.
Mira te explico. Resulta que la HP 49G (los primeros modelos que salieron) tienen un bug que hacen que no se pueda establecer comunicación mediante el puerto paralelo, por que hay algunos transistores que reciben mas corriente de la que deberian y no están preparados (Eso es lo que mas o menos he entendido xD). No obstante la solución que dan, es poner ese circuito entre el RS232 del PC y la calculadora para que atenue eso...
Y bueno pues ahí están mis dudas xD. Pongo la dirección del PDF donde encontré la info
http://www.hpcalc.org/hp49/docs/misc/buf49.zip
A ver si me podeis echar una manita
. Un saludo.
EDITO: También hay otra solución que consiste en hacer unos puentes internos a la calculadora.
En este PDF explican algo mejor el porque del bug:
http://www.hpcalc.org/hp49/docs/misc/serial.zip
"While working on a serial to IR converter, I discovered an error in the current hardware
design of the HP49G. The schematic of the serial buffer, made up of 4 transistors, a few
resistors and 2 diodes, differs from the original HP48 design. The driver should function as a
current limiter in case to much current flows into or out of the output (HP_TX).
When you take a look at the original schematic (page 3), you can see that a current
over 4mA (I = Ube / R? = 0.6V / 150R = 4mA) causes Q6 or Q7 to conduct and these will
'pinch' off Q5 or Q8 to limit the current that will flow to a few milliamps.
Now take a look at the new schematic drawing of the 49. Here you can see that R1 is not
connected between the base-emitter juntions of Q1 and Q4, and can therefore never protect
the port when to much current flows in or out of the port. And Q2 and Q4 look like they are
upside down!
Another error is made with the diodes. In the 48 design two separate schottky diodes were
used (glass melf). They have been replace by a dual schottky (L44 device marking, BAT54S
or equivalent) in a SOT23 package. One diode passes a high voltage (spacing) to the output
and the other one a low voltage (marking bits). When you take a look at Q2 and D1 you can
see that the two PN junctions are 'head to head'! The buffer schematic is definitely not doing
what it is supposed to do!"
Ahí viene explicado, pero no obstante, abrir la calculadora esta, es un trabajo de chinos aparte que no tengo un soldador para soldadura tan fina... Con lo cual casi descarto esta opción.