The accompanying file rs232-pl513.pbm is an 874x821 PBM file containing
the schematic for the circuit I built for interfacing between RS-232
and the PL513 from X-10.  (If you're FTPing this, I strongly recommend
you ask for rs232-pl513.pbm.gz instead - the FTP server knows how to
gzip it for you, and it's only 7489 bytes gzipped instead of 90321
bytes ungzipped.)

Components: all resistors are very low-wattage, probably 1/8 watt at
most; resistances are given in ohms (or, with K, kiloohms).  All diodes
are 1N914, except for the circled one, which is a LED.  All transistors
are generic silicon general-purpose (switching / small-signal amp)
transistors, the kind you can get a handful of for five bucks at Radio
Shack (most are NPN, but note that one is PNP).  Logic gates are as
described on the schematic.  The boxes marked "190" are 74LS190s.  The
21L02 is a 1024x1 static RAM obtained years ago at Radio Shack.  The
boxes marked "123" are each half a 74LS123.  They are all wired with
the CLR input held high.  If the input is labeled with a down-pointing
arrow, it's the A input, with the B input high; if an up-pointing
arrow, it's the B input, with the A input held low.  The text above
them gives the resistance and capacitance of the resistor and capacitor
connected to them.  An output labeled with a high-going pulse is the Q
output; low-going, Q-bar.  A down-pointing triangle formed of three
horizontal lines is circuit ground; a hollow up-pointing triangle with
a solid outline is the +5V supply.  In my case, this comes from a
leftover computer power supply, though you could probably just use a
7805, particularly if you use all LS logic and heatsink the regulator.

Note that this schematic describes the circuit I actually built.  There
are several things I intend to do to it before transferring it from a
breadboard to something more permanent:

- RS-232 pin 2 should get a pulldown to ground similar to the one pin
  20 has, for better noise immunity.

- The 190s (decade counters) should be replaced by 191s (binary
  counters), and a third 191 added to make use of the two currently
  unused address pins on the 21L02.

- The AND gate currently fed by the Dout pin on the 21L02 should be
  eliminated in favor of using one of the AND gates in the '51.  This
  would free up the gate to be used in place of the diode connected
  from the RESET switch to the cross-coupled NAND gates - TTL can
  generally tolerate having an output grounded, but it's really not the
  best way to do what I want.

- Resistor values should be rationalized.

- A circuit should be added to reset automatically on power up.

- Some way should be added to tell the computer when the transmission
  is finished.  I haven't experimented with this yet; it may not be
  possible to do with just a +5 supply.

					der Mouse

			    mouse@collatz.mcrcim.mcgill.edu