🚀 go-pugleaf

A while back I posted a request for ideas for a circuit to simulate my
TV's remote control on-off output so I could modify my clock radio to
turn on my TV.

I got several suggestions, but most were not very elegant or
practical.  Several suggested some kind of contraption to press the
on-off button on my present remote.  But at least one person agreed
with me that the best plan was to somehow decode the output of my
present remote and build something to duplicate it.  (As you may
remember, the original IC was no longer available.)

One big problem I had (and would guess everyone would have) is the
difficulty in deciphering the pulse location and spacing with a
conventional oscilloscope.  You get a lot of fuzzy jumble and can't
easily find the starting point of the sequence.  Eventually I was able
to use a pulse timing instrument to get a clean display, and was able
to identify the leader, custom and data codes.  Now all I needed was a
circuit to produce it.

My first idea was to use a 555 timer for the 40 KHz carrier, and
other ICs for a clock, address counter and ROM to generate the actual
code.  This added up to a lot of parts, so a second idea was to use a
PAL to reduce the parts count.  I could have also used a bunch of one-
shots etc., but that would have meant even more parts.

Once I was up to PAL complexity, a microcomputer looked attractive.
With advice from our department's uP expert, I ended up selecting the
Signetics 87C751.  It was the smallest programable part I could find
(24 pin "Skinny DIP"), and its 16K of EPROM was plenty for the task.

Having selected the 751, we looked into having it generate the whole
signal, carrier and all.  A nice one-IC solution.  The problem with
this idea is that you have to generate an 8.3 uS minimum pulse width
(1/40 KHz, 30% duty cycle), and a very long bit sequence.  We could
handle the long sequence, but even at a 12 MHz clock rate, the 751
didn't have time to do all of the necessary subroutines.  So, I
resigned myself to using the 555 for the carrier and the 751 to
modulate it.  Two ICs to replace one custom circuit - not bad.

Next was how to use the 751 to generate all of the required pulse
widths and delays.  Its internal timers seemed obvious.  However, the
I2C timer doesn't produce the correct code, and the fixed rate timer
doesn't output to an external pin.  It could be used to time the
various pulses and delays, but it was easier just to program timing
loops in the software.

When the circuit was all done, the thing didn't work.  After much
debugging, the problem was found to be a missing bit on the end of the
code.  It was real easy to miss it.

I had one more problem after I got the circuit installed in the radio.  When
the alarm turned on, the radio put out a short blurp before the muting took
affect.  I fixed this by putting a low-ohm MOSFET in series with the speaker
and an R/C delay in its gate.  Now the whole thing works perfectly.

Jerry Willard,  Tektronix
jerryw@teksce.SCE.TEK.COM