wrote in message
...
On Jun 2, 2:49 am, "N_Cook" wrote:
Like breaking a combination lock you need a "tell " to work on.
Assuming a piece of kit, impossible to get a remote control for and
"universal " remotes do not register any change to the otherwise
dead kit.
If you broke into a ground or supply line to the microcontroller
and
monitored the supply current , would there be staged changes of
current when
exercised by various , but vast majority wrong, IR signals?
ie simple swept 30 to 50 KHz oscillator source would you pick up on
say
38KHz as the basic required "carrier" frequency ?
Varying mark/space of gated pulses at that "carrier" f, would you
pick up on
the correct mark/space
?
Then would there be a recognisable respone to various random
"nibbles" that
are parts of the required coding ?

Years back I wanted to use a Sony remote (had extras) to control
something completely different. I bought a 38KHz IR receiver and sat
down with a scope and watched what happened and quickly found there is
a method to the madness. Different width pulses, repeated inverted
patterns, preambles. It wasn't too big a deal to write assembly code
to receive and decode a stream. Later with an early HDTV (before
integrated tuners) and set top box I did the same again and this time
received, transcoded and transmitted the same command to the other
boxes. BTW both Sony and Samsung (10 years + old) used 32 bits 8
start, 8 'unit ID' and 16 data bits sent as an 8 bit block followed by
the same pattern inverted. I haven't looked at anything recent.

It's really nothing more than an optical modem but everybody has
different protocols. They don't WANT to make it easy .

The problem is the large variety of codes and carrier frequencies
available. There isn't enough room to to store all the variants in a
low power system. I believe the Logitech Harmony is programmed via the
PC so only needs a small amount of memory. I know they're expensive -
that's why I don't have one but many folks swear by them (and a few
swear at them).

G²

+++++++

The Logitech is only a URC with extra bells and whistles. It has the same
failing of all the others. If the maker does not have the code to laydown in
its library placed in ROM or via PC , and none of the other URCs have this
code to be copied across from. Then its as non-universal as the other
"universals", just costs more.

The only thing in out favour is the requirement for customer satisfaction to
have a speedy response to a button pusk.
And also the limited number of combinations possible if there are other
constraints.
See my listing on the 10 th in this thread
There is very limited number of vatiants if you include a 111111111 subcode
because of the 9 blocks limit in this structure. A few more for 8x1 etc

That PVR I mentioned , where I obtained an original secondhand zapper.
There is a piezo-tick when the clock changes digit , or any othe rfunction
like choosing a letter for titling of DVD tracks etc.
Shining that bad RoSS zapper (clamped down FAV function which with hindsight
contained a large subsection of the PVR code) and the "6in1" "universal "
some codes elicited a few of thses clicks , unkown function if anything.
Reading out blink code faxhion from the "6in1" one for something Alba and
one for something Philips
They both had the same use of double width pulses in the code. But now
having the genuine code for that PVR if must have been interpreting a double
width pulses as 11 and part of the ROSS