Our TV sits in an open bookshelf set at a 45 degree angle in the corner of the room.  

Recently, something got knocked off a lower shelf onto the floor behind.  I removed some books, stuck my head in to see behind, and removed several things that had been there for years, including a TV-like remote control.  I have often wanted to see what the signal from a remote looks like, so I probed it with my recently acquired oscilloscope.

I fully expected to see a bit stream encoded with Manchester Encoding, named for a technique used in an early computer at the University of Manchester.

The limited digital work I had done in my early experience used an available clock signal to time data, or otherwise synchronize digital logic.  I had wondered how things were done when a clock was not available, as when serial data is sent by radio, or by a TV remote control.  

Later, I was part of a project with a product that had five switches.  The associated 32 possible states were set by a 5-bit serial control word.  It was received by wire, but only one wire, and did not have a second wire for a clock signal.  The 5-bits were Manchester encoded.  

My lowly task was to test the 32 states by sending the control word from a test box (a "bit-box") having 5 toggle switches.  I don't remember the bit-box details, but I remember I used one-shots (i.e., a SN74123 integrated circuit), which any true digital designer would scorn.  I thought of each individual bit as being a pulse that had one of two possible trigger times, i.e., a pulse then no-pulse for a zero, or no-pulse then a pulse for a zero.  The more accepted way is to think of each bit is as a bi-phase signal, i.e., inverted for a zero, or not-inverted for a one.

Regardless of my crude approach for the bit-box, it worked, and life went on.   I was left with the lingering thought that Manchester encoding was a clever idea.  Now, decades later, came an opportunity to look at the signal in an infrared remote control.  The first presentation below shows a Micro-Cap simulation  for Manchester using 7400 Series TTL compatible ICs (still available for about 90 cents each) .

Well, the joke's on me.  I tested my remote and discovered it is not Manchester or bi-phase encoded, but instead uses Pulse Width Modulation.  This means the zero and one bits are not of equal time length, and so a word with more zeros is shorter than a word with more ones.   My first thought was,  "How distasteful."   But then I thought,  "Well, it's like Morse Code, the zero is a 'dit', and the one is a 'dah' ."    I feel much better, now.  

See the 2015 EDN article "IR Remote Control Basics" available here .

Also related:

Wikipedia:  Consumer IR

Wikipedia:  RC-5

"Infrared Remote Controller Knowledge Base", May 2021, by Holtek Semiconductor available  here .