Clocks and serial data transmission

I'm currious as to if there are any single line tranmission techniques that include the clock on the data line? I was thinking that one could "multiplex" many digital signals on one line by delaying them.

An analogy might be musical beats where the first beat is the clock, the second beat is the data for one device, the third is the data for another, etc...

something like

Clock +---+ +---+ | | | |

---+ +--------------------+ +--...

Data 1 +---+ | |

------------+ +------------------

Data 2 +---+ | |

---------------------+ +---------

Not sure if this is practical but just wondering...

Thanks, Jon

Reply to
Abstract Dissonance
Loading thread data ...

"Abstract Dissonance" schreef in bericht news: snipped-for-privacy@corp.supernews.com...

The start- and stopbits in asynchronous serial comminication are sometimes considered clocksignals. (They are actually synchronisation signals.) A well kown way of combining data and clock is Manchester encoding. Just Google for it.

petrus bitbyter

Reply to
petrus bitbyter

that

the

another,

You might say that X-10 data is "clocked" by the 60Hz power line it travels on. You should also look into Manchester encoding. Easy to generate, but a pain to receive.

Reply to
Anthony Fremont

Hi, Jon. Good idea, but it won't work. You'd then end up with a data rate that's 3 X the clock in your example above, and you'd have the difficulty of telling the different data bits apart, except at 3 X the rate of data transmission. Not only that, but you'd only get 2 X the data for a 3X increase in clock speed. You'd be wasting 1/3 of your bandwidth.

Sending data a bit at a time without a separate clock wire is called asynchronous serial communication (to paraphrase Jan, the Palmolive manicurist, "Youre posting with it!"). One way this is commonly done is for the sender and the receiver to be agreed on the rate of transmission, number of bits to be sent at a time, and other various communication parameters. The individual bits of each byte are then sent, with start and stop bits used at the beginning and end of each byte to keep resynchronizing the transmission if their internal timing is a little different. Various methods are then used to check the integrity of the data, and retransmit if necessary.

There's been quite a bit of work done on this subject over the past 100 years or so. It's a delightful trip from the telegraph to today. Also a long one, and a lot more than a series of posts on s.e.b. can handle.

If you want a start on the topic, you could do worse than looking at Jan Axelson's Lakeview Research Serial Port Central page. RS-232 is a good place to start for someone with some natural talent in practical electronics and computer experience like yourself. Just about every old PC made since about 1985 has a serial port built in, so grab a junker PC out of your garage, and you've got a hardware base for your research.

formatting link

Good luck Chris

Reply to
Chris

...

Fibre Channel is a newer serial communication protocol that does not have a separate clock line. Information is encoded so that the clocking can be recovered by the receiver. Fibre Channel works up to 4 gigabits/second, last time I looked, but they keep doubling the speed every time I check.

There are commodity parts you can buy off the shelf to handle the whole process. You can also find higher end workstations, and even disk drives, that have Fibre Channel built in.

Reply to
Don Taylor

Thanks,

I've found a few sites with the different encoding methods.

One thing I was always curious about was using multiple voltage levels. I saw that there was a encoding that does this for ethernet I think.

why doesn't one use many voltage levels to transmit data? lets say -4 to +4 for 8 bit data(with 0 being null I guess)? I know the more levels you have the larger a chance for corrupted data but lets say if one used +-50 volts... couldn't one just use DAC's and ADC's to transmit and recieve the signal at 8x faster than using 2 logic levels?

Seems like this is a way to increase the data throughput while keeping the data rate constant? Even if there were corruption one could retransmit and still have pretty could throughput?

Thanks, Jon

Reply to
Abstract Dissonance

Since you mentioned music: check out S/PDIF, it uses biphase mark encoding, similar to Manchester encoding. One drawback: (semi)constant clock rate only, or if you are using a phase locked loop to recover the clock then at least momentarily a reception of mess if you change the clock on the fly.

About using many voltage levels to transmit: it's already used, see for example quadrature amplitude modulation (QAM).

QAM is one way to increase throughput. But also it gradually makes everything more susceptible to background or circuit noise that will cause increased error rates.

formatting link

- Jan

Reply to
Jan Wagner

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.