Chapter 16. Line Encoding
The waveform pattern of voltage or current used to represent the 1s and 0s of a digital signal on a transmission link is called line encoding. The common types of line encoding are Unipolar, Polar, Bipolar and Manchester encoding.
Unipolar Encoding
Unipolar encoding has 2 voltage states, with one of the states being 0 volts. Since Unipolar line encoding has one of its states at 0 Volts, it is also called Return to Zero (RTZ). A common example of Unipolar line encoding is the TTL logic levels used in computers and digital logic.
Unipolar line encoding works well for inside machines--where the signal path is short-- but is unsuitable for long distances, due to the presence of stray capacitance in the transmission medium. On long transmission paths, the constant level shift from 0 to 5 volts, which causes the stray capacitance to charge up (remember, the capacitor charging formula is: 1-e-t/RC !). There will be a "stray" capacitor effect between any two conductors that are in close proximity to each other. For example, parallel running cables or wires are very suspect to stray capacitance.
If there is sufficient capacitance on the line (and a sufficient stream of 1s) a DC voltage component will be added to the data stream. Instead of returning to 0 volts, it would only return to 2 or 3 volts. The receiving station may not recognize a digital low at voltage of 2 volts!
Unipolar line encoding can have synchronization problems between the transmitter and receiver's clock oscillator. The receiver's clock oscillator locks on to the transmitted signal's level shifts (logic changes from 0 to 1) if there is a long series of logical 1s or 0s in a row. There is no level shift for the receiver's oscillator to lock to. The receiver oscillator's frequency may drift and become unsynchronized: it could lose track of where the receiver is supposed to sample the transmitted data!
Receive oscillator may drift during the period of all 1s
Polar Encoding
When the digital encoding is symmetrical--around 0 Volts--it is called a Polar Code. For example, the RS-232D interface uses Polar line encoding. The signal does not return to zero; it is either a +ve voltage or a -ve voltage. Polar line encoding is also called None Return To Zero (NRZ). Polar line encoding is the simplest pattern that eliminates most of the residual DC problem.
There is still a small residual DC problem, but Polar line encoding is a great improvement over Unipolar line encoding. Polar encoding has an added benefit in that it reduces the power required to transmit the signal by one-half.
RS-232D TXD
Polar and Unipolar line encoding both share the same synchronization problem: if there is a long string of logical 1s or 0s, the receive oscillator may drift and become unsynchronized.
Bipolar Line Encoding
Bipolar line encoding has 3 voltage levels. A low or 0 is represented by a 0 Volt level and a 1 is represented by alternating polarity pulses. By alternating the polarity of the pulses for 1s, the residual DC component cancels.
Bipolar Line Encoding
Synchronization of receive and transmit clocks is greatly improved--except if there is a long string of 0s transmitted. Bipolar line encoding is also called Alternate Mark Inversion (AMI).
Manchester Line Encoding
In Manchester Line Encoding, there is a transition at the middle of each bit period. The mid-bit transition serves as a clocking mechanism (and also as data): a low to high transition represents a 1 and a high to low transition represents a 0.
Manchester line encoding has no DC component and there is always a transition available for synchronizing receive and transmit clocks. Manchester line encoding is also called self-clocking line encoding. It has the added benefit of requiring the least amount of bandwidth compared to the other line encoding. Manchester line encoding requires 2 frequencies: the base carrier and 2 x the carrier frequency. All others require a range from 0 hertz to the maximum transfer rate frequency.
Manchester line encoding can detect errors during transmission: a transition is expected during every bit period. Therefore, the absence of a transition would indicate an error condition.
If this section was helpful, why not donate to further development?