When bauds are transmitted serially over distances greater than a few metres the logic levels output by ICs and microcontrollers are inadequate, and typically a line driver (also known as a transceiver) is needed.
Line driver ICs improve data transmission reliability and noise immunity by converting the logic levels into a form more suited for transmission to a cable, and by restoring the digital bits on reception from a cable. The signal may be sent as a volatge, a current, pulse of light, radio frequencey tone, etc. This course will only consider cabled connections representing a baud by a specific voltage level. Some common physical layers used for serial communications are EIA-232, EIA-432, and EIA-485.
Transmiting line drivers typicaly send a voltage eitehr +5 or larger; and 0 to -15V. Receivers for unbalanced transmission set a receive threshold that is relative to a common reference (ground). Common reference levels are +/- 3V. For balanced transmission, such as EIA-485, smaller voltages are often used on the line (e.g. 0/5V). Receivers for balanced transmission, use a threshold measured as the difference between the two receive lines.
Line receivers need to separate the signal from any noise introduced as the signal was transmitted along the cable. This task is made more difficult because the signal is attenuated by the resistance of the cable, and can be distorted by its capacitance/inductance. External signals (from other cables or radio emitters) also contribute to the unwanted signal. To mitigate these effects all line receivers implement a filter, that limits the slew-rate of the received signal. This low-pass filter removes unwanted higehr frequency signals. A line receiver often implements hystersis to reduce the effect of varying signal levels:
For example a threshold of 6V can be used with an ansynchronous system, such as EIA-232 and in EIA-485, a threshold of 200mV hysteresis establishes two thresholds 200mV apart.
In simplex communications, a driver operates either as an output at the sender or as an input at a receiver. Two wires and two sets of line drivers are sued for full-duplex communications. When half-duplex mode is used, the driver needs to alternate between input (receive) and output (transmit) modes, and only one wire is used.
Note: EIA-232 sends signals using a voltage level that can be larger than the IC supply voltage, this requires the line voltages to be generated by an internal charge pump.
Suppose we wish to transmit the byte "10110010" over a communications link, then we would expect the signal corresponding to the byte to be distorted as it travels along the cable.
In fact, there are many different ways a byte may be distorted, some of which are represented below at three points along the cable. The crucial question is: "Will the receiver be able to recognise the signal it receives and associate it with the correct transmitted bit sequence?"
If the byte is correctly interpreted by the receiver at the end of the cable, we say that we have achieved error-free communication. If however, the receiver sometimes makes mistakes in interpreting the received bits, we say that bit errors have been introduced by the link.
As the bits (actaully bauds) propagate along the cable, the signal will be subject to attenuation - due to signal loss (from the resistance of the cable); distortion - due to the reshaping caused by the cable capacitance and inductance and any noise/interference (which reduces the ability of the receiver to discriminate the baud symbols from other extraneous signals).The lower the energy per baud compared to the noise level, the harder it is to reliably determine the bit value.
The easiest effect to imagine is the reductiin in the signal as the signal travels along the cable with a certain resistance per meter. In this case, some of the signal fades away. In effect, the signal gets weaker, and weaker, the further it has to travel. This is known as attenuation.
Real cables not only have resistance, they also have capacitance and inductance. Thei means the attenuation is not linear, signal components at different frequencies suffer different levels of attenuation. If we view the frequency spectrum of the signal that was originally sent, we would see that some parts of the frequenct spectrum are attenuated more than other parts (typically the higher frequencies are more attenuated). This causes the shape of the received signal to change as it passes along the cable. For a specific cable, the level of distortion could be predicted. Most commonly the signal is band-limited by a filter at the sender/receiver. Sometimes this can be compensated for by applying non-linear amplification before and after the signal is sent (this is known as signal equalisation).
Attenuation and (predictable) distortion is not be a significant problem for error-free communication. Expressed in decibels (dB), loss of signal happens along the length of any cable. It is a natural phenomenon that occurs for any type of transmission: electrical power or network data. The longer the cable, the greater the loss. Some loss also occurs at all connection points along the cable such as connectors. All communications systems also experience some random noise which also appears at the receiver mixed with the received signal.
To provide reliable communications, the sender needs ensure the signal sent results in a signal at the receiver that is above the noise threshold. As a signal becomes attenuated, it can approach the level of noise (i.e., it becomes hard to differentiate the signal from the noise). Understanding this allows a link to be designed that can offer relaible communication with an acceptable maximum bit error ratio.
One way to improve the receiver signal to noise ratio is to send a larger signal at the transmitter. However, a larger signal requires more power. For a given slew-rate, doubling the signal voltage halves the maximum baud rate. This becomes a trade-off between ability to drive long cables (at a lower rate) and shorter cables (at a higher rate). A larger signal also results in more power dissipation, increasing the current-capacity needed for the cable.
Using a large diameter copper wire in the cable can reduce the resistance per metre. This therefore also improves the signal to noise ratio at the receiver. However, this increases the cable weight and can significantly increase the cost of the cable.
Another possibility is to amplify the signal as it travels along the cable. Simply useing an analogie amplifier will however also amplify any noise that was added to the signal before the amplifier. Any noise will grow with each stage of amplification. People often position an amplifier 1/3 of the distance along the cable to optimise this overall signal to noise level at a receiver. However, this approach has limited benefit.
A better approach for digital signals is to use regenerative digital repeaters. At each repeater, the signalm is amplified (and filtered). The received signal is then digitally sampled to decode each baud (i.e. recognises each baude in turn). Once decoded, the digital bauds are sent a fresh down the cable to the next repeater, regenerating a signal at the output of the repeater which is as good as the original sent signal. This method also reconstructs the clock signal, and can be used to retime signals. Unlike analogue amplifiers, digital regenerative repeaters can be plced at any point along the cable, and any number can be used.
By placing repeaters at suitable points along the cable (each time before the signal deteriorates below the level at which it can be reliably decoded), arbitrary long distances may be reached with little probability of bit errors.
See also:
Prof. Gorry Fairhurst, School of Engineering, University of Aberdeen, Scotland (2025).