|
|
|
| Authored by: Ian Al on Tuesday, October 16 2012 @ 06:24 AM EDT |
Resistor/Transistor (RTL) and Transistor/Transistor (TTL) logic use a +5Volt
power supply. The '0' symbol is represented by a voltage anywhere in the range 0
Volts to 0.5 Volts and the '1' symbol is represented by a voltage anywhere in
the range 2.7 Volts to 5 Volts. However, input ports accept a slightly wider
range of voltages in order to be sure that a valid output voltage is accepted by
inputs.
CMOS gates can be given supply voltages between 5V and 18V. The voltage ranges
related to each logical symbol are usually expressed as a proportion of the
voltage supplied to the chip rather than an absolute voltage.
Motherboards have a high-power 3.3 Volt supply to run the processor. This is
also the voltage required by mobile phone ARM processors and by ARM controllers.
That suggests that the processor logic voltages are reduced in a pro rata
manner.
Communication along busses can use the same logic values. However, the longer
the communication buss then the more important it is that the signal has no DC
component because of losses in the communication path. This is usually achieved
by using +ve and -ve logical voltages.
The same thing applies for high, logic-circuit data rates. Emitter Coupled Logic
(ECL) define the symbols 0 and 1 differently from the DC values in TTL and CMOS.
In fact, the outputs switch current between one of two wires. It is a form of
balanced or differential transmission circuit.
The earliest application of this principle of which I am aware is teleprinter
communication. It uses +80Volts and -80Volts. The International Alphabet No. 5
means that most characters have a DC component. However, at 50baud transmission,
the sum DC component is around zero volts over any length of time and the
communication efficiency is not reduced.
For modem transmission, the transmission path only passes the frequency band of
300 to 3400 Hz and so the DC logic levels are used to modulate tones in this
frequency range. AC signals formed in this way have no significant DC component.
However, it is still binary signals being represented and so one gets a sort of
off-white noise effect when listening to the data stream. As the modulation
technique gets more complex than simply switching between two frequencies, then
the frequency components generated use more and more of the available frequency
band which means higher data rates are possible.
Telephone dial pulses are at 10 pulses per second. This low rate means that a
50v supply at the exchange via a relay will reliably detect the dial short
circuiting and open circuiting the line. That's how the very old call box fraud
worked. Tapping the receiver rest mimicked the dial pulses. It enabled dial
pulses to be transmitted even though the dial mechanism was disabled until coins
were inserted. However, the dial was effectively sending a +50v signal to the
exchange.
The combination of inductive relays, relatively high, pulsing, DC currents and
voltage meant that the dial pulses induced interference in adjacent lines.
The pairs of tones used in two tone diallers, having low current and no DC
component do not have this problem.
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
|
|
|
|
