you use a hybrid to separate the transmit signal and the receive signal. this is done so that you can amplify either signal since amps only work one direction only.
Phones use 2 wire loops since that is the most economical to the phne companies (putting the copper cables in the ground or up on poles is expensive) but the transmission of signals over long distances is done by what is refered to as 4 wire operation (2 wires a tip and ring or a signal and ground are used for each direction)
I will try explaining how the 2/4 wire conversion worked with the older analog devices so you can understand the conversion that takes place inside of a hybrid chip on the newer solid state circuits.
the two wire loop is sent into a hybrid that is immediately fed into a transformer, primary coil is the 2 wire and two secondary coils for the TX and RX pairs, the result of this is that any signal fed into any of the pairs result in 1/2 of that signal being output by the other two pairs (normal signal loss) There is a feedback control that prevents the receive pair from sending the same signal back on the transmit pair. the Transmit pair is then fed into an amplifier to recover the signal loss of the transformer. the receive pair was amplified prior to entering the transformer to prevent loss thru the transformer. the end result is that you have one pair (Transmit) that only has the signal being sent and one pair (receive) that only has receive signa. the 2 wire pair is a mix of both.
with a circuit divided into 4 wire you can now use multiplexer equipment to place many circuits on one path be it fiber cable, microwave, satellite, etc.
you were asking earlier about the 180Hz-3.2Khz bandwidth. the intiligence of a voice conversation takes place between about 300Hz and 3KHZ, some people talk with higher or lower voices and there are harmonics in a normal converstaion that you hear when you are talking one-to-one but are not really needed to pass the intelligence of your conversations. you really notice this when you are listening to a radio station that puts someone you know that has called in on a telephone call on the air, the quality of the voice is changed but you still know what they are saying.
this reduction of the bandwidth is done to allow the maximum number of circuits to be multiplexed together on an available bandwidth. again going to methods used in the past a circuit would be modulated with a "carrier" that caused the circuit to shift in frequency upward by 4,8,16 etc KHZ 4 KHZ bandwidth shift was used to provide a "buffer" between circuits but still maximize the number of channels. The old microwave systems used groups of 12 voice channels for a total of 64KHZ per group. groups themselves were multiplexed again using Frequency Division Multiplexing into 5 group supergroups.
Newer diagital multiplexing still uses the basic 4KHZ bandwidth per channel but samples that 4KHZ bandwith 8000 times each second and assignes the voltage value at each of those 8000 samples a value from 0 to 255 and sends those sample values in a serial stream of 64,000 bits per second. at this point the circuit is in the digital world and multiplexing is just a simple case of taking smaller and smaller samples of each bit and streaming them with other sample streams (I am really simplifing this part since it wasen't part of your original questions). the receive circuit signal is taken at 64,000 bits per second and separated into 8 bit blocks, converted to a value of 0-255 and then a voltage at that value is sent out on the circuit for 1/8000th of a second, whereupon the next voltage sample is sent. this receive signal is sent into the hybrid receive amplifier-thru the transformer-and out the 2-wire pair of the hybrid.
OK I got a little further than the hybrid but take a look and I hope I have not thrown up too much confusion to you on this.
----------------------------
'Rule 29', "The enemy of my enemy is my enemy's enemy. No more. No less."
----------------------------
JerryReeve
Communication Systems Int'l
com-sys.com
