31

In this video, when A wanted to talk to B, a microphone was connected to a distant speaker using two wires (A → B).

When B talked to A, a copy of above was used in the other direction (B → A) using two more wires.

This required four wires:

Enter image description here

But then it says the two-way communication could be made by using just two wires:

Enter image description here

How was this possible? Wouldn't the current from the microphone affect the speaker on the same side?

Peter Mortensen
  • 1,676
  • 3
  • 17
  • 23
across
  • 1,100
  • 11
  • 25
  • 12
    You actually want the local speaker to respond to the microphone, otherwise the user keeps raising their voice, because they feel like their voice isn't carrying. But you don't want it to give respond too strongly or the user will keep lowering their voice until the far end user can't hear them. – The Photon Jan 19 '21 at 02:15
  • 4
    I was just reading about this. They call it a hybrid circuit. Googling that may shed some light. In this day and age it is certainly possible to subtract out the local contribution from the microphone using op-amp circuits. In the old days they used transformers. – user57037 Jan 19 '21 at 02:17
  • 3
    This may be of interest: https://www.sound-au.com/appnotes/an010.htm – user57037 Jan 19 '21 at 02:19
  • 4
    Old phones had transformer. When signal comes from microphone to line it also added in reversed phase to speaker, so you do not hear youself. On line signal is mix from two sides signal. – user263983 Jan 19 '21 at 02:19
  • 2
    In fact the "hybrid" circuit still exists in semiconductor form - these days it's called a "2 to 4 wire converter". The Elliot sound link above goes into the deatils very well. – danmcb Jan 19 '21 at 11:25
  • In fact you can get away with a *single* wire in some conditions — it's called “ground return”. – Reid Jan 20 '21 at 00:49
  • The answer https://electronics.stackexchange.com/questions/413798/how-is-it-that-two-electric-currents-can-travel-in-opposite-directions-on-the-sa/413799#413799 may be more fundamental in regarding why one can transmit information in opposite directions over a single channel (yes it also applies to waves not just wires and in optical, electrical and also acoustic domain): – Andreas H. Jan 20 '21 at 07:57
  • @ThePhoton, Re, "You...want the local speaker to respond..." There's a name for that. It's called "[side tone](https://www.collinsdictionary.com/us/dictionary/english/side-tone)." – Solomon Slow Jan 20 '21 at 22:40
  • TLDR: The instantaneous current in the two wire "loop" is proportional to the sum of the signals from the microphones at either end. The "hybrid block" (the transformer thingy shown in several of the answers below), effectively _adds_ the local microphone signal to whatever current is flowing in the loop, while simultaneously delivering the _difference_ between the local microphone signal and the loop signal to the local speaker. That difference that's delivered to the speaker is mathematically equal to the microphone signal from the far end.\* – Solomon Slow Jan 20 '21 at 22:46
  • \* That is, assuming that there's only two telephones in the loop. But if there's more than two, it still just works. The signal on the loop is the sum of _everybody's_ voices, and the signal that's delivered to your local speaker is everybody's voice minus your own voice. (plus, that little bit of side-tone to keep you from wanting to shout.) – Solomon Slow Jan 20 '21 at 22:49

4 Answers4

36

How was this possible? Wouldn't the current from microphone affect speaker on the same side?

The modern telephone is wired in a Wheatstone bridge arrangement like this: -

enter image description here

So, if you ensure that the telephone network line impedance (\$Z_{LINE}\$) is controlled then, theoretically, any signal produced by the microphone is dramatically reduced into the local earpiece. It won't be a perfect cancellation but it'll be pretty good.

Amended picture originally from here. It might be easier to understand this diagram: -

enter image description here

Picture from here.

And, all throughout the network there are line amplifiers that need to translate from 2 wire to 4 wires so that amplifier circuits can be added: -

enter image description here

Picture from here. Then another hybrid transformer is used to reconvert the 2-way (4-wire) amplified signals back 2-wire: -

enter image description here

Here's an example of an early telephone anti-sidetone circuit using the same principle as the hybrid transformer: -

enter image description here

Picture from this website.

Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • 2
    so the speaker is placed at the center of a balanced wheatstone bridge, then the signal from the mic would never reach the speaker. Clever! – across Jan 19 '21 at 12:34
  • @across correct – Andy aka Jan 19 '21 at 12:36
  • Did they have op-amps in those days? – user253751 Jan 19 '21 at 14:55
  • The first op-amp came out in the 1960s (uA741 circa 1968) but transistor amplifiers would have been used and, before that valves/tubes @user253751 – Andy aka Jan 19 '21 at 14:59
  • 1
    Not the very first ones. But OpAmps were build with tubes since the 40s. – jusaca Jan 19 '21 at 15:00
  • 2
    @Andyaka Normal telephones didn't have amplifiers until the 1970s or later. – grahamj42 Jan 19 '21 at 17:48
  • 1
    @grahamj42 correct and, the hybrid transformer was used before amplification. Of course these needed carbon granule microphones to get anything like a line level signal from the microphone. Not great but, as we who remember, know it worked (distortion and all). – Andy aka Jan 19 '21 at 18:33
  • Note that the lack of complete cancellation is considered a feature, not a bug: it gives the user the cues they need to avoid shouting into the microphone. – Mark Jan 20 '21 at 18:48
  • @Mark too much sidetone really does degrade telephone return loss and, as someone who has designed about 4 POTs in the 1980/90s, I can report that the British standards (BABT and BT) never had anything in their pages about ensuring sidetone is above a certain low level. In fact they had tests that tried to measure if there was too much (called a corner test). And, of course the tests were very hot on return loss and, too much sidetone can wreck return loss. – Andy aka Jan 20 '21 at 19:07
28

Remember that the telephone was invented over 30 years before the triode vacuum tube, and over 70 years before the transistor. Changing, combining, and subtracting audio had to be done with resistors, capacitors, and transformers - only. Early phones had a "network" inside, of which the major component was the hybrid transformer. This had four-to-six windings, connected such that the signal from the microphone both drove the phone line and was subtracted from the signal going to the earpiece. The subtraction was not perfect intentionally, allowing the talker to hear themself in their own earphone. This let the talker's brain act as an automatic volume control. The leaked audio signal is called sidetone.

Search for telephone network schematic to see examples.

AnalogKid
  • 16,865
  • 1
  • 13
  • 25
9

It is done with a circuit that sends the microphone signal to wires, but removes your own microphone signal from the wire signal before sending it to the speaker, so that only thing that can be heard from the speaker is the signal from the remote mic.

Such a circuit is called a telephone hybrid, so you can find more info if you like.

Justme
  • 127,425
  • 3
  • 97
  • 261
9

Ah yes Plain Old Telephone System or POTS is a thing of beauty. Who still uses it?

In order to have two baseband signals over shared 2 wires, one way is current modulation and the other uses voltage-modulation over a controlled impedance using hybrid centre-tapped series-parallel transformers to perform this.

The early mics were very near-field sensitive carbon-diaphragms that modulated impedance.

enter image description here credit

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182