I was perusing this circuit on Redcircuits.com when I noticed the capacitor C4 seemed to be backwards – the plus should be on the left. I contacted the author and he said that he would correct it.
The idea of talking and listening at the same time “full duplex”, which is technically a communications term, not an audio term, is a very old concept. Since the early days of telephones, the handset has had this feature for more than a hundred years. But for some reason it was seldom if ever applied to an intercom system.
In reality, the way it works is the amplified audio from the microphone is fed onto a bus with the speakers of both (or all in the case of 3 or more stations) outputting the audio to the listening area. The audio from the local microphone is cancelled out so there is no feedback howl, and you don’t hear your own voice in the speaker. This is called anti-sidetone in the telephone lingo.
In the schematic he shows the two stations connected together with shielded cable. There can be some major issues with this method. If the stations are separated by some distance, say 50 feet or 15 meters, the difference in the power line ground potentials can introduce hum between the two ends. One solution is to run the power supply wires from the master to the other station(s) and leave the other stations ungrounded. Hopefully this will help a lot in reducing the power line hum.
Another solution is to isolate the stations with a 1 : 1 transformer. This has safety advantages if the transformer is able to withstand power line voltages or even higher voltages. It also allows the connecting cable to be a much cheaper twisted pair telephone wire instead of more expensive shielded cable. And since the line is balanced, it can be run over considerable distances. This is how the telephone company can have telephones thousands of feet away from the central office with very low hum and interference.
One transformer that works is the 600 ohm 1 : 1 transformer that I wrote up in my blog. These have four windings so I suppose one could connect four intercom stations together. But no real need to do that, since every intercom station can have a transformer, and there will be a single twisted pair “bus” running between all stations.
The original schematic shows a single collector resistor R5 for both stations. If transformers are used, there will be no DC continuity between stations, so each station will have to have its own collector resistor. R5 is 2.7k, so each station could have its own 5.6k resistor.
Other issues The author uses the TDA7052 audio power amplifier chip to drive the speaker. This makes good sense because it’s inexpensive and simple, taking up very little circuit board space. However I’ve never seen these ICs for sale in any electronics store such as Radio Shack. I can get the LM386 audio power amplifier chip for a reasonable price from R-S and other similar stores. Or I can also build a simple audio amplifier out of discrete parts, a few transistors, resistors and capacitors, that will do an adequate job. The LM386 is a good choice for 1 watt of audio output which should be loud enough with a decent speaker.
One other issue is microphone volume. There are two factors here. Neither station has an adjustment for the microphone volume. I’ve found that there is a wide variation in the sensitivity between electret condenser microphones from the same maker. One may be high, the next may be low. This may have unwanted consequences for the intercom. Another factor is how close do the participants want to be in order to be heard well. If one station is in a high noise level environment, they may want the mic sensitivity to be very low, and will have to speak close to the microphone. The other station may be in a large room, and the participant may want to be heard at various locations, both close and far.
For a two station setup, these issues may be adequately managed by the volume controls as shown. There may be a problem for more than two stations, or if there is a long distance or there are large losses between stations.
One other factor that is often not addressed is the low frequency response of the telecom system. Low frequencies below 300 Hz do not carry much voice intelligence. This intercom system has good low frequency response, and that may be a problem if there is a lot of background noise from machinery or other low frequency generators. In that case, it may be necessary to limit the low frequency response. Reducing C1, C3 and C4 would be a good choice.
Back to experimenting…