Hey folks,
I need to build, because I cant afford to buy, an intercom.
I am going to start wired and possibly look at wireless extension later.

The unit is for a remote gate and I want to avoid having to push buttons if at all possible.

I would like to use a single 'star' config with the potential to have multiple stations active simultaneously contributing to the conversation.
like several speakerphones in a conference call.

I will probably be able to find / build a simple voice quality amps and already know how to implement audio over a balanced pair.

What I dont know is how to prevent feedback if a station is sending and receiving signals at the same time.

I know modules are available to do this that require only a single small speaker that also acts as a mike but I would really like to understand how they work and have a go at building one.

All suggestions welcome,
Al

 

Some work with VOX circuits (voice activates the transmit function) while others utilize differences in frequencies or modulation methods.

Go online to Nutone's website and see if you can pull down any of their schematics, they were always clearly written and often included the theory behind the design.

 

Hi,
I had a look at Nutone's site, thanks marshallf3, some good stuff although I couldnt find much technical detail.
Nice kit, pity I cant afford it ...

Anyway ... I think my question is actually more fundamental than TX RX strategys.

Asume my audio distribution circuit is simmilar to a telephone circuit:-
All instruments share a single pare of conductors carrying DC and a superimposed audio signal.
All instruments amplify the available audio signal and present it to the user.
All instruments accept audio from the user and place it on the line.
[When I was a kid we used to use old GPO telephone hansets, they had carbon dust speaking elliments, to make intercoms bt simply wiring them in series with a big 4.5V battery.] but I digress ...

In a conventional telephone you hear yourself in the handset, feedback dosnt occur because the speaker and mike in the handset do not interact.

But how dose it that work with a speakerfone ....
It cant be a function of the line, thats just a pair of conductors with all the instruments in paralell.

Is it a delay somewhere or something more complicated that cancles any input that looks like it is also output.

Anyone have any idea?

Al

 

To cancel feedback they often use phase cancellation in the enclosure &/or the circuit.

In another thread someone mentioned that they have been known to introduce a minor delay in the signals.

 

This datasheet gives the information needed to undertstand the principles behind analogue speakerphones: http://www.datasheetcatalog.org/datasheet/motorola/MC34118P.pdf

 

That looks like plenty to get me going ...
More complex than I thought, I wonder if they are available retail.

Thanks Rogs
Al

 

The attached file is from the 1994 National Semiconductor data sheet for the LM759. Too bad both the LM759 and its lower power brother are both obsolete and impossible to find. Concept should work with most any audio amplifier since the bi-directional function is in the front end.

 

Hi,

Here is a circuit I simulated awhile back. It just uses any opamp for audio. This is a basic, no-frills circuit to demo the magic of the plain old telephone.

Have Fun,
Ifixit

 

BillB3857 .... that great thanks.

I love the symplicty of that, th top and bottom of the balance pot move in opersit directions with a local input effectivly cancling in restect of its midpoint.

The top feeds the line

When a signal changes only the top of the balance pot its midpoint will also change frrding the amp.
That is correct is it?

Al

 

Also interesting Ifixit thanks for that.
I cant see how that telephone sym would cancel local input I am asuming it dosnt is that correct?

Very basic question now but I just realised I dont know .. Can I build an audio amp with only 1 rail .. if so how would the output be arranged ... a DC bias wouldnt work because of the constant current through the speeker right?

Al

 

There seems to be 2 directions developing here.
Certainly the intercom options shown are much simpler than the traditional speakerphone options, but they won't work with a standard 'telephone line' single pair.
In addition to the audio line, which is shown unbalanced, there needs to be a power supply. So a minimum of 3 cores (power and audio sharing the same common) and the possible noise problems of an unbalanced audio system, over a long run.
You will also have fun trying to introduce more units, simultaneously onto that kind of 'line', I suspect. Might be fun trying though!

The speakerphone option does function on a single 'telephone' type pair -- and you can probalby introduce 3 or 4 units onto the line before it all 'collapses under it's own impedance', so to speak!
Unfortunately, it's more complex. (I've also just noticed that the link I posted above only links to the first page of that data sheet. This:
http://www.freescale.com/files/timing_interconnect_access/doc/data_sheet/MC34118.pdf
is the whole thing, AFAIK).

The main problem is dealing with not only direction control, but background noise. With audio determining which direction the line is active, you don't want a continuous background noise 'locking' the system in one direction. Hence the extra complexity.

Now if you want to really address the problems with modern technology, take a look at echo cancelling full duplex digital speakerphones:
http://www.vocal.com/data_sheets/fullduplexspeakerphone.pdf

That really does get a bit hairy!

 

All good points I feal ... bear in mind I am talking gate intercom so I can put up with a lousy bandwith and background noise to some extent but not if it locks the system out.
I like the way the simple intercom works, essentally cancling output for any input 'it' generated whilst leaving what it puts on the line unchanged but it looks as if it may be hard to adjust and would probably drift.

I think I will try the concept with seperate amp stages rather than the transistor, anyway I have amps I think.

I am going to run cat 5 and use baluns, OK 600:600 line transformers, to couple the audio pair, I have done this loads of times in the past to extend 75Ohm and 1K unbalanced lines, the big advantage is that you have a comom mode choke at both ends so its usuly prety good in respect of noise.

Power will just sit on another pair because it can !

It all good stuff this, thanks so much.
If simple isnt good enough then I will now be abe to try complex.

I could still do with some basic cuidance RE opamp based audio amps, preferably on a single supply.
Always bought or was working with existing kit in the past.
It will be a good opertunity to laern.

Al

 

See Attached sketch for some ideas on a single supply, op amp based version of BillB3857's circuit.
Uses an electret mic capsule, and a more modern power amp ic.

I've not actually built it, but it should at least give a starting point!


Further edit -I have built it now. Works OK!

 

That looks as if it contains all the concepts I need to look at ..
Thanks, it will be a big help
Al