can i extend the dtmf digits from 16 to 25

[hejas.sp]

hai,

i am trying to do a project where it requires 25 different sound signal for 25 different error signal.
i need to encode the error signal to sound signal and also need to decode the same.
shall i use dtmf for this . by just adding 2 another freq to it.
so plz suggest me circut for the above requrement..


thank you

 

[betwixt]

It will not meet DTMF specifications if you do that but it is fairly easy to do. Use two conventional DTMF circuits, one with the 'normal' 3.579MHz clock and one with a higher frequency clock. The tones generated are proportional to the clock frequency so you will produce a whole new set of 16 tones giving you a potential of 256 signals. You don't have to use all these of course and in the decoding circuit you can ignore the output of bits you are not using. Beware that if you are sending these tones down a phone line, they still have to fall withing the voice band of frequencies and you might hit problems with some tones being used for internal purposes in the telephone network.

Brian.

 

[hejas.sp]

thank you for replaying sir.

i think am using a mobile phone for transmission of sound signal generated by dtmf. ie i am using a speaker at the dtmf encoder and this sound is transmitted over normal mobile call . can you plz suggest any circuit diagram .
once again thank you..

 

[betwixt]

Look at the data sheets for the HT9170 and the HT9200A IC, these are standard DTMF receivers and transmitters. The schematic are on the data sheets, all you need to do is make two transmitters and two receivers, one of each using a standard 3.579MHz crystal and the others using a crystal of around 4.8MHz. Link the outputs of the two transmitters together through 1K resistors and you should get one keypad producing normal DTMF tones and the other shifted tones, a little higher in pitch. Each receiver should only pick up a valid signal from the keypad using the same crystal frequency so you have 2 x 8 = 16 bits of data out.

Brian.

 

[hejas.sp]

thank you sir . i will try this ...

 

[YEAR_]

I think DTMF Frequencies are carefully selected so that no aliasing can occur so that notes are not misinterpreted , How is the mixing of two DTMF tones work smoothly? It is better to generate a whole new set of frequencies with a micro controller.

 

[betwixt]

As I pointed out, the extra tones will not meet DTMF specification but this is inevitable when 'add on' tones are used. The original DTMF frequency pairs were chosen because they were statistically unlikely to occur together in a human voice although I personally know someone who's voice regularly operates a DTMF decoder!

The frequency I suggested does produce a new pair of tones exactly as a microcontroller might do and by choosing a 4.8MHz crystal in conjunction with standard DTMF transmitter and receiver ICs, the new low group tones fall between the existing high and low group and the new high group are above the existing high group so it spreads the 16 tones evenly throughout the audio spectrum and maintains as much seperation as possible between individual tones.

To extend it beyond that many tones will require complex sound generation and detection or a move to a serial tones stream instead of parallel.

Brian.