Are there any "Data Over Voice Chip" exists?

[karaapak]

I wanna send data over audio,
i am searching methods for this but
if there is any readily available solution for this purpuse i can go for it.

Higher bitrates are more appretiated.
I need at least 8 kBps.

It is a hobby project,
i strongly wanna use and learn data over voice method.

 

[Ian Rogers]

Of course there is.... The old 56k modems did just that.... Also all the old home computers did....

MODEM chips are still largely available just search for modem chip!

 

[karaapak]

Thanks, i couldn't think searching it as a "modem chip"

 

[karaapak]

I searched the modem chips but i think it is not the solution for me.
If i understood modem chips' operation correctly,
they can do much more complicated operations than just sending data over sound.

What i wanna do is:
For example;
on one hand, i have a message board, maybe graphical,
i will send some text or graphics to it,
this message could be as high as 128 kbytes,
on the other hand, i have a device that can play sound,
this sound could come from a cellphone music player, a mp3 player, even call from an another cellphone, radio receiver, etc.
Finally, with playing sound i could send data to my message board.

Message board will be made by me,
sound that need to be played will be generated by me.

If there is some other cheap methods for data over voice, i wanna learn.
If it would be a solution for some methods, i can program and use mcu.

 

[Tony Stewart]

karaapak are you going to compensate for echoes, background noise, bit errors that turn into pixel errors ,, retry algorithm? for Forward error correcting?

Get with the program !

IRDA2 is the way to go. >1Mbps in 1 meter. 20 yr old technology. Use a method of encoding that ensure at least 1 transition per byte either synchronous packet burst with preamble and frame sync with addressable units or Serial Async. then make it by directional.

Or if you want slower.... use IRDA 115kbps with a USB serial port or a PIC Chip and FLash memory.

 

[karaapak]

At the begining i just wanna see the data to be send to the receiver by sound,
but at the end, of course i want to achive error free data over voice communication.

Thanks for the recommandation about irda,
i had searched infrared about 2-3 years ago and
i had made a working example, i think i learned it.

Why i wanna learn data over voice,
for transmitter side there is readily avaiable commonly used devices,
why not to use them.
At the receiver side, if we could solve it just with a simple microphone circuit and
a little bit an algorithm for mcu or a cheap one package solution chip,
it would be a preferrable solution for some wireless communication.

 

[Ian Rogers]

It is fairly simple... You just need two frequencies ... Say if you have 15Khz represent HIGH and 5Khz represent LOW.. The receiver just differentiates the signal and reconstructs the data.... Be aware that using audio will be quite slow...

 

[dougy83]

You can use an FSK demodulator IC such as:
* **broken link removed**
* **broken link removed**

The signal from the microphone is amplified and fed into the input of the FSK demodulator IC. The output contains the data that you sent

 

[Tony Stewart]

All POT type modems use 1200/2400 Hz but in order to get more than than 200 Baud or 1200 baud rates you have to use a protocol to compress bandwidth using n level amplitude and phase modulation with trellis encoding used all the way up to 64kbps. The training in modems is to equalize the group delay & spectral flatness error distortion in a 4kHz bandwidth. FOr this they used hardware DSP's , then later Winmodem software drivers. At 1200 bd I recall they used QPSK, then went up to 64 levels of amplitude and phase, which requires a higher SNR with forward error correction FEC. to get 64/4 = 16 baud per Hz compression. FCC complained about 64kbps so it was curtailed to 56kbps.

YOu can use FSK and tradeoff SNR and frequency response distortion with deviation ratio and SNR to optimize baud rate. Maybe get 8kbps. which is tight and like PSK or Manchester biphase encoding at 4 to 8 kHz is simple, beyond this you will get group delay distortion.