Robotic voice generator

Posted on

Voice generator scheme

Robotic voice generator does not need the variety of complex chips. You can assemble using only ISD2500 ChipCorder® family and few components.The family of chip-coders ISD2500 firm Winbond contains almost everything you need to record and play back voice messages. As chips have mic preamps with AGC, working with the cheapest electret microphones, the output amplifier that runs on the speaker, memory, oscillator, ADC and DAC.

In a family of four basic models: 2560, 2575, 2590 and 25 120. The number following the numerals 25 denotes a maximum recording time in seconds. The amount of memory chips all the same, and long duration recording is achieved by a lower sampling rate. Ie, the chip with the lowest recording time provides the best sound quality.

Steps to follow when writing

First of all, set switch S3 to the recording mode (low level at pin 27). When you press S2 recording mode, which will end after pressing S2. With the third click again to start recording, etc. So you can continue for as long as the record will be nothing, or until the LED D2, indicating that the memory is full.

Steps to follow when playing

Playback can be started immediately, including the S1 and S3 in the switching mode “play”. Now, with every touch of S2, will consistently play the recorded messages. To record a new message over the old, turn S1, S3 switched to “record”, and pressing the S2, start recording again.

Some flexibility in the organization of the play mode allows you to connect together fragments of messages. Each piece ends with the flag of EOM (End Of Message – End of Posts), which is stored in the memory chip. So, instead of storing entire phrases, such as the “obstacle in front,” is much more efficient to keep an “obstacle”, “front”, “right”, “left”, “rear”. Similarly, for the numbers: from individual elements of the “one” “two”, “hundred”, “point”, etc. You can collect any number.

The Robotic voice generator scheme, managed by a microcontroller, is shown in the figure. Signals A0, PD, / CE and / EOM connected to a microcontroller that controls the robot.

To play you must first set to “0” signal PD. To play the first message, the input / CE must submit a negative pulse. If this level of input A0 “0”, the playback will occur at normal speed, and if “1” – a rate 800 times greater – a sort of “fast forward”. If you want after the first report was made, for example, the third, the microcontroller must set the “1” to A0 and submit a negative pulse / CE, to “squander” the second message with great speed. Then, after waiting for the installation of the flag / EOM at a low level, A0 must be reset to “0” and pulse / CE start playing the third message.

Due to the fact that the pulse duration / EOM may be less than 10 ms, it is better to use it to interrupt the processor, but do not check status / EOM interviews.

Gravatar Image
Robotics Technology , Tutorial and News

Leave a Reply

Your email address will not be published. Required fields are marked *