Intelligent voice controlled doll circuit

This example introduces an intelligent voice-controlled doll that can not only talk to you, but also truly understand your words. She
will answer whatever questions you ask, will not answer questions you don’t ask, and can even communicate with people. This magical toy applies speech recognition and speech recording and playback technologies.
Voice integrated circuits have vigorous vitality and are currently undergoing continuous development and expansion. Gradually penetrate into people's life and study.
    Referring to the circuit shown in the figure, readers can not only make an intelligent voice-controlled doll by themselves, but also technologies such as the voice-controlled dialing
function of mobile phones and automatic stop announcers on buses are also made using similar principles.
    (1) Working principle of circuit chip This circuit is mainly composed of two chips. The principle is introduced as follows:
    1) H1.7300 is a speech recognition circuit. It has a microphone amplifier, A/D converter, key memory, speech processor and I/O
controller and other circuit units, HL7300 can recognize 12 channels of voice commands. The recognition phrase must be input in advance before use (the time does not exceed 1.5s).
The chip converts the recognition phrase into a digital signal and stores it in a dedicated memory. When using When the voice command issued by the speaker is consistent with the recognition phrase stored in a certain channel in the chip
, the corresponding output terminal outputs a high level. The APR9600 chip is a voice recording and playback circuit. The recording and playback time is 32~60s. The specific
value is related to the size of the external oscillation resistor R21. When 40k0 is selected, the recording and playback time is about 48s.
    2) The APR9600 chip stores sound signals in a dedicated flash memory in the form of analog quantities. Due to the use of analog/multi-level storage
technology, it can avoid the A/D and D/A conversion processes of ordinary digital recording circuits. distortion caused by quantization and compression, so the sound quality is very good. The non-volatile flash memory used in the circuit
is not afraid of power outage, and the stored voice will not be lost due to power outage. The entered information can be saved for 100 years. The chip directly
drives the speaker, and the sound output power is about 125mW. APR9600 has a very flexible control method.
Parallel direct access and serial sequential access can be achieved by setting the levels of @, @, and ⑧ pins. Here, it is connected into parallel 4 segments. In the parallel control state, each period is the same. So
each segment is 12 seconds. The content recorded in 4 segments can be played separately. You just need to use the small question as the recognition key w of the speech recognition circuit, and the answer
as the recording of the speech recording and playback circuit. In this way, as long as different small questions are asked, the corresponding output end triggers
the recording circuit through the interface circuit to send out the recording of the correct answer after being recognized by the recognition circuit.