NUVOTON NM-ISD3900

ISD3900 is a multi-message record chip featuring digital compression, comprehensive memory management, and an integrated analog/digital audio signal path. Its message management function is designed to simplify the message recording process, enable address-free operation, and streamline code development for playback-only applications. The chip utilizes Winbond 25X/25Q series flash memory to provide non-volatile audio recording and playback functionality in a dual-chip solution. Unlike other record chip families, the ISD3900 offers an I²S digital audio interface, faster digital recording speed, higher sampling frequencies, and a signal path with SNR equivalent to 12-bit resolution. The ISD3900 can receive digital audio data via I²S or SPI interface. When I²S input is selected, it replaces the analog audio input and supports sampling rates of 32, 44.1, or 48 kHz depending on the clock configuration. When the SPI interface is selected, the audio data transmitted must use one of the sampling rates supported by the ISD3900. The chip integrates analog audio input, an analog audio line driver, and a speaker driver output. Its two analog audio inputs are: 1) AUXIN, whose gain is fixed via SPI command; 2) ANAIN/ANAOUT, a fixed-gain amplifier whose gain is set by two external resistors. ANAIN/ANAOUT can also be used as a differential microphone input in conjunction with an AGC circuit configured via SPI command. There are three forms of analog output: 1) AUXOUT is a single-ended voltage output; 2) AUDOUT can be configured as either a single-ended voltage output or a single-ended current output; 3) BTL is a differential voltage output.

• Supports Winbond 25X/25Q series SPI flash as external memory
• Addressing capability up to 128Mbit; provides 64 minutes of recording time based on 8kHz/4-bit ADPCM encoding
• Fast digital programming with rates up to 1Mbit/s, primarily limited by flash write speed
• Message management: executes address-free recording, allocates memory for new recording requests, returns start address to host via SPI upon completion; stores pre-recorded audio using high-quality digital compression; index-based simple command playback; executes pre-programmed macro scripts designed to control device configuration and play back voice prompt sequences and message recordings
• Sample rates: seven recording and playback sampling frequencies at a given master sample rate. For example, when the device operates at a 32kHz master sample rate, available recording and playback sampling frequencies are 4, 5.3, 6.4, 8, 12.8, 16, and 32kHz. For I²S operation, available master sample rates are 32, 44.1, and 48kHz, with recording and playback sampling frequencies scaled accordingly
• Compression algorithms: recording supports ADPCM, μ-law, differential μ-law, and PCM; pre-recorded voice prompts support μ-law, differential μ-law, PCM, and enhanced ADPCM; variable bit rate optimized compression achievable at given SNR and background noise level specifications
• Oscillator: internal oscillator with internal reference at 2.048 MHz, ±10% deviation; internal oscillator with external resistor at 2.048 MHz when Rosc = 80kΩ, ±5% deviation; external crystal or clock input; I²S bit clock input; supports crystals and resonators for standard audio sample rates including 2.048, 4.096, 8.192, 12.288, and 11.2896 MHz
• Inputs: AUXIN, analog input with gain configured via 2-bit gain control through SPI command; ANAIN/ANAOUT, analog input with gain set by two external resistors from ANAOUT to ANAIN, or as microphone differential input; digital AGC for digitized data from analog inputs
• Outputs: PWM, Class-D speaker driver capable of directly driving 8Ω speaker or buzzer; PWM, Class-D speaker driver with typical output power; AUDOUT, configurable as current or voltage single-ended line driver; AUXOUT, single-ended voltage output; BTL, differential voltage output with typical output power
• I/O: SPI interface including MISO, MOSI, SCLK, SSB for command and digital audio data transfer