6 FUNCTIONAL DESCRIPTION
6.1 DETAILED DESCRIPTION
6.1.1 Audio Quality
Winbond’s patented ChipCorder® Multi-Level Storage (MLS) technology provides a natural,
high-quality record and playback solution on a single chip. The input voice signals are stored
directly in the Flash memory and are reproduced in their natural form without any of the
compression artifacts caused by digital speech solutions.
6.1.2 Message Duration
The ISD1740A/50A/60A devices offer record and playback duration from 26 seconds to 120
seconds. Sampling rate and message duration are determined by an external resistor
connected to the ROSC pin.
Table 6.1 Duration vs. Sampling Frequency
6.1.3 Flash Storage
The ISD1740A/50A/60A devices utilize embedded Flash memory to provide non-volatile
storage. A message can be retained for a minimum of 100 years without power. Additionally,
each device can be re-recorded over 10,000 times (typical).
6.2 MEMORY ARRAY ARCHITECTURE
The memory array provides storage for four special Sound Effects (SE) and the audio data. The
memory array is addressed by rows. A row is the minimum storage resolution by which the
memory can be addressed. The memory assignment is handled automatically by the internal
message management system.
The four sound effects occupy four rows of the first sixteen rows in the memory array. The
minimum storage resolution varies with the sampling frequency, as shown in Table 6.2.
Table 6.2 Minimum Storage Resolution vs. Sampling Frequency
6.4 kHz 5.3 kHz
Minimum Storage Resolution 83.3 msec 125 msec 156 msec 187 msec 250 msec
For example, at 8 kHz sampling frequency, the minimum storage resolution is 125 msec, so, each
Sound Effect (SE) is approximately 0.5 seconds long. The remaining memory is dedicated to audio
- 11 -
Publication Release Date: March 28, 2005