CML Microcircuits, innovator and provider of low-power semiconductors for Global Wireless and Wireline communication market has introduced new RALCWI (Robust Advanced Low Complexity Waveform Interpolation) vocoder IC.
The product is the result of a co-operation between CML Microcircuits and SPIRIT DSP.
The new CML IC is named as the CMX618 based upon SPIRIT's proprietary Low Bit-Rate Vocoder technology.
The co-operation combines CML's 39-year history of developing mixed-signal semiconductors for professional and leisure communication applications, with SPIRIT's extensive experience in embedded voice products.
CMX608 and CMX618 are flexible, high integration, high performance, RALCWI (Robust Advanced Low Complexity Waveform Interpolation) Vocoder ICs, providing near toll quality voice at low bit rates.
There are 3 programmable Vocoding bit rates: 2050bps, 2400bps or 2750bps, which when combined with the embedded FEC, provides an overall bit rate of 3600bps.
FEC packet lengths of 60ms and 80ms are supported providing maximum protection to channel errors. The CMX618 includes an integrated voice codec, providing a complete analogue voice-to-low-bit-rate vocoded data function.
The CMX608 and CMX618 are free from any license and royalty costs, fully meet the requirement of high performance low bit rate digital voice applications and are available in very small 48pin VQFN and LQFP packages.
The CMX608 and CMX618 are near toll quality half duplex Vocoders.
The voice encoding/decoding is performed by a new, low bit-rate algorithm technology called RALCWI (Robust Advanced Low Complexity Waveform Interpolation).
The data rate is selectable between 2050bps, 2400bps or 2750bps all without FEC or 3600bps with FEC. The frame length is selectable from 20ms, 40ms, 60ms or 80ms all without FEC and 60ms or 80ms with FEC protected frames.
The integral Viterbi decoder can optionally use soft decision metrics to improve its decoding ability if the data signal is applied in a 4-bit digital representation of the received/demodulated analogue signal.
The FEC function can also be used separately, so that data can be decoded/error-corrected then re-encoded and forwarded on. This is particularly useful in voice repeaters, providing optimum voice performance on the overall up/down link and the maximum range.
RALCWI technology uses different proprietary signal decomposition and parameter encoding methods, ensuring high speech quality at high compression ratios.
The speech quality of RALCWI, as estimated by independent listeners, is similar to that provided by standard codecs running at bit rates above 4000 bps.
The Mean Opinion Score (MOS) of speech quality for this vocoder is around 3.5-3.6.
This value was determined by a paired comparison method, performing listening tests of developed and standard speech coders.
The RALCWI Speech Encoder/Decoder operates on a frame-by-frame basis. The source speech frame consists of 160 samples of linear 16-bit PCM sampled at 8 kHz.
The Speech Encoder performs speech analysis at a high time resolution (8 times per frame) and forms a set of estimated parameters for each speech segment. All of the estimated parameters are quantised to produce 41-, 48- or 55-bit frames, using Vector Quantisation (VQ) of different types.
All the vector quantisers were trained on a mixed multi-language speech base, which contains speech samples in both Eastern and Western languages.
All speech parameters are coded with high efficiency by the Speech Encoder. Sensitivity to errors is estimated for each output bit and a bit frame is split into two parts: more sensitive to errors and less sensitive to errors.
To protect the speech parameters against channel errors without reducing speech quality, the more sensitive bits are protected.
Such protection is provided by the FEC Encoder/Decoder engine. The Speech Decoder de-quantises the parameters and re-synthesises the speech. Reconstructed speech is filtered and presented to the output gain block.
The CMX618 available from Caelera includes a microphone input connected to an ADC, and a DAC driving an earpiece output; a variable gain stage is associated with each converter.
The differential input stage includes 20dB of switchable gain, for use as a microphone pre-amplifier, while the differential output stage includes 6dB of switchable gain. By using the DAC output stage in a differential configuration, a further 6dB of gain can be achieved.
The CMX618 is a self contained RALCWI Vocoder only requiring control from a host microcontroller, therefore providing small footprint.
Control of the CMX608 and CMX618 is via a serial port (C-BUS), from the host µC. The CSEL pin allows the selection of an internal or external CODEC.
DTMF and single tones can be successfully transported through the Vocoder by special control codes.
The CMX608/CMX618 also features a DTX (Discontinuous Transmission) mode, where an integral Voice Activity Detector will send SID (Silence Insertion Description) data to the decoder, which will perform Comfort Noise Generation.
The CMX608/CMX618 flexibility extends to the selection of a Xtal or external clock source.
A number of Xtal/Clock frequencies in the range from 9.6MHz through to 24MHz can be pin-selected.
- Near toll quality RALCWI coding algorithm
- Half-duplex operation
- Multiple bit rate modes:
- 2050, 2400 or 2750 bps
- 3600 bps with FEC enabled
- CMX618 Only:
- Integrated audio codec
- Integrated input and output filters
- 4-bit Viterbi soft decision decoding
- 20ms Vocoder frame length
- 60ms and 80ms FEC frames
- No licensing or royalty payments
- Ancillary audio functions:
- Voice activity detector
- Comfort noise generator
- DTMF transport
- Single tone regeneration
- 1.8V Low power operation, 3.3V tolerant I/O
- Small 48-pin LQFP and VQFN packages.
- Digital professional PMR/LMR (DMR)
- Digital leisure PMR (dPMR)
- Digital trunking
- Voice scramblers / voice encryption
- Lift systems
- Military field radios
- Secure telephone
- Voice storage / playback:
- Voice annunciation
- Answering machines
- Alarm systems
- Voice pagers.