5

®

PCM63P

THEORY OF OPERATION

DUAL-DAC

COLINEAR ARCHITECTURE

Digital audio systems have traditionally used laser-trimmed,

current-source DACs in order to achieve sufficient accuracy.

However even the best of these suffer from potential low-

level nonlinearity due to errors at the major carry bipolar

zero transition. More recently, DACs employing a different

architecture which utilizes noise shaping techniques and

very high oversampling frequencies, have been introduced

(“Bitstream”, “MASH”, or 1-bit DACs). These DACs over-

come the low level linearity problem, but only at the expense

of signal-to-noise performance, and often to the detriment of

channel separation and intermodulation distortion if the

succeeding circuitry is not carefully designed.

The PCM63 is a new solution to the problem. It combines all

the advantages of a conventional DAC (excellent full scale

performance, high signal-to-noise ratio and ease of use) with

superior low-level performance. Two DACs are combined

in a complementary arrangement to produce an extremely

linear output. The two DACs share a common reference and

a common R-2R ladder to ensure perfect tracking under all

conditions. By interleaving the individual bits of each DAC

and employing precise laser trimming of resistors, the highly

accurate match required between DACs is achieved.

This new, complementary linear or dual-DAC Colinear

approach, which steps away from zero with small steps in

both directions, avoids any glitching or “large” linearity

errors and provides an absolute current output. The low level

performance of the PCM63P is such that real 20-bit resolu-

tion can be realized, especially around the critical bipolar

zero point.

Table I shows the conversion made by the internal logic of

the PCM63P from binary two’s complement (BTC). Also,

the resulting internal codes to the upper and lower DACs

(see front page block diagram) are listed. Notice that only

the LSB portions of either internal DAC are changing

around bipolar zero. This accounts for the superlative per-

formance of the PCM63P in this area of operation.

DISCUSSION OF SPECIFICATIONS

DYNAMIC SPECIFICATIONS

Total Harmonic Distortion + Noise

The key specification for the PCM63P is total harmonic

distortion plus noise (THD+N). Digital data words are read

into the PCM63P at eight times the standard compact disk

audio sampling frequency of 44.1kHz (352.8kHz) so that a

sine wave output of 991Hz is realized. For production

testing, the output of the DAC goes to an I to V converter,

then to a programmable gain amplifier to provide gain at

lower signal output test levels, and then through a 40kHz

low pass filter before being fed into an analog type distortion

analyzer. Figure 1 shows a block diagram of the production

THD+N test setup.

For the audio bandwidth, THD+N of the PCM63P is essen-

tially flat for all frequencies. The typical performance curve,

“THD+N vs Frequency,” shows four different output signal

levels: 0dB, –20dB, –40dB, and –60dB. The test signals are

derived from a special compact test disk (the CBS CD-1). It

is interesting to note that the –20dB signal falls only about

10dB below the full scale signal instead of the expected

20dB. This is primarily due to the superior low-level signal

performance of the dual-DAC Colinear architecture of the

PCM63P.

In terms of signal measurement, THD+N is the ratio of

the PCM63P, THD+N is 100% tested at all three specified

output levels using the test setup shown in Figure 1. It is

significant to note that this test setup does not include any

output deglitching circuitry. All specifications are achieved

without the use of external deglitchers.

Dynamic Range

Dynamic range in audio converters is specified as the measure

of THD+N at an effective output signal level of –60dB

referred to 0dB. Resolution is commonly used as a theoretical

measure of dynamic range, but it does not take into account

the effects of distortion and noise at low signal levels. The

INPUT CODE

LOWER DAC CODE

UPPER DAC CODE

ANALOG OUTPUT

(20-bit Binary Two’s Complement)

(19-bit Straight Binary)

(19-bit Straight Binary)

+Full Scale

011...111

111...111 + 1LSB*

111...111

+Full Scale – 1LSB

011...110

111...111 + 1LSB*

111...110

Bipolar Zero + 2LSB

000...010

111...111 + 1LSB*

000...010

Bipolar Zero + 1LSB

000...001

111...111 + 1LSB*

000...001

Bipolar Zero

000...000

111...111 + 1LSB*

000...000

Bipolar Zero – 1LSB

111...111

111...111

000...000

Bipolar Zero – 2LSB

111...110

111...110

000...000

–Full Scale + 1LSB

100...001

000...001

000...000

–Full Scale

100...000

000...000

000...000

*The extra weight of 1LSB is added at this point to make the transfer function symmetrical around bipolar zero.

TABLE I. Binary Two’s Complement to Colinear Conversion Chart.