REV. A
AD1838A
–11–
Table I. Coding Scheme
Code Level
0111 . . . . 11111 +FS
0000 . . . . 00000 0 (Ref Level)
1000 . . . . 00000 –FS
AD1838A CLOCKING SCHEME
By default, the AD1838A requires an MCLK signal that is
256 times the required sample frequency up to a maximum of
12.288 MHz. The AD1838A uses a clock scaler to double the
clock frequency for use internally. The default setting of the
clock scaler is Multiply by 2. The clock scaler can also be set
Multiply by 1 (bypass) or by 2/3. The clock scaler is controlled
by programming the bits in the ADC Control 3 register. The
internal MCLK signal, IMCLK, should not exceed 24.576 MHz
to ensure correct operation.
The MCLK of the AD1838A should remain constant during
normal operation of the DAC and ADC. If it is required to change
the MCLK rate, then the AD1838A should be reset. Additionally,
if MCLK scaler needs to be modified so that the IMCLK does not
exceed 24.576 MHz, this should be done during the internal reset
phase of the AD1838A by programming the bits in the first
3072 MCLK periods following the reset.
Selecting DAC Sampling Rate
The AD1838A DAC engine has a programmable interpolator
that allows the user to select different interpolation rates based
on the required sample rate and MCLK value available. Table II
shows the settings required for sample rates based on a fixed
MCLK of 12.288 MHz.
Table II. DAC Sample Rate Settings
Sample Rate Interpolator Rate DAC Control 1 Register
48 kHz 8 000000xxxxxxxx00
96 kHz 4 000000xxxxxxxx01
192 kHz 2 000000xxxxxxxx10
Selecting an ADC Sample Rate
The AD1838A ADC engine has a programmable decimator that
allows the user to select the sample rate based on the MCLK
value. By default, the output sample rate is IMCLK/512. To
achieve a sample rate of IMCLK/256, the sample rate bit in the
ADC Control 1 register should be set as shown in Table III.
Table III. ADC Sample Rate Settings
Sample Rate ADC Control 1 Register
IMCLK/512 1100000xx0xxxxxx (48 kHz)
IMCLK/256 1100000xx1xxxxxx (96 kHz)
To maintain the highest performance possible, it is recommended
that the clock jitter of the master clock signal be limited to less than
300 ps rms, measured using the edge-to-edge technique. Even at
these levels, extra noise or tones may appear in the DAC outputs if
the jitter spectrum contains large spectral peaks. It is highly recom-
mended that the master clock be generated by an independent
crystal oscillator. In addition, it is especially important that the
clock signal should not be passed through an FPGA or other large
digital chip before being applied to the AD1838A. In most cases,
this will induce clock jitter because the clock signal is sharing
common power and ground connections with other unrelated
digital output signals.
FUNCTIONAL OVERVIEW
ADCs
There are two ADC channels in the AD1838A, configured as a
stereo pair. Each ADC has fully differential inputs. The ADC
section can operate at a sample rate of up to 96 kHz. The ADCs
include on-board digital decimation filters with 120 dB stop-band
attenuation and linear phase response, operating at an oversam-
pling ratio of 128 (for 48 kHz operation) or 64 (for 96 kHz
operation).
ADC peak level information for each ADC may be read from the
ADC Peak 0 and ADC Peak 1 registers. The data is supplied
as a 6-bit word with a maximum range of 0 dB to –63 dB and a
resolution of 1 dB. The registers will hold peak information
until read; after reading, the registers are reset so that new peak
information can be acquired. Refer to the register description for
details of the format. The two ADC channels have a common
serial bit clock and a left-right framing clock. The clock signals
are all synchronous with the sample rate.
The ADC digital pins, ABCLK and ALRCLK, can be set to
operate as inputs or outputs by connecting the M/S pin to
ODVDD or DGND, respectively. When the pins are set as
outputs, the AD1838A will generate the timing signals.
When the pins are set as inputs, the timing must be generated
by the external audio controller.
DACs
The AD1838A has six DAC channels arranged as three inde-
pendent stereo pairs, with six fully differential analog outputs
for improved noise and distortion performance. Each channel has
its own independently programmable attenuator, adjustable in
1024 linear steps. Digital inputs are supplied through three
serial data input pins (one for each stereo pair) and a common
frame (DLRCLK) and bit (DBCLK) clock. Alternatively, one of
the packed data modes may be used to access all six channels on a
single TDM data pin. A stereo replicate feature is included where
the DAC data sent to the first DAC pair is also sent to the
other DACs in the part. The AD1838A can accept DAC data at
a sample rate of 192 kHz on DAC 1 only. The stereo repli-
cate feature can then be used to copy the audio data to the
other DACs.
Each set of differential output pins sits at a dc level of V
REF
and
swings ± 1.4 V for a 0 dB digital input signal. A single op amp
third-order external low-pass filter is recommended to remove
high frequency noise present on the output pins, as well as to
provide differential-to-single-ended conversion. Note that the use
of op amps with low slew rate or low bandwidth may cause high
frequency noise and tones to fold down into the audio band;
care should be exercised in selecting these components.
The FILTD pin should be connected to an external grounded
capacitor. This pin is used to reduce the noise of the internal
DAC bias circuitry, thereby reducing the DAC output noise. In
some cases, this capacitor may be eliminated with little effect on
performance.
DAC and ADC Coding
The DAC and ADC output data stream is in a twos complement
encoded format. The word width can be selected from 16 bit,
20 bit, or 24 bit. The coding scheme is detailed in Table I.
