REV. 0
AD9752
–15–
AD9752
IOUTA
IOUTB
C
OPT
200V
U1
V
OUT
= I
OUTFS
3 R
FB
I
OUTFS
= 10mA
R
FB
200V
Figure 32. Unipolar Buffered Voltage Output
POWER AND GROUNDING CONSIDERATIONS, POWER
SUPPLY REJECTION
Many applications seek high speed and high performance under
less than ideal operating conditions. In these circuits, the imple-
mentation and construction of the printed circuit board design
is as important as the circuit design. Proper RF techniques must
be used for device selection, placement and routing as well as
power supply bypassing and grounding to ensure optimum
performance. Figures 42-47 illustrate the recommended printed
circuit board ground, power and signal plane layouts which are
implemented on the AD9752 evaluation board.
One factor that can measurably affect system performance is the
ability of the DAC output to reject dc variations or ac noise
superimposed on the analog or digital dc power distribution
(i.e., AVDD, DVDD). This is referred to as Power Supply
Rejection Ratio (PSRR). For dc variations of the power supply,
the resulting performance of the DAC directly corresponds to a
gain error associated with the DAC’s full-scale current, I
OUTFS
.
AC noise on the dc supplies is common in applications where
the power distribution is generated by a switching power supply.
Typically, switching power supply noise will occur over the
spectrum from tens of kHz to several MHz. PSRR vs. frequency
of the AD9752 AVDD supply, over this frequency range, is
given in Figure 33.
FREQUENCY – MHz
90
80
60
1.00.26
PSRR – dB
0.5 0.75
70
Figure 33. Power Supply Rejection Ratio of AD9752
Note that the units in Figure 33 are given in units of (amps out)/
(volts in). Noise on the analog power supply has the effect of
modulating the internal switches, and therefore the output
current. The voltage noise on the dc power, therefore, will be
added in a nonlinear manner to the desired I
OUT
. Due to the
relative different sizes of these switches, PSRR is very code depen-
dent. This can produce a mixing effect which can modulate low
frequency power supply noise to higher frequencies. Worst case
PSRR for either one of the differential DAC outputs will occur
when the full-scale current is directed towards that output. As a
result, the PSRR measurement in Figure 33 represents a worst
case condition in which the digital inputs remain static and the
full scale output current of 20 mA is directed to the DAC out-
put being measured.
An example serves to illustrate the effect of supply noise on the
analog supply. Suppose a switching regulator with a switching
frequency of 250 kHz produces 10 mV rms of noise and for
simplicity sake (i.e., ignore harmonics), all of this noise is con-
centrated at 250 kHz. To calculate how much of this undesired
noise will appear as current noise super imposed on the DAC’s
full-scale current, I
OUTFS
, one must determine the PSRR in dB
using Figure 33 at 250 kHz. To calculate the PSRR for a given
R
LOAD
, such that the units of PSRR are converted from A/V to
V/V, adjust the curve in Figure 33 by the scaling factor 20 × Log
(R
LOAD
). For instance, if R
LOAD
is 50 , the PSRR is reduced
by 34 dB (i.e., PSRR of the DAC at 1 MHz which is 74 dB in
Figure 33 becomes 40 dB V
OUT
/V
IN
).
Proper grounding and decoupling should be a primary objective
in any high speed, high resolution system. The AD9752 features
separate analog and digital supply and ground pins to optimize
the management of analog and digital ground currents in a
system. In general, AVDD, the analog supply, should be de-
coupled to ACOM, the analog common, as close to the chip as
physically possible. Similarly, DVDD, the digital supply, should
be decoupled to DCOM as close as physically as possible.
For those applications that require a single +5 V or +3 V supply
for both the analog and digital supply, a clean analog supply
may be generated using the circuit shown in Figure 34. The
circuit consists of a differential LC filter with separate power
supply and return lines. Lower noise can be attained using low
ESR type electrolytic and tantalum capacitors.
100mF
ELECT.
10-22mF
TANT.
0.1mF
CER.
TTL/CMOS
LOGIC
CIRCUITS
+5V OR +3V
POWER SUPPLY
FERRITE
BEADS
AVDD
ACOM
Figure 34. Differential LC Filter for Single +5 V or +3 V
Applications
Maintaining low noise on power supplies and ground is critical
to obtaining optimum results from the AD9752. If properly
implemented, ground planes can perform a host of functions on
high speed circuit boards: bypassing, shielding, current trans-
port, etc. In mixed signal design, the analog and digital portions
of the board should be distinct from each other, with the analog
ground plane confined to the areas covering the analog signal
traces, and the digital ground plane confined to areas covering
the digital interconnects.
All analog ground pins of the DAC, reference and other analog
components should be tied directly to the analog ground plane.
The two ground planes should be connected by a path 1/8
to 1/4 inch wide underneath or within 1/2 inch of the DAC to
REV. 0
AD9752
–16–
maintain optimum performance. Care should be taken to ensure
that the ground plane is uninterrupted over crucial signal paths.
On the digital side, this includes the digital input lines running
to the DAC as well as any clock signals. On the analog side, this
includes the DAC output signal, reference signal and the supply
feeders.
The use of wide runs or planes in the routing of power lines is
also recommended. This serves the dual role of providing a low
series impedance power supply to the part, as well as providing
some “free” capacitive decoupling to the appropriate ground
plane. It is essential that care be taken in the layout of signal and
power ground interconnects to avoid inducing extraneous volt-
age drops in the signal ground paths. It is recommended that all
connections be short, direct and as physically close to the pack-
age as possible in order to minimize the sharing of conduction
paths between different currents. When runs exceed an inch in
length, strip line techniques with proper termination resistor
should be considered. The necessity and value of this resistor
will be dependent upon the logic family used.
For a more detailed discussion of the implementation and
construction of high speed, mixed signal printed circuit boards,
refer to Analog Devices’ application notes AN-280 and
AN-333.
FREQUENCY – Hz
–30
–40
–100
1M
600k
AMPLITUDE – dBm
800k
–50
–60
–70
–80
–90
Figure 35a. Notch in Missing Bin at 750 kHz is Down
>60 dB. (Peak Amplitude + 0 dBm).
FREQUENCY – MHz
–30
–40
–100
5.15
4.85
AMPLITUDE – dBm
5
–50
–60
–70
–80
–90
–110
Figure 35b. Notch in Missing Bin at 5 MHz is Down
>60 dB. (Peak Amplitude + 0 dBm).
APPLICATIONS
VDSL Applications Using the AD9752
Very High Frequency Digital Subscriber Line (VDSL) technol-
ogy is growing rapidly in applications requiring data transfer
over relatively short distances. By using QAM modulation and
transmitting the data in multiple discrete tones, high data rates
can be achieved.
As with other multitone applications, each VDSL tone is ca-
pable of transmitting a given number of bits, depending on the
signal-to-noise ratio (SNR) in a narrow band around that tone.
The tones are evenly spaced over the range of several kHz to
10 MHz. At the high frequency end of this range, performance
is generally limited by cable characteristics and environmental
factors, such as external interferers. Performance at the lower
frequencies is much more dependent on the performance of the
components in the signal chain. In addition to in-band noise,
intermodulation from other tones can also potentially interfere
with the recovery of data for a given tone. The two graphs in
Figure 35 represent a 500 tone missing bin test vector, with
frequencies evenly spaced from 400 Hz to 10 MHz. This test is
very commonly done to determine if distortion will limit the
number of bits which can be transmitted in a tone. The test
vector has a series of missing tones around 750 kHz, which is
represented in Figure 35a and a series of missing tones around
5 MHz which is represented in Figure 35b. In both cases, the
spurious free range between the transmitted tones and the empty
bins is greater than 60 dB.
Using the AD9752 for Quadrature Amplitude Modulation
(QAM)
QAM is one of the most widely used digital modulation
schemes in digital communication systems. This modulation
technique can be found in FDM as well as spreadspectrum (i.e.,
CDMA) based systems. A QAM signal is a carrier frequency
that is modulated in both amplitude (i.e., AM modulation) and
phase (i.e., PM modulation). It can be generated by indepen-
dently modulating two carriers of identical frequency but with a
90° phase difference. This results in an in-phase (I) carrier com-
ponent and a quadrature (Q) carrier component at a 90° phase
shift with respect to the I component. The I and Q components
are then summed to provide a QAM signal at the specified car-
rier frequency.
A common and traditional implementation of a QAM modu-
lator is shown in Figure 36. The modulation is performed in the
analog domain in which two DACs are used to generate the
baseband I and Q components, respectively. Each component is
then typically applied to a Nyquist filter before being applied to
a quadrature mixer. The matching Nyquist filters shape and
limit each component’s spectral envelope while minimizing
intersymbol interference. The DAC is typically updated at the
QAM symbol rate or possibly a multiple of it if an interpolating
filter precedes the DAC. The use of an interpolating filter typi-
cally eases the implementation and complexity of the analog
filter, which can be a significant contributor to mismatches in
gain and phase between the two baseband channels. A quadra-
ture mixer modulates the I and Q components with in-phase
and quadrature phase carrier frequency and then sums the two
outputs to provide the QAM signal.
REV. 0
AD9752
–17–
AD9752
(“I DAC”)
AD9752
(“Q DAC”)
IOUTA
IOUTB
QOUTA
QOUTB
DCOM
FSADJ
REFIO
SLEEP
R
SET2
1.9kV
0.1mF
CLK
Q DATA
INPUT
I DATA
INPUT
DVDD
AVDD
100W
500V
100V
C
FILTER
100V
C
FILTER
100V
500V
500V
500V500V
500V
500V
634V
0.1mF
+5V
VPBF
BBIP
BBIN
BBQP
BBQN
AD8346
PHASE
SPLITTER
LOIP
LOIN
VOUT
500mV p-p WITH
V
CM
=1.2V
NOTE: 500V RESISTOR NETWORK - OHMTEK ORN5000D
100V RESISTOR NETWORK - TOMC1603-100D
REFLO
ACOM
REFLO
AVDD
REFIO
FSADJ
R
SET1
2kV
R
CAL
220V
U1
U2
AVDD
1.82V
LATCHES
500V
DAC
DAC
+
LATCHES
Figure 37. Baseband QAM Implementation Using Two AD9752s
AD9752
0
90
AD9752
CARRIER
FREQUENCY
12
12
TO
MIXER
DSP
OR
ASIC
NYQUIST
FILTERS
QUADRATURE
MODULATOR
S
Figure 36. Typical Analog QAM Architecture
In this implementation, it is much more difficult to maintain
proper gain and phase matching between the I and Q channels.
The circuit implementation shown in Figure 37 helps improve
upon the matching and temperature stability characteristics
between the I and Q channels, as well as showing a path for up-
conversion using the AD8346 quadrature modulator. Using a
single voltage reference derived from U1 to set the gain for both
the I and Q channels will improve the gain matching and stabil-
ity. R
CAL
can be used to compensate for any mismatch in gain
between the two channels. This mismatch may be attributed to
the mismatch between R
SET1
and R
SET2
, effective load resistance
of each channel, and/or the voltage offset of the control ampli-
fier in each DAC. The differential voltage outputs of U1 and U2
are fed into the respective differential inputs of the AD8346 via
matching networks.
Using the same matching techniques described above, Figure 38
shows an example of the AD9752 used in a W-CDMA transmit-
ter application using the AD6122 CDMA 3 V transmitter IF
subsystem. The AD6122 has functions, such as external gain
control and low distortion characteristics, needed for the supe-
rior Adjacent Channel Power (ACP) requirements of W-CDMA.
CDMA
Carrier Division Multiple Access, or CDMA, is an air transmit/
receive scheme where the signal in the transmit path is modu-
lated with a pseudorandom digital code (sometimes referred to
as the spreading code). The effect of this is to spread the trans-
mitted signal across a wide spectrum. Similar to a DMT wave-
form, a CDMA waveform containing multiple subscribers can
be characterized as having a high peak to average ratio (i.e.,
crest factor), thus demanding highly linear components in the
transmit signal path. The bandwidth of the spectrum is defined
by the CDMA standard being used, and in operation is imple-
mented by using a spreading code with particular characteristics.
Distortion in the transmit path can lead to power being trans-
mitted out of the defined band. The ratio of power transmitted
in-band to out-of-band is often referred to as Adjacent Channel
Power (ACP). This is a regulatory issue due to the possibility of
interference with other signals being transmitted by air. Regula-
tory bodies define a spectral mask outside of the transmit band,
and the ACP must fall under this mask. If distortion in the
transmit path cause the ACP to be above the spectral mask,
then filtering, or different component selection is needed to
meet the mask requirements.

AD9752ARZ

Mfr. #:
Manufacturer:
Analog Devices Inc.
Description:
Digital to Analog Converters - DAC 12-Bit 100 MSPS
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union