LC2MOS High Speed, mP-Compatible
a
FEATURES
Fast Conversion Time: 660 ns max
100 kHz Track-and-Hold Function
1 MHz Sample Rate
Unipolar and Bipolar Input Ranges
Ratiometric Reference Inputs
No External Clock
Extended Temperature Range Operation
Skinny 20-Pin DlPs, SOIC and 20-Terminal
Surface Mount Packages
GENERAL DESCRIPTION
The AD7821 is a high speed, 8-bit, sampling, analog-to-digital
converter that offers improved performance over the popular
AD7820. It offers a conversion time of 660 ns (vs. 1.36 µs for
the AD7820) and 100 kHz signal bandwidth (vs. 6.4 kHz). The
sampling instant is better defined and occurs on the falling edge
of
WR or RD. The provision of a VSS pin (Pin 19) allows the
part to operate from ±5 V supplies and to digitize bipolar input
signals. Alternatively, for unipolar inputs, the V
grounded and the AD7821 will operate from a single +5 V supply, like the AD7820.
The AD7821 has a built-in track-and-hold function capable of
digitizing full-scale signals up to 100 kHz max. It also uses a
half-flash conversion technique that eliminates the need to generate a CLK signal for the ADC.
The AD7821 is designed with standard microprocessor control
signals (
outputs capable of interfacing to high speed data buses. An
overflow output (
achieve higher resolution.
The AD7821 is fabricated in Linear-Compatible CMOS
(LC
precision bipolar circuits with low power CMOS logic. The part
features a low power dissipation of 50 mW.
CS, RD, WR, RDY, INT) and latched, three-state data
OFL) is also provided for cascading devices to
2
MOS), an advanced, mixed technology process combining
pin can be
SS
8-Bit ADC with Track/Hold Function
AD7821
FUNCTIONAL BLOCK DIAGRAM
PRODUCT HIGHLIGHTS
1. Fast Conversion Time
The half-flash conversion technique, coupled with fabrication
on Analog Devices’ LC
version time. The conversion time for the WR-RD mode is
660 ns, with 700 ns for the RD mode.
2. Built-In Track-and-Hold
This allows input signals with slew rates up to 1.6 V/µs to be
converted to 8-bits without an external track-and-hold. This
corresponds to a 5 V peak-to-peak, 100 kHz sine wave signal.
3. Total Unadjusted Error
The AD7821 features an excellent total unadjusted error figure of less than ±1 LSB over the full operating temperature
range.
4. Unipolar/Bipolar Input Ranges
The AD7821 is specified for single supply (+5 V) operation
with a unipolar full-scale range of 0 to +5 V, and for dual supply (±5 V) operation with a bipolar input range of ±2.5 V.
Typical performance characteristics are given for other input
ranges.
5. Dynamic Specifications for DSP Users
In addition to the traditional ADC specifications, the
AD7821 is specified for ac parameters, including signal-tonoise ratio, distortion and slew rate.
2
MOS process, enables a very fast con-
REV. A
Information furnished by Analog Devices is believed to be accurate and
reliable. However, no responsibility is assumed by Analog Devices for its
use, nor for any infringements of patents or other rights of third parties
which may result from its use. No license is granted by implication or
otherwise under any patent or patent rights of Analog Devices.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 617/329-4700 Fax: 617/326-8703
VDD = +5 V 6 5%, GND = 0 V. Unipolar Input Range: VSS = GND, V
V
AD7821–SPECIFICA TIONS
V
(–) = –2.5 V. These test conditions apply unless otherwise stated. All specifications T
REF
(–) = GND. Bipolar Input Range: VSS = –5 V 6 5%, V
REF
MIN
to T
unless otherwise noted. Specifications
MAX
(+) = 2.5 V,
REF
apply for RD Mode (Pin 7 = 0 V).
REF
REF
(–)/V
REF
(–)/V
1
B, T Versions Units Comments
V
DD
(+) VSS/V
REF
(+) V
REF
REF
(–)/V
(+) V min/V max
REF
(–)/V
V min/V max
DD
(+) V min/ max
REF
fa (84.72 kHz) and fb (94.97 kHz) Full-Scale Sine Waves
with f
SOURCE
SINK
SINK
(V
REF
= 500 kHz
SAMPLING
= 360 µA
= 1.6 mA
= 2.6 mA
(+) = 4.75 V max for Unipolar Mode)
Parameter K Version
UNIPOLAR INPUT RANGE
Resolution
Total Unadjusted Error
2
3
8 8 Bits
±1 ±1 LSB max
Minimum Resolution for which
No Missing Codes are Guaranteed 8 8 Bits
BIPOLAR INPUT RANGE
Resolution
2
8 8 Bits
Zero Code Error ±1 ±1 LSB max
Full Scale Error ±1 ±1 LSB max
Signal-to-Noise Ratio (SNR)
3
45 45 dB min VIN = 99.85 kHz Full-Scale Sine Wave with f
Total Harmonic Distortion (THD)3–50 –50 dB max VIN = 99.85 kHz Full-Scale Sine Wave with f
Peak Harmonic or Spurious Noise3–50 –50 dB max VIN = 99.85 kHz Full-Scale Sine Wave with f
Intermodulation Distortion (IMD)
3
–50 –50 dB max Second Order Terms
Slew Rate, Tracking
3
–50 –50 dB max Third Order Terms
1.6 1.6 V/µs max
2.36 2.36 V/µs typ
REFERENCE INPUT
Input Resistance 1.0/4.0 1.0/4.0 kΩ min/kΩ max
V
(+) Input Voltage Range V
REF
V
(–) Input Voltage Range VSS/V
REF
ANALOG INPUT
Input Voltage Range V
Input Leakage Current ±3 ±3 µA max –5 V ≤ VIN ≤ +5 V
Input Capacitance 55 55 pF typ
LOGIC INPUTS
CS, WR, RD
V
INH
V
INL
I
(CS, RD)11µA max
INH
I
(WR)33µA max
INH
I
INL
Input Capacitance
4
2.4 2.4 V min
0.8 0.8 V max
–1 –1 µA max
8 8 pF max Typically 5 pF
MODE
V
INH
V
INL
I
INH
I
INL
Input Capacitance
4
3.5 3.5 V min
1.5 1.5 V max
200 200 µA max 50 µA typ
–1 –1 µA max
8 8 pF max Typically 5 pF
LOGIC OUTPUTS
DB0–DB7, OFL, INT
V
OH
V
OL
I
(DB0–DB7) ±3 ±3 µA max Floating State Leakage
OUT
Output Capacitance4 (DB0–DB7) 8 8 pF max Typically 5 pF
4.0 4.0 V min I
0.4 0.4 V max I
RDY
V
OL
I
OUT
Output Capacitance
POWER SUPPLY
5
I
DD
I
SS
Power Dissipation 50 50 mW typ
4
0.4 0.4 V max I
±3 ±3 µA max Floating State Leakage
8 8 pF max Typically 5 pF
15 20 mA max CS = RD = 0 V
100 100 µA max CS = RD = 0 V
Power Supply Sensitivity ±1/4 ±1/4 LSB max ±1/16 LSB typ, VDD = 4.75 V to 5.25 V,
NOTES
1
Temperature Ranges are as follows: K Version = –40°C to +85°C; B Version = –40°C to +85°C; T Version = –55°C to +125°C.
2
1 LSB = 19.53 mV for both the unipolar (0 V to +5 V) and bipolar (–2.5 V to +2.5 V) input ranges.
3
See Terminology.
4
Sample tested at +25°C to ensure compliance.
5
See Typical Performance Characteristics.
Specifications subject to change without notice.
(+) = 5 V,
REF
SAMPLING
SAMPLING
SAMPLING
= 500 kHz
= 500 kHz
= 500 kHz
–2–
REV. A
AD7821
TIMING CHARACTERISTICS
1
(VDD = +5 V ± 5%, VSS = 0 V or –5 V ± 5%; Unipolar or Bipolar Input Range)
Limit at Limit at
Parameter (All Versions) (K, B Versions) (T Version) Units Conditions/Comments
Limit at +258CT
t
CSS
t
CSH
t
RDY
t
CRD
t
ACC0
t
INTH
t
DH
t
P
t
WR
t
RD
t
READ1
t
ACC1
2
3
2
4
3
0 0 0 ns min CS to RD/WR Setup Time
0 0 0 ns min CS to RD/WR Hold Time
70 85 100 ns max CS to RDY Delay. Pull-Up
700 875 975 ns max Conversion Time (RD Mode)
t
+ 25 t
CRD
t
+ 50 t
CRD
50 – – ns typ RD to INT Delay (RD Mode)
80 85 90 ns max
15 15 15 ns min Data Hold Time
60 70 80 ns max
350 425 500 ns min Delay Time Between Conversions
250 325 400 ns min Write Pulse Width
10 10 10 µs max
250 350 450 ns min Delay Time between WR and RD Pulses
160 205 240 ns min RD Pulse Width (WR-RD Mode, see Figure 12b)
160 205 240 ns max CL = 20 pF
, T
MIN
MAX
+ 30 t
CRD
+ 65 t
CRD
T
, T
MIN
MAX
Resistor 5 k1.
Data Access Time (RD Mode)
+ 35 ns max CL = 20 pF
CRD
+ 75 ns max CL = 100 pF
CRD
Determined by t
Data Access Time (WR-RD Mode, see Figure 12b)
ACC1
185 235 275 ns max CL = 100 pF
t
RI
t
INTL
t
READ2
t
ACC2
t
IHWR
t
ID
2
3
2
3
150 185 220 ns max RD to INT Delay
380 – – ns typ WR to INT Delay
500 610 700 ns max
65 75 85 ns min RD Pulse Width (WR-RD Mode, see Figure 12a)
Determined by t
Data Access Time (WR-RD Mode, see Figure 12a)
ACC2
65 75 85 ns max CL = 20 pF
90 110 130 ns max CL = 100 pF
80 100 120 ns max WR to INT Delay (Stand-Alone Operation)
Data Access Time after INT
(Stand-Alone Operation)
30 35 40 ns max CL = 20 pF
45 60 70 ns max CL = 100 pF
NOTES
1
Sample tested at +25°C to ensure compliance. All input control signals are specified with tr = tf = 5 ns (10% to 90% of +5 V) and timed from a voltage level of 1.6 V.
2
CL = 50 pF.
3
Measured with load circuits of Figure 1 and defined as the time required for an output to cross 0.8 V or 2.4 V.
4
Defined as the time required for the data lines to change 0.5 V when loaded with the circuits of Figure 2.
Specifications subject to change without notice.
Test Circuits
a. High Z to V
OH
Figure 1. Load Circuits for Data Access Time Test
a. VOH to High Z b. VOL to High Z
Figure 2. Load Circuits for Data Hold Time Test
REV. A
b. High Z to V
OL
ORDERING GUIDE
Total
Temperature Unadjusted Package
Range Error (LSB) Option
Model
1
AD7821KN –40°C to +85°C ±1 max N-20
AD7821KP –40°C to +85°C ±1 max P-20A
AD7821KR –40°C to +85°C ±1 max R-20
AD7821BQ –40°C to +85°C ±1 max Q-20
AD7821TQ –55°C to +125°C ±1 max Q-20
AD7821TE –55°C to +125°C ±1 max E-20A
NOTES
1
To order MIL-STD-883, Class B processed parts, add /883B to part
number. Contact local sales office for military data sheet.
2
E = Leadless Ceramic Chip Carrier; N = Plastic DIP; P = Plastic Leaded
Chip Carrier; Q = Cerdip; R = SOIC.
–3–
2
AD7821
WARNING!
ESD SENSITIVE DEVICE
ABSOLUTE MAXIMUM RATINGS*
VDD to GND . . . . . . . . . . . . . . . . . . . . . . . . . . . –0.3 V, + 7 V
V
to GND . . . . . . . . . . . . . . . . . . . . . . . . . . . . +0.3 V, + 7 V
SS
Digital Input Voltage to GND
(Pins 6–8, 13) . . . . . . . . . . . . . . . . . . . –0.3 V, V
+ 0.3 V
DD
Digital Output Voltage to GND
(Pins 2–5, 9, 14–18) . . . . . . . . . . . . . . . –0.3 V, V
V
(+) to GND . . . . . . . . . . . . . . . VSS – 0.3 V, VDD + 0.3 V
REF
V
(–) to GND . . . . . . . . . . . . . . . VSS – 0.3 V, VDD + 0.3 V
REF
V
to GND . . . . . . . . . . . . . . . . . . . VSS – 0.3 V, VDD + 0.3 V
IN
+ 0.3 V
DD
Operating Temperature Range
Industrial (B Version) . . . . . . . . . . . . . . . . –40°C to +85°C
Extended (T Version) . . . . . . . . . . . . . . . –55°C to +125°C
Storage Temperature Range . . . . . . . . . . . . –65°C to +150°C
Lead Temperature (Soldering, 10 secs) . . . . . . . . . . . +300°C
Power Dissipation (Any Package) to +75°C . . . . . . . . 450 mW
Derates above +75°C by . . . . . . . . . . . . . . . . . . . . . 6 mW/°C
*Stresses above those listed under “Absolute Maximum Ratings” may cause
permanent damage to the device. This is a stress rating only and functional
operation of the device at these or any other conditions above those indicated in the
operational sections of this specification is not implied. Exposure to absolute
maximum rating conditions for extended periods may affect device reliability.
Commercial (K Version) . . . . . . . . . . . . . . –40°C to +85°C
CAUTION
ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily
accumulate on the human body and test equipment and can discharge without detection.
Although the AD7821 features proprietary ESD protection circuitry, permanent damage may
occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD
precautions are recommended to avoid performance degradation or loss of functionality.
PIN CONFIGURATIONS
DIP AND SOIC LCCC PLCC
TERMINOLOGY
LEAST SIGNIFICANT BIT (LSB)
An ADC with 8-bit resolution can resolve one part in 28 (1/256
of full scale). For the AD7821 operating in either the unipolar
or bipolar input range with 5 V full scale, one LSB is 19.53 mV.
TOTAL UNADJUSTED ERROR
This is a comprehensive specification which includes relative
accuracy, offset error and full-scale error.
SLEW RATE
Slew Rate is the maximum allowable rate of change of input
signal such that the digital sample values are not in error.
TOTAL HARMONIC DISTORTION
Total harmonic distortion is the ratio of the square root of the
sum of the squares of the rms value of the harmonics to the rms
value of the fundamental. For the AD7821, total harmonic distortion (THD) is defined as
V
20 log
()
2
2
+V
2
2
+V
3
5
V
1
2
+V
6
dB
where V1 is the rms amplitude of the fundamental and V2, V3,
V
, V5, V6, are the rms amplitudes of the individual harmonics.
4
INTERMODULATION DISTORTION
With inputs consisting of sine waves at two frequencies, fa and
fb, any active device with nonlinearities will create distortion
products, of order (m+n), at sum and difference frequencies of
mfa+nfb, where m, n = 0, 1, 2, 3,- - - -. Intermodulation terms
are those for which m or n is not equal to zero. For example, the
second order terms include (fa + fb) and (fa – fb), and the third
order terms include (2fa + fb), (2fa – fb), (fa + 2fb) and
(fa – 2fb). For the AD7821 intermodulation distortion is calculated separately for both the second and third order terms.
SIGNAL-TO-NOISE RATIO
Signal-to-noise ratio (SNR) is measured signal-to-noise at the
output of the ADC. The signal is the rms magnitude of the fundamental. Noise is the rms sum of all nonfundamental signals
(excluding dc) up to half the sampling frequency. SNR is dependent on the number of quantization levels used in the digitization process. The theoretical SNR for a sine wave input is
given by:
SNR = (6.02 N + 1.76) dB
where N is the number of bits in the ADC. Thus, for an ideal
8-bit ADC, SNR = 50 dB.
PEAK HARMONIC OR SPURIOUS NOISE
Peak harmonic or spurious noise is the rms value of the largest
nonfundamental frequency (excluding dc) up to half the sampling frequency to the rms value of the fundamental.
–4–
REV. A
Typical Performance Curves–
AD7821
Conversion Time (RD Mode) vs.
Temperature
Accuracy vs. t
RD
Power Supply Current vs. Temperature (Not Including Reference Ladder)
Accuracy vs. t
P
Accuracy vs. t
Accuracy vs. V
[V
= V
REF
(+) – V
REF
REF
WR
REF
(–)]
Effective Number of Bits vs. Input
±
Signal (
2.5 V) Frequency
REV. A
t
, Internal Time Delay vs.
INTL
Temperature
–5–
Output Current vs. Temperature
AD7821
PIN FUNCTION DESCRIPTION
Pin Mnemonic Description
1VINAnalog Input: Range V
V
(+)
REF
(–) ≤ VIN ≤
REF
2 DB0 Three-State Data Output (LSB).
3–5 DB1–DB3 Three-State Data Outputs.
6
WR/RDY WRITE control input/READY status
output. See Digital Interface section.
7 MODE Mode Selection Input. It determines
whether the device operates in the
WR-RD or RD mode. This input is internally pulled low through a 50 µA
current source. See Digital Interface
section.
8
RD READ Input. RD must be low to access
data from the part. See Digital Interface
section.
9
INT INTERRUPT Output. INT going low
indicates that the conversion is complete.
INT returns high on the rising edge of
CS or RD. See Digital Interface section.
10 GND Ground.
11 V
12 V
13
(–) Lower limit of reference span.
REF
(+) Upper limit of reference span.
REF
Range: V
Range: V
SS
REF
≤ V
REF
(–) < V
(–) ≤ V
(+) ≤ VDD.
REF
REF
(+).
CS Chip Select Input. The device is selected
when this input is low.
14–16 DB4–DB6 Three-State Data Outputs.
17 DB7 Three-State Data Output (MSB).
18
OFL Overflow Output. If the analog input is
higher than (V
(+) – 1/2 LSB), OFL
REF
will be low at the end of conversion. It is
a non-three-state output which can be
used to cascade 2 or more devices to
increase resolution.
19 V
20 V
SS
DD
Negative supply voltage.
V
= 0 V; Unipolar Operation.
SS
V
= –5 V; Bipolar Operation.
SS
Positive supply voltage, +5 V.
CIRCUIT INFORMATION
BASIC DESCRIPTION
The AD7821 uses a half flash conversion technique (see Functional Block Diagram), whereby two 4-bit flash ADCs are used
to achieve an 8-bit result. Each 4-bit flash ADC contains 15
comparators, which compare an unknown input voltage to the
reference ladder, to achieve a 4-bit result. The MS (most significant) flash ADC converts an unknown analog input voltage
(V
) to provide the 4 MS data bits. An internal DAC, driven by
IN
the 4 MS data bits, then recreates an analog approximation of
the input voltage. The DAC output voltage is subtracted from
the analog input, and the difference is converted by the LS
(least significant) ADC to provide the 4 LS data bits. The MS
flash ADC also has one additional comparator to detect overrange on the analog input.
OPERATING SEQUENCE
The AD7821 has two operating modes. The RD mode allows a
conversion to be started and data to be read with a single, extended, READ operation (i.e.,
CS and RD are taken low). The
conversion process is timed out by internal one-shots. The WRRD mode uses
WR to start a conversion and RD to read the
data and allows the conversion timing to be externally controlled. The operating sequence for the WR-RD mode is shown
in Figure 3.
Figure 3. Operating Sequence (WR-RD Mode)
A conversion is initiated and the analog input signal (VIN)
sampled on the falling edge of
mode). A setup time (t
WR (falling edge of RD, RD
, delay time between conversions) of
P
350 ns is required prior to this falling edge. See Digital Interface
section for more details. When
WR is low, the internal MS
(most significant) ADC compares the sampled analog input with
the reference ladder to provide the 4 MS data bits. A minimum
of 250 ns is required for this comparison. On the rising edge of
WR, the MS data result is latched internally and the LS (least
significant) conversion begins, to yield the 4 LS data bits.
goes low typically 380 ns after the rising edge of
WR. This indi-
INT
cates the LS conversion is complete and that both the LS and
MS data results are latched into the output buffer.
RD going
low then enables the output data. If a faster conversion time is
required, the
RD line can be brought low 250 ns after WR goes
high. This latches both the LS and MS data bits and outputs the
conversion result on DB0–DB7.
REFERENCE AND INPUT
The V
(–) and V
REF
(+) reference inputs on the AD7821 are
REF
fully differential and define the zero and full-scale input range of
the ADC. The transfer characteristic of the part is defined by
the integer value of the following expression:
Data (LSBs) = 256
V
V
IN−VREF
(+)−V
REF
REF
(−)
(−)
+0.5
As a result, the analog input (VIN) Of the device can easily be set
up to provide both unipolar and bipolar operation. The data
output code for unipolar and bipolar operation is Natural Binary
and Offset Binary, respectively.
The span of the analog input voltage can easily be varied.
By reducing the reference span, V
REF
(+) – V
(–), to less than
REF
5 V the sensitivity of the converter can be increased (i.e., if
V
= 2 V then 1 LSB = 7.8 mV). The reference flexibility also
REF
allows the input span for unipolar operation to be offset from
zero (V
(–) > GND). Additionally, the input/reference ar-
REF
rangement facilitates ratiometric operation.
Figures 4 and 5 show some configurations which are possible.
For minimum noise a 47 µF capacitor in parallel with a 0.1 µF
capacitor should be connected between the reference inputs and
GND.
–6–
REV. A
Figure 4. Power Supply as Reference.
Unipolar Operation (0 to + 5 V)
AD7821
The input capacitors must charge to the input voltage through
the on resistance of the analog switches (about 2 kΩ to 5 kΩ).
In addition, about 12 pF of input stray capacitance must be
charged.
The analog input can be modeled as an equivalent RC network
as shown in Figure 7. As R
input capacitance takes longer to charge.
The comparators track the analog input between conversions. A
minimum delay time (t
sions to allow for voltage source settling and comparator tracking time. This allows input time constants of 50 ns without
settling time problems. Typical total input capacitance values of
55 pF allow R
to be 0.9 kΩ without lengthening tP to give V
S
more time to settle.
Figure 7. RC Network Model
(source impedance) increases, the
S
) of 350 ns is required between conver-
P
IN
Figure 5. External Reference.
Bipolar Operation (–2.5 V to +2.5 V)
INPUT CURRENT
The analog input of the AD7821 behaves somewhat differently
to conventional A/D converters. This is due to the ADC’s
sampled data comparators, which take varying amounts of input
current depending on the cycle of the converter.
The equivalent input circuit of the AD7821 is shown in Figure
6. When a conversion ends (e.g., falling edge of
mode, t
RD
> t
) all the input switches are closed and VIN is
INTL
INT, WR-RD
connected to the comparators of the internal LS and MS ADCs.
Therefore, V
is connected to 31 one-pF input capacitors
IN
simultaneously .
INPUT TRANSIENTS
Transients on the analog input signal caused by charging
current flowing into V
will not normally degrade the ADC’s
IN
performance. In effect, the AD7821 does not “look” at the input when these transients occur. The comparators’ inputs track
V
and are not sampled until the falling edge of WR (WR-RD
IN
Mode) or
RD (RD Mode), so at least 350 ns (tP) is provided to
charge the ADC’s input capacitance. It is, therefore, not necessary to filter out these transients with an external capacitor at
the V
terminal.
IN
INHERENT TRACK-AND-HOLD
A major benefit of the AD7821’s input structure is its ability to
measure a variety of high-speed signals without the help of an
external track-and-hold. Any ADC which does not have a builtin track-and-hold, regardless of its speed, requires the analog input to remain stable to at least 1/2 LSB for the duration of the
conversion to maintain full accuracy. This requires the use of a
track-and-hold whenever the input is a high-speed signal. The
AD7821’s sampled-data comparators, by nature of their input
switching, inherently accomplish this track-and-hold function.
Although the conversion time for the AD7821 is 660 ns
(WR-RD mode, t
be stable to 1/2 LSB is much smaller. The AD7821 tracks V
+ tRD + t
WR
), the time for which VIN must
ACC1
IN
between conversions only, and its value on the falling edge of
WR or RD in the WR-RD or RD modes, respectively, is the
measured value.
REV. A
Figure 6. AD7821 Equivalent Input Circuit
SINUSOIDAL INPUTS
The bandwidth of the built-in track-and-hold is 100 kHz max
(150 kHz typ, 5 V p-p). This is limited by the analog bandwidth
of the comparators and timing skew between the comparator
switches. This means that the analog input frequency can be up
to 100 kHz without the aid of an external track-and-hold. The
Nyquist criterion requires that the sampling rate be at least
twice the input frequency (i.e., ≥2 3 100 kHz). This requires an
ideal antialiasing filter with an infinite roll-off. To ease the prob-
–7–
AD7821
lem of antialiasing filter design, the sampling rate is usually set
much greater than the Nyquist criterion. The maximum sampling rate (f
(t
< t
RD
INTL
f
MAX
f
MAX
tWR = Write Pulse Width
t
RD
t
RI
= Delay Time between Conversions
t
P
) for the AD7821 in the WR-RD mode,
MAX
) can be calculated as follows:
=
t
WR+tRD+tRI+tP
=
0.25E − 6 +0.25E − 6 + 0.15E − 6 +0.35E −6
= Delay Time between WR and RD Pulses
= RD to INT Delay
1
1
This permits a maximum sampling rate for the AD7821 of
1 MHz, which is much greater than the Nyquist criterion for
sampling a 100 kHz analog input signal.
DIGITAL SIGNAL PROCESSING APPLICATIONS
In Digital Signal Processing (DSP) application areas like voice
recognition, echo cancellation and adaptive filtering, the dynamic characteristics (Signal-to-Noise Ratio, Harmonic Distortion, Intermodulation Distortion) of an ADC are critical. Since
the AD7821 is a very fast ADC with a built-in track-and-hold
function, it is specified dynamically as well as with standard dc
specifications (Total Unadjusted Error, etc.).
SIGNAL-TO-NOISE RATIO AND DISTORTION
The dynamic performance of the AD7821 is evaluated by applying a very low distortion sine wave signal to the analog input
(V
) which is then sampled at a 512 kHz sampling rate. A Fast
IN
Fourier Transform (FFT) plot is then generated from which
Signal-to-Noise Ratio (SNR) and harmonic distortion data are
obtained.
Figure 8 shows a 2048 point FFT plot of the AD7821 with an
input signal of 100.25 kHz. The SNR is 49.1 dB. It should be
noted that the harmonics are taken into account when calculating the SNR. The theoretical relationship between SNR and
resolution (N) is expressed by the following equation:
SNR = (6.02 N + 1.76) dB . . . . . . . . . . . . . . . . . . . . . (1)
INTERMODULATION DISTORTION
For intermodulation distortion (IMD), an FFT plot consisting
of very low distortion sine waves at two frequencies is generated
by sampling an analog input applied to the ADC. Figure 9
shows a 2048 point plot for IMD.
Figure 9. FFT Plot for IMD
HISTOGRAM PLOT
When a sine wave of specified frequency is applied to the VIN input of the AD7821 and several thousand samples are taken, it is
possible to plot a histogram showing the frequency of occurrence of each of the 256 ADC codes. A perfect ADC produces a
probability density function described by the equation:
P(V )=
π(A
1
2−V2)1/2
where A is the peak amplitude of the sine wave and P(V) is the
probability of occurrence at a voltage V.
If a particular step is wider than the ideal 1 LSB width, then the
code associated with that step will accumulate more counts than
for the code for an ideal step. Likewise, a step narrower than the
ideal width will have fewer counts. Missing codes are easily seen
because a missing code means zero counts for a particular code.
The absence of large spikes in the plot indicates small differential nonlinearity.
Figure 10 shows a histogram plot for the AD7821, which corresponds very well with the ideal shape. The plot indicates very
small differential nonlinearity and no missing codes for an input
frequency of 100.25 kHz.
Figure 8. AD7821 FFT Plot
EFFECTIVE NUMBER OF BITS
By working backwards from Equation (1) it is possible to get a
measure of ADC performance expressed in effective number of
bits (N). A plot of the effective number of bits versus input frequency is given in the Typical Performance Characteristics section. The effective number of bits typically falls between 7.7 and
7.9, corresponding to SNR figures of 48.1 and 49.7 dB.
–8–
Figure 10. AD7821 Histogram Plot
REV. A
AD7821
In digital signal processing applications, where the AD7821 is
used to sample ac signals, it is essential that the signal sampling
occurs at exactly equal intervals. This minimizes errors due to
sampling uncertainty or jitter. A precise timer or clock source,
to start the ADC conversion process, is the best method of generating equidistant sampling intervals.
The two modes of operation given in the data sheet are suitable
for DSP applications because the sampling instant of the
AD7821 is well defined. V
is sampled on the falling edge of
IN
WR or RD in the WR-RD or RD modes, respectively.
DIGITAL INTERFACE
The AD7821 has two basic interface modes which are determined by the status of the MODE pin. When this pin is low, the
converter is in the RD mode, with this pin high, the AD7821 is
set up for the WR-RD mode.
The RD mode is designed for microprocessors that can be
driven into a WAIT state. A READ operation (i.e.,
CS and RD
are taken low) starts a conversion and data is read when the
conversion is complete. The WR-RD mode does not require microprocessor WAIT states. A WRITE operation (i.e.,
CS and
WR are taken low) initiates a conversion, and a READ opera-
tion reads the result when the conversion is complete.
RD Mode (MODE = 0)
The timing diagram for the RD mode is shown in Figure 11.
This mode is intended for use with microprocessors which have
a WAIT state facility, whereby a READ instruction cycle can be
extended to accommodate slow memory devices. A conversion
is started by taking
CS and RD low (READ operation). Both
CS and RD are then kept low until output data appears.
INT typically goes low within 380 ns after the rising edge of
WR. It indicates that conversion is complete and that the data
result is in the output latch. With
(DB0–DB7) are activated when
the rising edge of
RD or CS.
Figure 12a. WR-RD Mode (tRD > t
CS low, the data outputs
RD goes low. INT is reset by
)
INTL
The alternative option can be used to shorten the conversion
time. This is a method for bypassing the internal time-out
circuit. The
250 ns after the rising edge of
INT line is ignored and RD can be brought low
WR. In this case RD going low
transfers the data result into the output latch and activates the
data output (DB0–DB7).
of
RD and is reset on the rising edge of RD or CS. The timing
INT is driven low on the falling edge
for this interface is shown in Figure 12b.
Figure 11. RD Mode
In this mode, Pin 6 of the AD7821 is configured as a status output, RDY. This RDY output can be used to drive the processor
READY or WAIT input. It is an open drain output (no internal
pull-up device) which goes low after the falling edge of
goes high impedance at the end of conversion. An
CS and
INT line is
also provided which goes low when a conversion is complete.
INT returns high on the rising edge of CS or RD.
WR-RD Mode (MODE = 1)
In the WR-RD mode, Pin 6 is configured as a WRITE (WR) input for the AD7821. With
falling edge of
WR. Two options exist for reading data from the
CS low, conversion is initiated on the
converter.
In the first of these options the processor waits for the
INT sta-
tus line to go low before reading the data (see Figure 12a).
REV. A
–9–
Figure 12b. WR-RD Mode (tRD < t
INTL
)
The AD7821 can also be used in stand-alone operation in the
WR-RD mode.
tiated by bringing
t
) after the rising edge of WR. The timing diagram for this
ID
CS and RD are tied low, and a conversion is ini-
WR low. Output data is valid 530 ns (t
INTL
+
mode is shown in Figure 13.
Figure 13. WR-RD Mode Stand-Alone Operation,
CS
= RD = 0
AD7821
MICROPROCESSOR INTERFACING
The AD7821 is designed for easy interfacing to microprocessors
as a memory mapped peripheral or an I/O device. This reduces
to a minimum the amount of external logic required for
interfacing.
AD7821 – 68008 INTERFACE
Figure 14 shows an AD7821 interface to the 68008 microprocessor. The ADC is configured for the RD interface mode. This
means that one read instruction starts a conversion and reads
the result when the conversion is completed. The read cycle is
stretched out over the entire conversion period by taking the
INT line back to the DTACK input of the 68008. Starting a
conversion and reading the relevant data consists of a <MOVE
B Dn, addr> instruction, where addr is the decoded ADC address and Dn is the data register into which the result is placed.
AD7821 – TMS32010 INTERFACE
A typical interface to the TMS32010 is shown in Figure 16. The
AD7821 is mapped at a port address and the interface is designed
for the maximum TMS32010 clock frequency of 20 MHz. In this
case, the AD7821 is configured in the WR-RD interface mode.
This means that a write instruction starts a conversion and a
read instruction reads the result when the conversion is completed. A precise timer or clock source is used to start a conversion in applications requiring equidistant sampling intervals.
The scheme used, whereby the AD7821 generates an interrupt
to the TMS32010, is limited in that it does not allow the
AD7821 to be sampled at its maximum rate. This is because the
time between samples has to be long enough to allow the
TMS32010 to service its interrupt and read data from the
AD7821. Constant interruption of the TMS32010 by the
AD7821, every time the ADC completes a conversion, is not a
very efficient use of the processor time. To overcome these
problems, some buffer memory or FIFO could be placed between the AD7821 and the TMS32010. The
AD7821 could be used to trigger a pulse which drives its
and
RD lines and places the AD7821 data into a FIFO or buffer
memory. The microprocessor can then read a batch of data
from the FIFO or buffer memory at some convenient time.
Reading data from the AD7821, after an
ceived, consists of <IN A, PA> instruction (PA is the decoded
ADC address).
INT line of the
CS
INT has been re-
Figure 14. AD7821 to 68008 Interface
AD7821 – 8088 INTERFACE
A typical interface to the 8088 is shown in Figure 15. The
AD7821 is configured for the RD interface mode. One read instruction starts a conversion and reads the result. The read cycle
is stretched out over the entire conversion period by taking the
RDY line back to the READY input of the 8088. Starting a conversion and reading the result consists of a <MOV AX, (addr)>
instruction, where addr is the decoded ADC address and AX is
the 8088 data register into which the conversion result is placed.
Figure 15. AD7821 to 8088 Interface
Figure 16. AD7821 to TMS32010 Interface
AD7821 – 8051 INTERFACE
Figure 17 shows the AD7821 interface to the 8051 microcomputer. The AD7821 is configured in the WR-RD interface mode
and is connected to the 8051 ports. The processor starts conversion and then polls
conversion result. Data is read from the AD7821 by using the
<MOV A, 90H> instruction (90H is the address for Port 1).
INT, until it goes low, before reading the
–10–
Figure 17. AD7821 to 8051 Interface
REV. A
AD7821
APPLYING THE AD7821
The AD7821 is specified for a unipolar input range of 0 to +5 V
and a bipolar input range of –2.5 V to +2.5 V. The V
V
(+) voltages required for these input ranges are outlined be-
REF
REF
(–) and
low. See the Typical Performance Characteristics section for operation with unspecified input voltage ranges.
UNIPOLAR OPERATION
Figure 18 gives the configuration and reference voltages required for 0 V to +5 V operation. The nominal transfer characteristic for this input range is shown in Figure 19. The output
code is Natural Binary with 1 LSB = (5/256) V = 19.5 mV.
Figure 18. AD7821 Unipolar/Bipolar Operation
BIPOLAR OPERATION
Figure 18 gives the configuration and reference voltages
required for –2.5 V to +2.5 V operation. The nominal transfer
characteristic for this input range is shown in Figure 20.
The output code is Offset Binary with 1 LSB = ([+2.5 –
(–2.5)]/256) V = 19.5 mV.
Figure 20. Nominal Transfer Characteristic for Bipolar
(–2.5 V to +2.5 V) Operation
16-CHANNEL TELECOM A/D CONVERTER
The fast sampling rate (1 MHz) and bipolar operation of the
AD7821 makes it useful in Telecom applications for sampling a
number of input channels using a multiplexer. Figure 21 shows
a circuit for such an application.
The maximum signal frequency required for acceptable quality
in Telecom applications is 3 kHz. The circuit given in Figure 21
permits each of the 16-input channels to be sampled at a rate of
16 kHz maximum. The sampling rate takes account of such
multiplexer parameters as t
, settling time etc. The circuit also
ON
eases the problem of the antialiasing filter design by sampling at
a rate much greater than that required by the Nyquist criterion.
Figure 19. Nominal Transfer Characteristic for Unipolar
(0 V to +5 V) Operation
REV. A
Figure 21. 16-Channel Telecom A/D Converter System
–11–
AD7821
SIMULTANEOUS SAMPLING A/D CONVERTERS
The AD7821’s inherent track-and-hold and well-defined sampling instant makes it useful, in such applications as sonar,
where a number of input channels are required to be sampled
simultaneously. Figure 22 shows a circuit for such an application.
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
20-Pin Plastic DIP (N-20)
C1186–10–4/88
20-Pin Cerdip (Q-20)
Figure 22. Simultaneous Sampling A/D Converters
The actual sampling instant, which is the instant at which VIN is
measured, occurs approximately 50 ns after the falling edge of
WR or RD in the WR-RD or RD modes, respectively, due to in-
ternal logic delays. However, the internal logic delay and, therefore, the sampling instant can vary from device to device, but is
typically within ±5 ns. This means that a maximum common input sine wave of ±2.5 V at 32 kHz, applied to any number of
AD7821s in the circuit of Figure 22, will yield a maximum difference between the converter outputs of typically ± 1/4 LSB.
20-Terminal Plastic Leaded Chip Carrier
(P-20A)
20-Terminal Leadless Ceramic Chip Carrier
(E-20A)
PRINTED IN U.S.A.
–12–
REV. A
Loading...