The MAX114/MAX118 are microprocessor-compatible,
8-bit, 4-channel and 8-channel analog-to-digital converters (ADCs). They operate from a single +5V supply
and use a half-flash technique to achieve a 660ns conversion time (1Msps). A power-down (PWRDN) pin
reduces current consumption typically to 1µA. The
devices return from power-down mode to normal operating mode in less than 200ns, allowing large supplycurrent reductions in burst-mode applications (in burst
mode, the ADC wakes up from a low-power state at
specified intervals to sample the analog input signals).
Both converters include a track/hold, enabling the ADC
to digitize fast analog signals.
____________________________Features
♦ Single +5V Supply Operation
♦ 4 (MAX114) or 8 (MAX118) Analog Input Channels
♦ Low Power: 40mW (operating mode)
5µW (power-down mode)
♦ Total Unadjusted Error ≤1LSB
♦ Fast Conversion Time: 660ns per Channel
♦ No External Clock Required
♦ Internal Track/Hold
♦ 1MHz Full-Power Bandwidth
♦ Internally Connected 8th Channel Monitors
Reference Voltage (MAX118)
Microprocessor (µP) interfaces are simplified because the
ADC can appear as a memory location or I/O port without
external interface logic. The data outputs use latched,
three-state buffer circuitry for direct connection to an 8-bit
parallel µP data bus or system input port. The
MAX114/MAX118 input/reference configuration enables
ratiometric operation.
The 4-channel MAX114 is available in a 24-pin DIP or
SSOP. The 8-channel MAX118 is available in a 28-pin
DIP or SSOP. For +3V applications, refer to the
MAX113/MAX117 data sheet.
________________________Applications
High-Speed DSPRemote Data Acquisition
Portable EquipmentCommunications Systems
______________Ordering Information
PART
MAX114CNG
MAX114CAG
MAX114C/D0°C to +70°C
MAX114ENG
MAX114EAG
MAX114MRG-55°C to +125°C
Ordering Information continued on last page.
*Dice are specified at T
**Contact factory for availability.
Stresses beyond those listed under “Absolute Maximum Ratings” may cause permanent damage to the device. These are stress ratings only, and functional
operation of the device at these or any other conditions beyond those indicated in the operational sections of the specifications is not implied. Exposure to
absolute maximum rating conditions for extended periods may affect device reliability.
Note 4: Input control signals are specified with tr = tf= 5ns, 10% to 90% of 5V, and timed from a voltage level of 1.6V.
Note 5: See Figure 1 for load circuit. Parameter defined as the time required for the output to cross 0.8V or 2.4V.
Note 6: See Figure 2 for load circuit. Parameter defined as the time required for the data lines to change 0.5V.
Note 7: Also defined as the Minimum Address-Valid to Convert-Start Time.
Analog Input Channel 6IN6—
Analog Input Channel 5IN5—2
Analog Input Channel 4IN41
Analog Input Channel 3
Analog Input Channel 2IN23
Analog Input Channel 1
Mode Selection Input. Internally pulled low with a 50µA current source. MODE = 0 acti-
vates read mode; MODE = 1 activates write-read mode (see
Three-State Data Output (LSB)D06
Three-State Data OutputsD1, D2, D37, 8, 9
Read Input. RD must be low to access data (see
Interrupt Output. INT goes low to indicate end of conversion (see
section).
GroundGND12
Lower Limit of Reference Span. REF- sets the zero-code voltage. Range is GND ≤
V
< V
REF-
Upper Limit of Reference Span. REF+ sets the full-scale input voltage. Range is V
< V
REF+
Write-Control Input/Ready-Status Output (see
Chip-Select Input. CS must be low for the device to recognize WR or RD inputs.CS
Three-State Data OutputsD4, D5, D617, 18, 19
Three-State Data Output (MSB)D720
Multiplexer Channel Address Input (MSB)A2—
Multiplexer Channel Address Input A121
Multiplexer Channel Address Input (LSB)A022
Power-Down Input. PWRDN reduces supply current when low.PWRDN
Positive Supply, +5VV
The MAX114/MAX118 use a half-flash conversion technique (see
Functional Diagram
ADC sections achieve an 8-bit result. Using 15 comparators, the flash ADC compares the unknown input
voltage to the reference ladder and provides the upper
four data bits. An internal digital-to-analog converter
(DAC) uses the four most significant bits (MSBs) to
generate both the analog result from the first flash conversion and a residue voltage that is the difference
between the unknown input and the DAC voltage. The
residue is then compared again with the flash comparators to obtain the lower four data bits (LSBs).
An internal analog multiplexer enables the devices to
read four (MAX114) or eight (MAX118) different analog
voltages under microprocessor (µP) control. One of the
MAX118’s analog channels, IN8, is internally hardwired and always reads V
DATA
OUTPUTS
RL = 3k
a) HIGH-Z TO V
Figure 1. Load Circuits for Data-Access Time Test
DATA
OUTPUTS
3k
a) V
Figure 2. Load Circuits for Data-Hold Time Test
OH
TO HIGH-Z b) V
OH
Converter Operation
) in which two 4-bit flash
when selected.
REF+
V
DD
RL = 3k
C
b) HIGH-Z TO V
V
DD
3k
10pF
TO HIGH-Z
OL
C
10pF
L
DATA
OUTPUTS
DATA
OUTPUTS
L
OL
In burst-mode or low sample-rate applications, the
Power-Down Mode
MAX114/MAX118 can be shut down between conversions, reducing supply current to microamp levels (see
Typical Operating Characteristics
). A logic low on the
PWRDN pin shuts the devices down, reducing supply
current typically to 1µA when powered from a single +5V
supply. A logic high on PWRDN wakes up the
MAX114/MAX118, and the selected analog input enters
the track mode. The signal is fully acquired after 360ns
(this includes both the power-up delay and the track/hold
acquisition time), and a new conversion can be started.
If the power-down feature is not required, connect
PWRDN to VDD. For minimum current consumption, keep
digital inputs at the supply rails in power-down mode.
Refer to the
Reference
section for information on reduc-
ing reference current during power-down.
___________________Digital Interface
The MAX114/MAX118 have two basic interface modes,
which are set by the MODE pin. When MODE is low,
the converters are in read mode; when MODE is high,
the converters are set up for write-read mode. The A0,
A1, and A2 inputs control channel selection, as shown
in Table 1. The address must be valid for a minimum
time, t
Table 1. Truth Table for Input Channel
Selection
In read mode, conversions and data access are controlled by the RD input (Figure 3). The comparator
inputs track the analog input voltage for the duration of
t
ACQ
With µPs that can be forced into a wait state, hold RD
low until output data appears. The µP starts the conversion, waits, and then reads data with a single read
instruction.
+5V, 1Msps, 4 & 8-Channel,
8-Bit ADCs with 1µA Power-Down
In read mode, WR/RDY is configured as a status output
(RDY), so it can drive the ready or wait input of a µP.
RDY is an open-collector output (no internal pull-up)
that goes low after the falling edge of CS and goes high
at the end of the conversion. If not used, the WR/RDY
pin can be left unconnected. The INT output goes low
at the end of the conversion and returns high on the rising edge of CS or RD.
Write-Read Mode (MODE = 1)
Figures 4 and 5 show the operating sequence for writeread mode. The comparator inputs track the analog
input voltage for the duration of t
. The conversion is
ACQ
initiated by a falling edge of WR. When WR returns
high, the result of the four-MSBs flash is latched into the
output buffers and the conversion of the four-LSBs flash
starts. INT goes low, indicating conversion end, and the
MAX114/MAX118
lower four data bits are latched into the output buffers.
The data is then accessible after RD goes low (see
Timing Characteristics
A minimum acquisition time (t
).
) is required from INT
ACQ
going low to the start of another conversion (WR going
low).
Options for reading data from the converter include
using internal delay, reading before delay, and pipelined
operation (discussed in the following sections).
Using Internal Delay
The µP waits for the INT output to go low before reading
the data (Figure 4). INT goes low after the rising edge of
WR, indicating that the conversion is complete and the
result is available in the output latch. With CS low, data
outputs D0–D7 can be accessed by pulling RD low. INT
is then reset by the rising edge of CS or RD.
Fastest Conversion:
Reading Before Delay
Figure 5 shows an external method of controlling the
conversion time. The internally generated delay (t
INTL
varies slightly with temperature and supply voltage,
and can be overridden with RD to achieve the fastest
conversion time. RD is brought low after the rising edge
of WR, but before INT goes low. This completes the
conversion and enables the output buffers that contain
the conversion result (D0–D7). INT also goes low after
the falling edge of RD and is reset on the rising edge of
RD or CS. The total conversion time is therefore: tWR+
tRD+ t
pipelined operation can be achieved by connecting
WR to RD (Figure 6). With CS low, driving WR and RD
low initiates a conversion and concurrently reads the
result of the previous conversion.
_____________Analog Considerations
MAX114
MAX118
Reference
Figures 7a, 7b, and 7c show typical reference connections. The voltages at REF+ and REF- set the ADC’s
analog input range (see Figure 10). The voltage at REFdefines the input that produces an output code of all
zeros, and the voltage at REF+ defines the input that
produces an output code of all ones.
The internal resistance from REF+ to REF- can be as
low as 1kΩ, and current will flow through it even when
the MAX114/MAX118 are shut down. Figure 7d shows
how an N-channel MOSFET can be connected to REF-
Figure 7d. An N-channel MOSFET switches off the reference
load during power-down
Page 10
+5V, 1Msps, 4 & 8-Channel,
8-Bit ADCs with 1µA Power-Down
to break this current path during power-down. The FET
should have an on-resistance of less than 2Ω with a 5V
gate drive. When REF- is switched, as in Figure 7d, a
new conversion can be initiated after waiting a period
of time equal to the power-up delay (tUP) plus the Nchannel FET’s turn-on time.
Although REF+ is frequently connected to VDD, the circuit of Figure 7d uses a low-current, low-dropout,
4.096V voltage reference: the MAX874. Since the
MAX874 cannot continuously furnish enough current for
the reference resistance, this circuit is intended for
applications where the MAX114/MAX118 are normally
in standby and are turned on in order to make measurements at intervals greater than 65µs. C1 (the
capacitor connected to REF+) is slowly charged by the
MAX874 during the standby period, and furnishes the
MAX114/MAX118
reference current during the short measurement period.
C1’s 3.3µF value ensures a voltage drop of less than
1/2LSB when performing four to eight successive conversions. Larger capacitors reduce the error still further.
Use ceramic or tantalum capacitors for C1.
Initial Power-Up
When power is first applied, perform a conversion to initialize the MAX114/MAX118. Disregard the output data.
Bypassing
Use a 4.7µF electrolytic in parallel with a 0.1µF ceramic
capacitor to bypass VDDto GND. Minimize capacitor
lead lengths.
Bypass the reference inputs with 0.1µF capacitors, as
shown in Figures 7a, 7b, and 7c.
Figure 8 shows the equivalent circuit of the MAX114/
Analog Inputs
MAX118 input. When a conversion starts and WR is
low, V
is connected to sixteen 0.6pF capacitors.
IN_
During this acquisition phase, the input capacitors
charge to the input voltage through the resistance of
the internal analog switches. In addition, about 22pF of
stray capacitance must be charged. The input can be
modeled as an equivalent RC network (Figure 9). As
source impedance increases, the capacitors take
longer to charge.
The typical 32pF input capacitance allows source resistance as high as 800Ω without setup problems. For
larger resistances, the acquisition time (t
ACQ
) must be
increased.
Internal protection diodes, which clamp the analog
input to VDDand GND, allow the channel input pins to
swing from GND - 0.3V to VDD+ 0.3V without damage.
However, for accurate conversions near full scale, the
inputs must not exceed VDDby more than 50mV or be
lower than GND by 50mV.
If the analog input exceeds 50mV beyond the supplies, limit the input current to no more than 2mA,
as excessive current will degrade the conversion
accuracy of the on channel.
Track/Hold
The track/hold enters hold mode when a conversion
starts (RD low or WR low). INT goes low at the end of
the conversion, at which point the track/hold enters
track mode. The next conversion can start after the minimum acquisition time, t
Figure 10 shows the MAX114/MAX118’s nominal trans-
Transfer Function
fer function. Code transitions occur halfway between
successive-integer LSB values. Output coding is binary
with 1LSB = (V
REF+
- V
REF-
) / 256.
Conversion Rate
The maximum sampling rate (f
MAX118 is achieved in write-read mode (tRD< t
) for the MAX114/
MAX
INTL
and is calculated as follows:
f=
MAX
t + t + t + t
WRRDRIACQ
f
=
MAX
250ns 250ns 150ns 160ns
f 1.23MHz
=
MAX
1
1
+++
where tWR= the write pulse width, tRD= the delay
between write and read pulses, tRI= RD to INT delay,
and t
= minimum acquisition time.
ACQ
Signal-to-Noise Ratio and
Effective Number of Bits
Signal-to-noise plus distortion (SINAD) is the ratio of the
fundamental input frequency’s RMS amplitude to all
other ADC output signals. The output spectrum is limited to frequencies above DC and below one-half the
ADC sample rate.
The theoretical minimum analog-to-digital noise is
caused by quantization error, and results directly from
the ADC’s resolution: SNR = (6.02N + 1.76)dB, where
N is the number of bits of resolution. Therefore, a perfect 8-bit ADC can do no better than 50dB.
The FFT Plot (see
Typical Operating Characteristics
shows the result of sampling a pure 195.8kHz sinusoid
at a 1MHz rate. This FFT plot of the output shows the
output level in various spectral bands.
The effective resolution (or “effective number of bits”)
the ADC provides can be measured by transposing the
equation that converts resolution to SNR: N = (SINAD -
1.76) / 6.02 (see
Typical Operating Characteristics
).
Total Harmonic Distortion
Total harmonic distortion (THD) is the ratio of the RMS
sum of all harmonics of the input signal (in the frequency
OUTPUT CODE
11111111
11111110
11111101
FULL-SCALE
TRANSITION
),
1LSB =
00000011
00000010
00000001
00000000
V
REF-
Figure 10. Transfer Function
123
INPUT VOLTAGE (LSBs)
FS - 1LSB
band above DC and below one-half the sample rate) to
the fundamental itself. This is expressed as:
where V
+++
VVV...V
22324
THD = 20log
is the fundamental RMS amplitude, and V
1
2
V
1
through VNare the amplitudes of the 2nd through Nth
harmonics.
)
Spurious-Free Dynamic Range
Spurious-free dynamic range (SFDR) is the ratio of the
fundamental RMS amplitude to the amplitude of the
next largest spectral component (in the frequency
band above DC and below one-half the sample rate).
Usually the next largest spectral component occurs at
some harmonic of the input frequency. However, if the
ADC is exceptionally linear, it may occur only at a random peak in the ADC’s noise floor. See the Signal-toNoise Ratio graph in
Maxim cannot assume responsibility for use of any circuitry other than circuitry entirely embodied in a Maxim product. No circuit patent licenses are
Maxim cannot assume responsibility for use of any circuitry other than circuitry entirely embodied in a Maxim product. No circuit patent licenses are
implied. Maxim reserves the right to change the circuitry and specifications without notice at any time.
implied. Maxim reserves the right to change the circuitry and specifications without notice at any time.
12
__________________Maxim Integrated Products, 120 San Gabriel Drive, Sunnyvale, CA 94086 (408) 737-7600
12
__________________Maxim Integrated Products, 120 San Gabriel Drive, Sunnyvale, CA 94086 (408) 737-7600