The MAX378 8-channel single-ended (1-of-8) multiplexer
and the MAX379 4-channel differential (2-of-8) multiplexer
use a series N-channel/P-channel/N-channel structure to
provide significant fault protection. If the power supplies to
the MAX378/MAX379 are inadvertently turned off while
input voltages are still applied,
all
channels in the muxes
are turned off, and only a few nanoamperes of leakage current will flow into the inputs. This protects not only the
MAX378/MAX379 and the circuitry they drive, but also the
sensors or signal sources that drive the muxes.
The series N-channel/P-channel/N-channel protection
structure has two significant advantages over the simple
current-limiting protection scheme of the industry’s firstgeneration fault-protected muxes. First, the Maxim protection scheme limits fault currents to nanoamp leakage
values rather than many milliamperes. This prevents damage to sensors or other sensitive signal sources. Second,
the MAX378/MAX379 fault-protected muxes can withstand
a
continuous
±60V input, unlike the first generation, which
had a continuous ±35V input limitation imposed by power
dissipation considerations.
All digital inputs have logic thresholds of 0.8V and 2.4V,
ensuring both TTL and CMOS compatibility without requiring pull-up resistors. Break-before-make operation is
guaranteed. Power dissipation is less than 2mW.
________________________Applications
Data Acquisition Systems
Industrial and Process Control Systems
Avionics Test Equipment
Signal Routing Between Systems
____________________________Features
♦ Fault Input Voltage ±75V with Power Supplies Off
♦ Fault Input Voltage ±60V with ±15V Power Supplies
♦ All Switches Off with Power Supplies Off
♦ On Channel Turns OFF if Overvoltage Occurs on
Input or Output
♦ Only Nanoamperes of Input Current Under All
Fault Conditions
♦ No Increase in Supply Currents Due to Fault
Conditions
♦ Latchup-Proof Construction
♦ Operates from ±4.5V to ±18V Supplies
♦ All Digital Inputs are TTL and CMOS Compatible
♦ Low-Power Monolithic CMOS Design
______________Ordering Information
PART
MAX378CPE
MAX378CWG
MAX378CJE0°C to +70°C
MAX378C/D0°C to +70°CDice**
MAX378EPE
MAX378EWG-40°C to +85°C
MAX378EJE
MAX378MJE-55°C to +125°C
MAX378MLP-55°C to +125°C20 LCC*
Ordering Information continued at end of data sheet.
* Contact factory for availability.
**The substrate may be allowed to float or be tied to V+ (JI CMOS).
Call toll free 1-800-998-8800 for free samples or literature.
1
High-Voltage, Fault-Protected
Analog Multiplexers
ABSOLUTE MAXIMUM RATINGS
Voltage between Supply Pins..............................................+44V
V+ to Ground...................................................................+22V
V- to Ground......................................................................-22V
Digital Input Overvoltage:
VEN, V
Analog Input with Multiplexer Power On..............................±65V
Analog Input with Multiplexer Power Off..............................±80V
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 2: When the analog signal exceeds +13.5V or -12V, the blocking action of Maxim’s gate structure goes into operation. Only
leakage currents flow and the channel ON resistance rises to infinity.
Note 3: The value shown is the steady-state value. The transient leakage is typically 50µA. See
Note 4: Guaranteed by other static parameters.
Note 5: Digital input leakage is primarily due to the clamp diodes. Typical leakage is less than 1nA at +25°C.
Note 6: Leakage currents not tested at T
Note 7: Electrical characteristics, such as ON Resistance, will change when power supplies other than ±15V are used.
Figure 6 shows a typical data acquisition system
using the MAX378 multiplexer. Since the multiplexer
is driving a high-impedance input, its error is a function of its own resistance (R
plexer leakage current (I
bias current (I
V
ERR
= R
BIAS
DS(ON)
):
x (I
OUT(ON)
OUT(ON)
= 2.0kΩ x (2nA + 30pA)
= 18.0µV maximum error
In most cases, this error is low enough that preamplification of input signals is not needed, even with very
low-level signals such as 40µV/°C from type J thermocouples.
In systems with fewer than eight inputs, an unused channel can be connected to the system ground reference
point for software zero correction. A second channel
connected to the system voltage reference allows gain
correction of the entire data acquisition system as well.
A MAX420 precision op amp is connected as a programmable-gain amplifier, with gains ranging from 1 to
10,000. The guaranteed 5µV unadjusted offset of the
MAX420 maintains high signal accuracy, while programmable gain allows the output signal level to be scaled to
the optimum range for the remainder of the data acquisition system, normally a Sample/Hold and A/D. Since
the gain-changing multiplexer is not connected to the
external sensors, it can be either a DG508A multiplexer
or the fault-protected MAX358 or MAX378.
High-Voltage, Fault-Protected
Analog Multiplexers
Input switching, however, must be done with a faultprotected MAX378 multiplexer, to provide the level of
protection and isolation required with most data acquisition inputs. Since external signal sources may continue to supply voltage when the multiplexer and system
power are turned off, non-fault-protected multiplexers,
or even first-generation fault-protected devices, will
allow many milliamps of fault current to flow from outside sources into the multiplexer. This could result in
damage to either the sensors or the multiplexer. A nonfault-protected multiplexer will also allow input overvoltages to appear at its output, perhaps damaging
Sample/Holds or A/Ds. Such input overdrives may also
cause input-to-input shorts, allowing the high current
output of one sensor to possibly damage another.
The MAX378 eliminates all of the above problems. It
not only limits its output voltage to safe levels, with or
without power applied (V+ and V-), but also turns all
channels off when power is removed. This allows it to
draw only sub-microamp fault currents from the inputs,
and maintain isolation between inputs for continuous
input levels up to ±75V with power supplies off.
_______________Detailed Description
Fault Protection Circuitry
The MAX378/MAX379 are fully fault protected for continuous input voltages up to ±60V, whether or not the V+
and V- power supplies are present. These devices use
a “series FET” switching scheme which not only protects the multiplexer output from overvoltage, but also
limits the input current to sub-microamp levels.
Figures 7 and 8 show how the series FET circuit protects against overvoltage conditions. When power is
off, the gates of all three FETs are at ground. With a -60V
input, N-channel FET Q1 is turned on by the +60V gate-
+60V
OVERVOLTAGE
N-CHANNEL MOSFET
IS TURNED OFF
GS
= -60V
BECAUSE V
Q
1
S
D
G
Q
2
D
S
G
Q
3
D
S
G
Figure 8. +60V Overvoltage with Multiplexer Power OFF
-60V
OVERVOLTAGE
N-CHANNEL MOSFET
IS TURNED OFF
GS
= +45V
BECAUSE V
-15V+15V-15V
-60V
Q
2
+15V FROM
DRIVERS
Q
1
-15V FROM
DRIVERS
P-CHANNEL
MOSFET IS OFF
+60V FORCED
ON COMMON
OUTPUT
LINE BY
EXTERNAL
CIRCUITRY
Q
3
N-CHANNEL
MOSFET IS OFF
Figure 9. -60V Overvoltage on an OFF Channel with
Multiplexer Power Supply ON
MAX378/MAX379
-60V
OVERVOLTAGE
N-CHANNEL MOSFET
IS TURNED ON
GS
= +60V
BECAUSE V
Q
1
S
G
P-CHANNEL
MOSFET IS OFF
Q
-60V
2
D
S
D
G
Q
3
D
S
G
Figure 7. -60V Overvoltage with Multiplexer Power OFF
Figure 10. +60V Overvoltage Input to the ON Channel
+13.5V
OUTPUT
High-Voltage, Fault-Protected
Analog Multiplexers
to-source voltage. The P-channel device (Q2), however, has +60V VGSand is turned off, thereby preventing
the input signal from reaching the output. If the input
voltage is +60V, Q1 has a negative VGS, which turns it
off. Similarly, only sub-microamp leakage currents can
flow from the output back to the input, since any voltage will turn off either Q1 or Q2.
Figure 9 shows the condition of an OFF channel with
V+ and V- present. As with Figures 7 and 8, either an
N-channel or a P-channel device will be off for any
input voltage from -60V to +60V. The leakage current
with negative overvoltages will immediately drop to a
few nanoamps at +25°C. For positive overvoltages,
that fault current will initially be 40µA or 50µA, decaying
over a few seconds to the nanoamp level. The time
constant of this decay is caused by the discharge of
MAX378/MAX379
stored charge from internal nodes, and does not compromise the fault-protection scheme.
Figure 10 shows the condition of the ON channel with
V+ and V- present. With input voltages less than ±10V,
all three FETs are on and the input signal appears at the
output. If the input voltage exceeds V+ minus the Nchannel threshold voltage (VTN), then the N-channel
FET will turn off. For voltages more negative than Vminus the P-channel threshold (VTP), the P-channel
device will turn off. Since VTNis typically 1.5V and V
is typically 3V, the multiplexer’s output swing is limited
to about -12V to +13.5V with ±15V supplies.
The
Typical Operating Characteristics
graphs show typical leakage vs. input voltage curves. Although the maximum rated input of these devices is ±65V, the
MAX378/MAX379 typically have excellent performance
up to ±75V, providing additional margin for the unknown
transients that exist in the real world. In summary, the
MAX378/MAX379 provide superior protection from all
fault conditions while using a standard, readily produced junction-isolated CMOS process.
Switching Characteristics
and Charge Injection
Table 1 shows typical charge-injection levels vs.
power-supply voltages and analog input voltage. Note
that since the channels are well matched, the differential charge injection for the MAX379 is typically less
than 5pC. The charge injection that occurs during
switching creates a voltage transient whose magnitude
is inversely proportional to the capacitance on the multiplexer output.
The channel-to-channel switching time is typically 600ns,
with about 200ns of break-before-make delay. This 200ns
break-before-make delay prevents the input-to-input short
that would occur if two input channels were simultaneous-
ly connected to the output. In a typical data acquisition
system, such as in Figure 6, the dominant delay is not the
switching time of the MAX378 multiplexer, but is the settling time of the following amplifiers and S/H. Another limiting factor is the RC time constant of the multiplexer
R
plus the signal source impedance multiplied by
DS(ON)
the load capacitance on the output of the multiplexer.
Even with low signal source impedances, 100pF of capacitance on the multiplexer output will approximately double
the settling time to 0.01% accuracy.
Operation with Supply Voltage
Other than ±15V
The main effect of supply voltages other than ±15V is
the reduction in output signal range. The MAX378 limits
the output voltage to about 1.5V below V+ and about 3V
above V-. In other words, the output swing is limited to
+3.5V to -2V when operating from ±5V. The
Operating Characteristics
graphs show typical R
for ±15V, ±10V, and ±5V power supplies. Maxim tests
and guarantees the MAX378/MAX379 for operation from
±4.5V to ±18V supplies. The switching delays are
increased by about a factor of 2 at ±5V, but breakbefore-make action is preserved.
The MAX378/MAX379 can be operated with a single +9V
to +22V supply, as well as asymmetrical power supplies
TP
such as +15V and -5V. The digital threshold will remain
approximately 1.6V above GND and the analog characteristics such as R
are determined by the total voltage
DS(ON)
difference between V+ and V-. Connect V- to 0V when
operating with a +9V to +22V single supply.
This means that the MAX378/MAX379 will operate with
standard TTL-logic levels, even with ±5V power supplies. In all cases, the threshold of the EN pin is the
same as the other logic inputs.
Table 1a. MAX378 Charge Injection
Supply Voltage Analog Input Level Injected Charge
±5V
±10V
±15V
Test Conditions: CL= 1000pF on multiplexer output; the tabulated analog input level is applied to channel 1; channels 2
through 8 are open circuited. EN = +5V, A1 = A2 = 0V, A0 is
toggled at 2kHz rate between 0V and 3V. +100pC of charge
creates a +100mV step when injected into a 1000pF load
capacitance.
Test Conditions: CL= 1000pF on Out A and Out B; the tabulated analog input level is applied to inputs 1A and 1B; channels
2 through 4 are open circuited. EN = +5V, A1 = 0V, A0 is toggled from 0V to 3V at a 2kHz rate.
Analog
Input Level
+1.7V
0V
-1.7V
+5V
0V
-5V
+10V
0V
-10V
The typical digital threshold of both the address lines
and the EN pin is 1.6V, with a temperature coefficient of
about -3mV/°C. This ensures compatibility with 0.8V to
2.4V TTL-logic swings over the entire temperature
range. The digital threshold is relatively independent of
the supply voltages, moving from 1.6V typical to 1.5V
typical as the power supplies are reduced from ±15V to
±5V. In all cases, the digital threshold is referenced to
GND.
The digital inputs can also be driven with CMOS-logic
levels swinging from either V+ to V- or from V+ to GND.
The digital input current is just a few nanoamps of leakage at all input voltage levels, with a guaranteed maximum of 1µA. The digital inputs are protected from ESD
by a 30V zener diode between the input and V+, and
can be driven ±4V beyond the supplies without drawing
excessive current.
Operation as a Demultiplexer
The MAX378/MAX379 will function as a demultiplexer,
where the input is applied to the OUT pin, and the input
pins are used as outputs. The MAX378/MAX379 provide both break-before-make action and full fault protection when operated as a demultiplexer, unlike earlier
generations of fault-protected multiplexers.
Channel-to-Channel Crosstalk,
Off Isolation, and Digital Feedthrough
At DC and low frequencies, channel-to-channel
crosstalk is caused by variations in output leakage cur-
Injected Charge
Out AOut B
+105pC
+73pC
+48pC
+215pC
+135pC
+62pC
+525pC
+180pC
+55pC
+107pC
+74pC
+50pC
+220pC
+139pC
+63pC
+530pC
+185pC
+55pC
Differential
A-B
-2pC
-1pC
-2pC
-5pC
-4pC
-1pC
-5pC
-5pC
0pC
Digital Interface Levels
rents as the off-channel input voltages are varied. The
MAX378 output leakage varies only a few picoamps as
all seven off inputs are toggled from -10V to +10V. The
output voltage change depends on the impedance level
at the MAX378 output, which is R
DS(ON)
plus the input
signal source resistance in most cases, since the load
driven by the MAX378 is usually a high impedance. For
a signal source impedance of 10kΩ or lower, the DC
crosstalk exceeds 120dB.
Table 2 shows typical AC crosstalk and off-isolation performance. Digital feedthrough is masked by the analog
charge injection when the output is enabled. When the
output is disabled, the digital feedthrough is virtually
unmeasurable, since the digital pins are physically isolated from the analog section by the GND and V- pins.
The ground plane formed by these lines is continued
onto the MAX378/MAX379 die to provide over 100dB
isolation between the digital and analog sections.
Table 2a. Typical Off-Isolation
Rejection Ratio
Frequency100kHz500kHz1MHz
One Channel Driven
All Channels Driven
Test Conditions: VIN= 20V
= 1.5kΩ between OUT and GND, EN = 0V.
R
L
OIRR = 20 Log ____________
74dB72dB66dB
64dB48dB44dB
at the tabulated frequency,
P-P
20V
P-P
V
OUT (P-P)
Table 2b. Typical Crosstalk
Rejection Ratio
Frequency
FL= 1.5k
RL= 10k
Test Conditions: Specified RLconnected from OUT to GND,
EN = +5V, A0 = A1 = A2 = +5V (Channel 1 selected). 20V
at the tabulated frequency is applied to Channel 2. All other
channels are open circuited. Similar crosstalk rejection can be
observed between any two channels.
MAX379CWG
MAX379CJE0°C to +70°C
MAX379C/D0°C to +70°CDice**
MAX379EPE
MAX379EWG-40°C to +85°C
MAX379EJE
MAX379MJE-55°C to +125°C
MAX379MLP-55°C to +125°C20 LCC*
* Contact factory for availability.
**The substrate may be allowed to float or be tied to V+ (JI CMOS).
TEMP. RANGEPIN-PACKAGE
0°C to +70°C
0°C to +70°C
-40°C to +85°C16 Plastic DIP
-40°C to +85°C16 CERDIP
16 Plastic DIP
24 Wide SO
16 CERDIP
24 Wide SO
16 CERDIP
_________________Chip Topographies
MAX378
IN8 OUT
IN7
IN7
IN6
IN5
V+
GND
A2A1EN
0.151"
(3.835mm)
NOTE: Connect substrate to V+ or leave it floating.
MAX379
OUTB OUTA
IN4B
IN4
IN3
0.229"
(5.816mm)
IN2
IN1
V-
A0
IN4A
MAX378/MAX379
IN3B
IN2B
IN1B
V+
GND
A0
A1EN
0.151"
(3.835mm)
NOTE: Connect substrate to V+ or leave it floating.
________________________________________________________Package Information
INCHESMILLIMETERS
DIM
D
A
0.101mm
e
B
A1
0.004in.
C
0°- 8°
L
A
A1
B
C
E
e
H
L
MIN
0.093
0.004
0.014
0.009
0.291
0.394
0.016
0.050
MAX
0.104
0.012
0.019
0.013
0.299
0.419
0.050
MIN
2.35
0.10
0.35
0.23
7.40
10.00
0.40
1.27
MAX
2.65
0.30
0.49
0.32
7.60
10.65
1.27
MAX378/MAX379
HE
D1
D
A2
A
L
A1
Wide SO
SMALL-OUTLINE
PACKAGE
(0.300 in.)
E
E1
A3
α
DIM
D
D
D
D
D
DIM
A
A1
A2
A3
B
B1
C
D
D1
E
E1
e
e
e
L
α
A
B
INCHESMILLIMETERS
PINS
MIN
MAX
–
0.413
0.463
0.512
0.614
0.713
MIN
10.10
11.35
12.60
15.20
17.70
MIN
–
0.38
3.18
1.40
0.41
1.27
0.20
18.92
0.13
7.62
6.10
–
2.92
0˚
0.398
16
0.447
18
0.496
20
0.598
24
0.697
28
INCHESMILLIMETERS
MAX
MIN
0.200
–
0.015
0.150
0.125
0.080
0.055
0.022
0.016
0.065
0.050
0.012
0.008
0.765
0.745
0.030
0.005
0.325
0.300
0.280
0.240
0.100 BSC
0.300 BSC
0.400
–
0.150
0.115
15˚
0˚
21-0042A
2.54 BSC
7.62 BSC
MAX
10.50
11.75
13.00
15.60
18.10
MAX
5.08
–
3.81
2.03
0.56
1.65
0.30
19.43
0.76
8.26
7.11
10.16
3.81
15˚
21-587A
16-PIN PLASTIC
e
B1
B
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