www.fairchildsemi.com
Application Note AN-3001
Optocoupler Input Drive Circuits
An optocoupler is a combination of a light source and a
photosensitive detector. In the optocoupler, or photon
coupled pair, the coupling is achieved by light being
generated on one side of a transparent insulating gap and
being detected on the other side of the gap without an
electrical connection between the two sides (except for a
minor amount of coupling capacitance). In the Fairchild
Semiconductor optocouplers, the light is generated by an
infrared light emitting diode, and the photo-detector is a
silicon diode which drives an amplifier, e.g., transistor. The
sensitivity of the silicon material peaks at the wavelength
emitted by the LED, giving maximum signal coupling.
Where the input to the optocoupler is a LED, the input
characteristics will be the same, independent of the type of
detector employed. The LED diode characteristics are shown
in Figure 1. The forward bias current threshold is shown at
approximately 1 volt, and the current increases exponen-
tially, the useful range of I
F
between 1 mA and 100 mA
being delivered at a V
F
between 1.2 and 1.3 volts. The
dynamic values of the forward bias impedance are current
dependent and are shown on the insert graph for R
DF
and
∆
R as defined in the figure. Reverse leakage is in the nano-
ampere range before avalanche breakdown.
The LED equivalent circuit is represented in Figure 2, along
with typical values of the components. The diode equations
are provided if needed for computer modeling and the con-
stants of the equations are given for the IR LED’s. Note that
the junction capacitance is large and increases with applied
forward voltage. An actual plot of this capacitance variation
with applied voltage is shown on the graph of Figure 3. It is
this large capacitance controlled by the driver impedance
which influences the pulse response of the LED. The capaci-
tance must be charged before there is junction current to
create light emission. This effect causes an inherent delay of
10-20 nanoseconds or more between applied current and
light emission in fast pulse conditions.
The LED is used in the forward biased mode. Since the
current increases very rapidly above threshold, the device
should always be driven in a current mode, not voltage
driven. The simplest method of achieving the current drive is
to provide a series current-limiting resistor, as shown in
Figure 4, such that the difference between V
APP
and V
F
is
dropped across the resistor at the desired I
F
, determined from
other criteria. A silicon diode is shown installed inversely
parallel to the LED. This diode is used to protect the reverse
breakdown of the LED and is the simplest method of achiev-
ing this protection. The LED must be protected from exces-
sive power dissipation in the reverse avalanche region. A
small amount of reverse current will not harm the LED, but it
must be guarded against unexpected current surges.
The forward voltage of the LED has a negative temperature
coefficient of 1.05 mV/°C and the variation is shown in
Figure 5.
The brightness of the IR LED slowly decreases in an expo-
nential fashion as a function of forward current (I
F
) and time.
The amount of light degradation is graphed in Figure 6
which is based on experimental data out to 20,000 hours.
A 50% degradation is considered to be the failure point.
This degradation must be considered in the initial design of
optoisolator circuits to allow for the decrease and still remain
within design specifications on the current-transfer-ratio
(CTR) over the design lifetime of the equipment. Also, a
limitation on I
F
drive is shown to extend useful lifetime of
the device.
In some circumstances it is desirable to have a definite
threshold for the LED above the normal 1.1 volts of the
diode V
F
. This threshold adjustment can be obtained by
shunting the LED by a resistor, the value of which is
determined by a ratio between the applied voltage, the
series resistor, and the desired threshold. The circuit of
Figure 7 shows the relationship between these values.
The calculations will determine the resistor values required
for a given I
FT
and V
A
. It is also quite proper to connect
several LED’s in series to share the same I
F
. The V
F
of the
series is the sum of the individual V
F
’s. Zener diodes may
also be used in series.
Where the input applied voltage is reversible or alternating
and it is desired to detect the phase or polarity of the input,
the bipolar input circuit of Figure 8 can be employed. The
individual optocouplers could control different functions or
be paralleled to become polarity independent. Note that in
this connection, the LED’s protect each other in reverse bias.
REV. 4.00 4/30/02
AN-3001
APPLICATION NOTE
V
F
- FORWARD VOLTAGE (VOLTS)
∆R
= 300Ω
1.5
1.4
1.3
1.2
30Ω
3Ω
0.3Ω
I
F
R
S
R
DF
=
13Ω
LED
EQUIVALENT
CIRCUIT
FORWARD BIAS
I
F
mA
100
80
SLOPE
V
=
R
DF
=
I
V
F
R
P
C
j
D
V
j
D - IDEAL DIODE
TA = 25˚C
120Ω
1KΩ
10KΩ
1.1
1.0
0.9
0.8
0.1
1.0
10
60
SLOPE
100
V
F
I
F
C
j
-5
-
55
0
-
-
1
-
10
-
V
I
F
- FORWARD CURRENT (mA)
AVALANCHE
20
V
R
18
16
14
12
10
8
6
4
2
0
40
20
0.5
0.01
1.0
=
∆R
=
∆V
∆I
100 mA
-
1.3
-
0.3
-
pF
V
nA
Ω
Ω
100 300 500
1.0
1.1
-
30
-
1.2
-
3
-
V
F
- V
FT
k
I
F
I
FT
1.5
V
F
VOLTS
V
j
I
R
R
S
R
P
>10
9
<10
0
∞
-
RANGE OF
BV
R
REVERSE BIAS
0.1
THRESHOLD
1.0
10
100
I
F
= I
FT
exp
NOTE CHANGE OF SCALES
mA
I
R
V
F
= V
FT
+ k log
For IRLED (940nm)
V
FTH
= 0.98V
I
FTH
= 0.10mA
K = 0.360
R
S
=
0.03V
I
F
(A)
Figure 1. Characteristics of IR LED
Figure 2. Equivalent Circuit Equations
350
JUNCTION CAPACITANCE (Cj) - pF
300
250
200
I
F
R
150
LED
V
APP
100
V
F
R=
V
APP
- V
F
I
F
NOTE SCALE CHANGE
Figure 4. Typical LED Drive Circuit
50
V
F
1.5
1.0
0.5
0 1 2 3
APPLIED VOLTAGE
V
R
4
5
6
7
8
Figure 3. Voltage Dependence of Junction Capacitance
2
REV. 4.00 4/30/02
APPLICATION NOTE
AN-3001
1.4
I
FT
R
1
R
2
PROVIDES A THRESHOLD
V
V -V
I
FT
=
F
R
1
=
A F
R
2
I
FT
R
2
FORWARD VOLTAGE - V
F
(VOLTS)
1.3
T = -55°C
1.2
V
A
V
F
Figure 7. LED Threshold Adjustment
1.1
T = +25°C
1.0
T = +100°C
I
F
0.9
1
V
F
2
R
1
0.8
0.1
V
A
0.2
0.5
1
2
5
10
20
50 100
FORWARD CURRENT - I
F
(mA)
R
2
V
F
Figure 5. IR Forward Voltage vs. Forward
Current and Temperature
Figure 8. Bipolar Input Selects LED
NORMALIZED CTR DEGRADATION - %
50
40
30
20
I
F
=1
00
R
1
mA
10
7
I
F
=
5m
A
V
A
R
2
V
F
I
F
=
8
6
4
2
60
mA
I
F
= 3
0 mA
Figure 9. High Threshold Bipolar Input
T
A
= 25°C
I
F
= 10 mA
1
10
100
1000
TIME - HOURS
10,000
100,000
Figure 6. Brightness Degredation vs.
Forward Current and Time
I
F
R
1
EXTERNAL
SWITCH
DEVICE
120V
RMS
60 Hz
+
-
C
1
R
3
V
R
2
R
2
+
-
Figure 10. AC Input to LED Drive Circuit
Another method of obtaining a high threshold for high level
noise immunity is shown in Figure 9, where the LED’s are in
inverse series with inverse parallel diodes to conduct the
opposite polarity currents. In this circuit, the V
F
is the total
forward drop of the LED and silicon diode in series. The
resistors serve their normal threshold and current limiting
functions. The silicon diodes could be replaced by LED’s
from other optocouplers or visible signal indicators.
REV. 4.00 4/30/02
3
AN-3001
APPLICATION NOTE
AC Mains Monitoring
In some situations it may be necessary to drive the LED from
a 120 VRMS, 60 Hz or 400 Hz source. Since the LED
responds in nanoseconds, it will follow the AC excursions
faithfully, turning on and off at each zero-crossing of the
input. If a constant output is desired from the optocoupler
detector as in AC to logic coupling, it is necessary to rectify
and filter the input to the LED. The circuit of Figure 10
illustrates a simple filtering scheme to deliver a DC current
to the LED. In some cases the filter could be designed into
the detector side of the optocoupler, allowing the LED to
pulse at line frequency. In the circuit of Figure 10, the value
of C
1
is selected to reduce the variations in the I
F
between
half cycles below the current that is detectable by the detec-
tor portion. This condition usually means that the detector is
functioning in saturation, so that minor variations of I
F
will
not be sensed. The values of R
1
, R
2
and R
3
are adjusted to
optimize the filtering function, R
3
C
1
time constant, etc.
Speed of turn-off may be a determining factor. More compli-
cated transistor filtering may be required, such as that shown
in Figure 11, where a definite time delay, rise time and fall
time can be designed in. In this circuit, C
1
and R
3
serve the
same basic function as in Figure 10. The transistor provides a
high impedance load to the R
4
C
2
filter network, which once
reaching the V
F
value, suddenly turns on the LED and pulls
the transistor quickly into saturation. The turn-off transient
consists of the discharge of C
1
, through R
3
and the LED.
Logic to Logic Interface
In logic-to logic coupling using the optocoupler, a simple
transistor drive circuit can be used as shown in Figure 12.
In the normally-off situation, the LED is energized only
when the transistor is in saturation. The design equations are
given for calculating the value of the series current limiting
resistor. With the transistor off, only minor collector leakage
current will flow through the LED. If this small leakage is
detectable in the optocoupler detector, the leakage can be
bypassed around the LED by the addition of another resistor
in parallel with the LED shown as R
1
. The value of R1 can
be large, calculated so that the leakage current develops less
than threshold V
F
(~0.8 volt) from Figure 5. The drive
transistor can be the normal output current sink of a TTL or
DTL integrated circuit, which will sink 16 mA at 0.2 volt
nominal and up to 50 mA in saturation.
If the logic is not capable of sinking the necessary I
F
, an aux-
iliary drive transistor can be employed to boost current capa-
bility. The circuit of Figure 13 shows how a PNP transistor is
connected as an emitter follower, or common collector, to
obtain current gain. When the output of the gate (G
1
) is low,
Q
1
is turned on and current flows through the LED. The
calculation of R
1
must now include the base-emitter forward
biased voltage drop, V
BE
, as shown in the figure.
+
R
3
I
F
DC
INPUT
FROM
BRIDGE
RECTIFIER
R
4
C
1
10K=R
1
-
C
2
R
2
=10K
V
F
Figure 11. R-C-Transistor Filter Circuit
V
CC
+
I
F
R
V
CC
= 5 V
I
F
= 20 mA
V
SAT
= 0.4 V
V
F
= 1.2 V
R=
V
CC
- V
F
- V
SAT
I
F
R
1
=
V
CC
- V
F
- V
BE
- V
CE(SAT)GATE
I
F
V
CC
V
CC
V
BE(Q1)
= 0.6 V
V
CE(SAT)(G1)
= 0.4 V
R
1
I
F
10K
OUTPUT
R
1
V
F
= 5 - 1.2 - 0.4 = 3.4
20
20
R = 170Ω
INPUT
V
F
(V
SAT)
G
1
V
BE
Q
1
V
CE(SAT)
Figure 12. Transistor Drive, Normally Off
Figure 13. Logic to LED Series Booster
REV. 4.00 4/30/02
4
APPLICATION NOTE
AN-3001
In the normally on situation of Figure 14, the transistor is
required to shunt the I
F
around the LED, with a V
SAT
of less
than threshold V
F
. Typical switching transistors have satura-
tion voltages less than 0.4 volts at I
C
=20 mA or less. The
value of the series resistor is determined to provide the
required I
F
with the transistor off.
Again, if the logic cannot sink the I
F
, a booster transistor
can be employed as shown in Figure 15. With the output of
the gate low, the transistor Q
1
will be on and the sum of
V
CE
(SAT) of G
1
and V
BE
of Q
1
, will be less than the
threshold V
F
of the LED. With the gate high, Q
1
is not
conducting and LED is on. The value of R
1
is calculated
normally, but shunt current will be greater than I
F
. The
normally-on or normally-off conditions are selected depend-
ing on the required function of the detector portion of the
optocoupler and fail-safe operation of the circuits.
In many applications it is found necessary to pulse drive the
LED to values beyond the DC ratings of the device. In these
situations a “pulse” is defined as an on-off transient occur-
ring and ending before thermal equilibrium is established
between the LED, the lead frame, and the ambient. This
equilibrium will normally occur within one millisecond.
For a pulse width in the microsecond range, the I
F
can be
driven above the DC ratings, if the duty cycle is low.
The chart of Figure 16 shows the relationship between the
amount of overdrive, duty cycle, and pulse width. The over-
drive is normalized to the I
DC
value listed as maximum on
the device data sheet. Average power dissipation is the limit-
ing parameter at high duty cycles and short pulse widths. For
longer pulse widths, the equilibrium temperature occurs at
lower duty cycle values, and peak power is the limiting
parameter.
For duty cycles of 1% or less the pulse becomes similar to a
nonrecurrent surge allowing additional ratings such as the I
2
t
used in rectifier diodes. Average current is used for lifetime
calculation. The pulse response of the detector must be con-
sidered in choosing drive conditions.
R=
V
CC
- V
F
I
F
3.8
=
= 190Ω
20
V
CC
R
1
=
V
CC
- V
F
I
F
V
CC
V
CC
I
F
R
R
1
I
F
10K
OUTPUT
INPUT
(V
SAT
)
V
F
G
1
V
BE
Q
1
V
CE(SAT)
Figure 14. Transistor Drive, Normally On
Figure 15. Logic to LED Shunt Booster
100
1 µS
5 µS
10 µS
I
PK
I
DC
PW = 30 µsec
10
100 µS
300 µS
1
0.1
1.0
10
100
DUTY CYCLE - %
Figure 16. Maximum Peak I
F
Pulse Normalized to Max I
DC
for Pulse Width (PW) and Duty Cycle (%)
REV. 4.00 4/30/02
5
京公网安备 11010802033920号
EEWORLD
