Application Note: AN-201 CPC1580 Application Technical Information

Application Note: AN-201
INTEGRATED CIRCUITS DIVISION
CPC1580 Application
Technical Information
AN-201-R01
www.ixysic.com
1
AN-201
INTEGRATED CIRCUITS DIVISION
1
Using the CPC1580 Isolated Gate Driver IC
The CPC1580 is an excellent choice for remote
switching of DC and low frequency loads where
isolated power is unavailable. The device uses external
components to satisfy design switching requirements,
which enables the designer to choose from a great
number of MOSFETs. The designer also has several
options when designing over-voltage protection
circuitry. The case studies look at only two of many
methods, but each has unique constraints that should
prove useful to many other designs.
Figure 1 shows a typical DC application circuit for
using the CPC1580 gate driver. The driver allows the
user to turn on the gate of a MOSFET and keep it on
until the LED current is turned off. The application
circuit uses a bootstrap diode (internal to the part) and
2
2.1
storage capacitor (CST) to provide the charge needed
for fast turn-on switching of an external MOSFET
device. When the MOSFET is on, the photo current
from the LED keeps the MOSFET gate biased to the
device’s specified gate to source voltage (VGS)
continuously.
The CPC1580 uses charge from the load voltage when
turning off to recover the MOSFET gate switching
charge for the next turn-on event. The transistor will
turn on even without this recovery of charge (in the
case of no load voltage), although the turn-on will be
much slower because only internal photo current will
be charging the gate of the MOSFET. This feature can
be exploited during system startup.
Application Component Selection
Storage Capacitor Selection CST
The storage capacitor (CST) enables the part to turn on
quickly by holding a reservoir of charge to be
transferred to the gate of the MOSFET. The turn-off
cycle doesn't depend on the storage capacitor.
CPC1580 can deliver adequate peak current to drive
32nC total gate charge at the rated operating speed,
and will operate with much higher capacitive loads
(<4F), or larger gate charge, with a slower turn-on and
turn-off time.
Note: Care must be taken to minimize any
capacitor-to-ground leakage current path between
pins 7 and 8, MOSFET gate current, and between
pins 5 and 6. Leakage currents will discharge the
storage capacitor, and, even though the device is
already on, will become a load to the photocurrent
which keeps the gate voltage on. The gate voltage
will be reduced if >500nA of leakage is present,
therefore the combined impedance from pin 8 to
pin 7, pin 5 and pin 6, capacitor current, and
MOSFET current must be >20M over the
temperature rating of the part.
Equation 1: Charge Storage Capacitor Calculation:
CST >
QG
(FARADS)
VLOAD - VCAP
QG is the MOSFET’s total gate charge; VCAP > 15V.
Equation 1 shows that the storage capacitor needs to
deliver enough charge to the gate without going below
the 15V required for switching the MOSFET. The
Figure 1
CPC1580 DC Application Circuit Diagram with Over-Voltage Protection
CPC1580
1
4
NC
+VLOAD
When Q1 Off
8 VCAP
ROVP
CST
7
LOAD
VD
NC
+VLOAD
COVP
5
VIN
2
3
VG
6
VS
2
www.ixysic.com
ZOVP
Q1
-VLOAD
R01
AN-201
INTEGRATED CIRCUITS DIVISION
2.2
Transistor Selection
The CPC1580 charges and discharges an external
MOSFET transistor. The selection of the MOSFET is
determined by the user to meet the specific power
requirements for the load. The CPC1580 output
voltage is listed in the specification, but, as mentioned
earlier, there must be little or no gate leakage.
Another parameter that plays a significant role in the
determination of the transistor is the gate drive voltage
available from the part. The CPC1580 uses
photovoltaic cells to collect the optical energy
generated by the LED, and, to generate more voltage,
the photovoltaic diodes are stacked. As such, the
voltage of the photovoltaic stack reduces with
increased temperature. The user must select a
transistor that will maintain the load current at the
maximum temperature, given the VGS in the CPC1580
specification.
The case studies below use "logic-level" MOSFETs for
each design to maintain the load described.
2.2.1
Transistor Switching Characteristics
The primary characteristics of the application switching
are tON, tOFF, tRISE, tFALL, and the recovery time of the
storage capacitor, tCHG. These parameters are
dependent on the MOSFET selection and need to be
reviewed in light of the application requirements.
The CPC1580 turns on the MOSFET to the datasheet
VGS after the tON delay. Similarly the tOFF delay is the
amount of time until the LED is turned off and the
capacitive load discharges to the level in the CPC1580
specification. For MOSFETs with larger or smaller
required gate charge the tON and tOFF will be
proportionately faster and slower, but it is not a linear
relationship.
To calculate the nominal rise and fall times of the
MOSFET's drain voltage:
Equation 2: Rise Time Calculation
tRISE,VD
~
VLOAD • CRSS
(SECONDS)
IG_SINK
Equation 3: Fall Time Calculation
tFALL,VD
R01
~
VLOAD • CRSS
IG_SOURCE
(SECONDS)
Where CRSS is the MOSFET gate-drain capacitance
(averaged over the switching voltage range) found in
the MOSFET datasheet, IG_SINK is the gate sinking
current of the CPC1580, and IG_SOURCE is the gate
driving ability. The maximum value of tRISE is limited by
the CPC1580 unloaded discharge characteristic, and
should be reviewed in light of the final application
component selections, if critical.
To calculate the value for the charge time, tCHG, which
is due to external component selection:
Equation 4: Storage Capacitor Charge Recovery Time
(seconds):
tCHG
(
~ - (400 + ROVP) • (CST + COVP) • ln (VLOAD - VFINAL) • CST
QG
)
where (VLOAD -VFINAL) is the difference in voltage
between the required load voltage and the potential the
capacitor will charge up to. The voltage at the storage
capacitor is VLOAD - (QG/CST) when the MOSFET is on,
where charge, QG, is the amount of charge required to
switch the MOSFET gate from 0V to the final voltage
out of the CPC1580 (VGS specification). VFINAL is the
capacitor voltage when it charges back up from when
the MOSFET is off.
ROVP and COVP form the overvoltage protection RC
filter. The RC filter is used to reduce the peak power
dissipation in the MOSFET by controlling the rate of
rise of the drain voltage. Note that the RC circuit will
reduce the switching speed of the MOSFET.
Note: Obviously, the logarithm doesn't work if
VFINAL = VLOAD because of the exponential nature
of R-C charging. That subsequently affects the
next cycle, so CST is more critical and should be
larger if the switching frequency is faster. Selecting
the term inside the logarithm to be 0.05 yields 3
equivalent time-constants.
Using this information, the maximum switching
frequency will be calculated in each application case
study below.
Note:The CPC1580 is ideal to use where remote
power is otherwise unavailable. If the LED is also
powered remotely, care must be taken to ensure
that parasitic transient signals are reliably filtered
from the input control signal. Large transient
currents will mutually couple energy between
cables and a simple R-C filtering of the CPC1580
input may be sufficient to suppress false turn-on.
www.ixysic.com
3
AN-201
INTEGRATED CIRCUITS DIVISION
3
Application Switching Losses
During the transition intervals, the application and load
components change energy states and during the
process incur switching losses. These losses are
manifested as heat in the application circuit, and must
be addressed by the designer to ensure that no one
component exceeds its power rating. The designer
must understand the details of load behavior in order to
adequately size and protect the application circuit.
There are three general cases to observe: (1) purely
resistive loads, (2) inductive/resistive loads, and (3)
loads with significant capacitance. Inductors and
capacitors are energy storage elements that require
special consideration for switching.
During switching periods, the energy stored in the load
inductor is discharged through the switching MOSFET,
load capacitance and the over-voltage-protection
circuitry.
3.2
Inductive/Resistive Loads
If the load is resistive and inductive, and the inductance
doesn't saturate, then the load current during turn-off is
described by:
Equation 8: Resistive/Inductive Load Current during
tRISE (Amps)
ILOAD(t) =
VLOAD
RLOAD
-
IG_SINK
LLOAD • CRSS
•
( )
2
LLOAD
RLOAD
•
[
RLOAD
LLOAD
-R LOAD
•t
LLOAD
• t-1+e
]
The drain voltage during turn-off is:
Equation 9: MOSFET Drain Voltage during tRISE (V)
VDRAIN(t) =
IG_SINK
CRSS
• t
At turn-on, the inductor energy is zero, and so the
capacitive energy in the load and parasitic elements of
the switching application must be dissipated by the
MOSFET in order for the load to change state.
The instantaneous power in the MOSFET will be the
product of the two equations and the energy will be the
integral of the power over time.
Equation 5: Stored Inductive Energy (Joules)
3.3
EL =
3.1
1
2
• L • ILOAD2
Resistive Load Losses: The Ideal Case
For purely resistive loads, the energy dissipated by
changing states occurs primarily in the MOSFET. The
equation describing MOSFET energy dissipation is:
The energy absorbed by the MOSFET for loads that
are more capacitive in nature occurs during the
MOSFET turn-on as opposed to the turn-off. The
energy absorbed by the MOSFET will be a function of
the load, the Transient Voltage Suppressor (TVS) or
other protector, and the MOSFET drain capacitance.
Equation 10: MOSFET Energy: EFALL (Joules)
Equation 6: MOSFET Energy: ERISE (Joules)
EMOSFET > VLOAD2 •
CRSS
IG_SINK
•
ILOAD
6
=
PLOAD
6
EFALL =
• tRISE,VD
The average power of the MOSFET for any load type
is:
Equation 7: MOSFET Average Power (Watts)
PAVG = ILOAD2 • RDSAT • D + fSWITCH • (ERISE + EFALL)
Where fSWITCH is the application switching frequency,
RDSAT is the MOSFET’s on-resistance, D is the switch's
operational duty cycle: D = tON/(tON+tOFF). ERISE and
EFALL are the energy dissipated during the rise and fall
times.
4
Capacitive Loads
1
2
• (CTVS + COSS + CLOAD) • VLOAD2
COSS is the MOSFET output capacitance found in the
datasheet. As mentioned earlier, the MOSFET
switching losses occur at different times, either rising or
falling, so loads with a combination of inductance and
capacitance can also be calculated by the energy
equations described above.
The MOSFET can dissipate the repeated avalanche
energy, (EAR), as specified in the datasheet. However
that energy must be reduced for increased ambient
temperature. For a 150°C MOSFET, the energy
reduction at TJ,MAX is:
www.ixysic.com
R01
AN-201
INTEGRATED CIRCUITS DIVISION
Equation 11: MOSFET Energy Adjustment for
Operating conditions (Joules):
E(TJ,MAX) < E(25°C) •
(150°C - TJ,MAX)
(150°C - 25°C)
TJ,MAX is the junction temperature of the die, so it must
include the temperature increase caused by power
dissipation of the load and the thermal impedance of
the package/application. E(25°C) is the EAR
specification found in the MOSFET datasheet at 25°C.
3.4
dV/dt Characteristics
The application shown in Figure 1 and described in
section 6.1 “Case 1: 24 V Loading Application”
dissipates significant energy caused by large dV/dt
events. Fault voltages across the MOSFET will turn it
on for the same reason the part turns off slowly.
4
For dV/dt events > IG_SINK/CRSS (from Equation 2) the
application circuit will dissipate energy proportional to
the CRSS and gFS (forward conductance) of the
selected transistor. CRSS is a function of the transistor's
on-resistance and current/power capability, so higher
load designs are more sensitive.
The CPC1580 provides an internal clamp to protect the
gate of the MOSFET from damage during such an
event. It can withstand 100mA for short periods, for
instance, during dV/dt transients.
Note:The CPC1580 is ideal to use where remote
power is otherwise unavailable. If the LED is also
powered remotely, care must be taken to ensure
that parasitic transient signals are reliably filtered
from the input control signal. Large transient
currents will mutually couple energy between
cables and a simple R-C filtering of the CPC1580
input may be sufficient to suppress false turn-on.
Design Switching Frequency
The over-voltage protection and storage capacitor play
a significant role in determining the switching
frequency. The maximum switching frequency is a
function of the gate charge of the MOSFET, the storage
capacitor (CST), and ROVP. The maximum switching
frequency relationship is:
Equation 12: Maximum Switch Operation (Hz)
FMAX <
5
1
-1
• (tON + tOFF + tRISE,VD | tCHG + tFALL,VD)
M
There is no minimum switching frequency because the
CPC1580 uses photovoltaic diode current to keep the
output charged as long as LED current flows.
CPC1580 Over-Voltage Protection
Over-voltage protection is generally required for the
CPC1580 because of parasitic inductance in the load,
wires, board traces, and axial leads of protectors.
Purely resistive loads or loads with low voltage
switching may be able to rely on the transistor to
handle any parasitic energy and thereby not require
protection for the CPC1580. For very low-inductance
loads and traces, over-voltage suppression may be
handled with a simple RC filter consisting of ROVP and
COVP, or by use of a free-wheeling diode. For more
moderate load inductance, or remote switching of a
load (i.e. through a long cable) a voltage suppressor
can be used. For heavily inductive loads only a freewheeling diode, DOVP, connected across the load
element is recommended, see Figure 2.
R01
where M=3 (a multiplication factor for temperature and
process variations); tON and tOFF are CPC1580
datasheet parameters; tRISE,VD is the rise time of the
drain voltage; tCHG is the charge time of the storage
capacitor (CST) and over-voltage protection circuitry
(COVP and ROVP); and tFALL,VD is the fall time across
the transistor. For this calculation, choose the greater
of tRISE,VD or tCHG.
The energy not consumed in switching losses must be
absorbed by the over-voltage protection element. Most
protective devices are designed to withstand certain
peak power in the case of a TVS, or maximum
avalanche energy in the case of a MOSFET. To reduce
the amount of stored inductive energy, a larger
capacitor can be added in parallel with the gate-drain
connection of the MOSFET. However care must be
taken so that the rise time and peak current do not
exceed the Safe Operating Area (SOA) rating of the
transistor. A consequence of increasing the gate-drain
effective capacitance is reduced dV/dt tolerance.
www.ixysic.com
5
AN-201
INTEGRATED CIRCUITS DIVISION
Figure 2
CPC1580 Over-Voltage Protection for Inductive Loads
CPC1580
RLED
VIN
1
8
2
7
3
6
DOVP
CST
ROVP
ZLOAD
V+
COVP
VLOAD
4
5
Q1
V-
Other Protection Techniques 1, 2
5.1
For applications in which higher inductance loads are
switched, the designer must consider other circuit
techniques, device ratings, or protector types. Of
paramount importance is that the designer know the
characteristics of the load being switched.
1
An excellent source describing power electronic devices
and switching behavior is: Power Semiconductor
Devices, by B. Jayant Baliga, ISBN 0-543-94098-6
2
For more over-voltage protection circuit techniques
consult: Switchmode Power Supply Handbook, 2nd
Edition, Keith Billings, ISBN 0-07-006719-8, or Power
MOSFET Design, B. E. Taylor, ISBN 0-471-93802-5.
6
www.ixysic.com
R01
AN-201
INTEGRATED CIRCUITS DIVISION
6
Design Examples
Table 1: shows two sample application component selections each with different over-voltage protection strategies.
Table 1:
Sample Application Components
Device
Q1
CST
Case 1: 24V/10A Value/Rating
SUD45N05-20L
>0.01F/100V
ZOVP
3
Case 2: 48V/5A Value/Rating
SUD23N06-31L
>0.01F/100V
3
5% Ceramic Disk
Littelfuse TVS-style protector
3
ROVP
SA24A
1K
COVP
0.001F, 50V
0.001F, 100V
5%, 1/8 Watt (60Hz Switching
Frequency or less)
5% Ceramic Disk
RLED
680, 1/8 Watt
680, 1/8 Watt
0 - 5V Switching
3
6.1
SA48A
5.1K
Comment
MOSFET
3
Use of the SUD45N05-20L, SUD23N06-31L, SA24A, and SA48A product datasheets is necessary to completely understand the
examples given.
Case 1: 24V Load Switching
In this example, the over-voltage protection circuitry is
quite simple. The CPC1580 is guaranteed for 60V
operation and the protector is rated for 45.4V @ 11.2A
peak pulse current, well below the 60V. The transistor
(Q1) is a 50V MOSFET, which guarantees the TVS
clamps before the transistor breakdown. Assuming
there will be load inductance in both the VLOAD+ and
Figure 3
VLOAD- traces, a TVS is selected to clamp the residual
10A not otherwise dissipated in the turn-off of the
MOSFET and parasitic TVS capacitance. ROVP and
COVP are optional for this load condition; however, their
inclusion will ease layout and critical placement of the
CPC1580.
Case 1 Application Circuit
CPC1580
RLED
VIN
1
CST
8
0.01μF/100V
2
ROVP
LOAD
7
680Ω
1kΩ
3
4
COVP
6
SA24A
0.001μF/50V
5
+VLOAD
ZOVP
SUD45N05-20L
-VLOAD
For this test case, the maximum switching frequency
for the design is FMAX = 0.333 (40s + 600s + (40s
| 42s) + 0.87s)-1 < ~475Hz. The components
selected were used for in-lab testing. Other
components with smaller package sizes and wattage
will also work, if calculations are performed to meet
component specifications.
Example:
• RLED=680
• Minimum voltage drop across the LED is 1.0V
R01
• Switching voltage, SwVON, when on, is 5V
• IF=Forward current of the LED
SwVON - Min LED Volt
RLED
5V - 1V
IF =
680
IF =
IF = 0.005882A = 5.9mA
The recommended IF is between 2mA and 10mA. The
IF calculated above meets this requirement.
www.ixysic.com
7
AN-201
INTEGRATED CIRCUITS DIVISION
The power dissipated, PD, is:
(12A)2 0.02+ 475/s 3.3mJ = 4.5 Watts, assuming
very high operational duty cycles.
PD=IF2 • R
PD = (5.9mA)2 • (680)
PD = 0.024W = 24mW
These calculations show that a 0603 resistor, which is
1/16 Watt, can be selected. The 1/16 Watt still provides
an adequate design margin: 0.0625W where only
0.024W is required.
6.1.1
Measured Results
Figure 4 shows the discharge of the storage capacitor
due to the gate switching on. The calculated voltage
drop (VLOAD - VCAP) using CST = 10 nF and (QG = 43nC
from the Q1 datasheet) from Equation 1 is 4.3 Volts.
From Equation 1: Charge Storage Capacitor
Calculation:
QG
CST >
VLOAD - VCAP
VLOAD • CRSS
~
tRISE ~ (24V-5V) 190pF/.0036 A ~ 1S
From Equation 3: Fall Time Calculation
tFALL,VD
~
IG_SOURCE
4
(SECONDS)
tFALL ~ (24V-5V) 190pF/0.00022 A ~ 16S
All other calculated / measured data is summarized in
Table 2:
Table 2:
The power absorbed by the TVS can be calculated
from the characteristic of the waveform shown in
Figure 10:
Energy = ½ L I2 = [(VTVS-VLOAD) tDSCHG]2/(2 L)
The example listed demonstrates the need to have an
accurately characterized load so that the energy due to
the switching event does not exceed the rating of the
MOSFET or TVS protector.
(SECONDS)
IG_SINK
VLOAD • CRSS
Again this assumes that the magnetics do not saturate,
however for the graphs shown in Figure 10 and
Figure 11, the current equation above only applies
after the magnetic flux leaves saturation and becomes
inductive again. As such, the load current is dominated
by VLOAD and RLOAD in Figure 10.
which is ½ 800H (0.45A)2 = 81J. This current
(0.45A) agrees well with the turn-off characteristic
shown in the graph where the magnetics leave
saturation at ~0.5A.
(FARADS)
From Equation 2: Rise Time Calculation
tRISE,VD
This circuit load was modified to include an 800H
inductor that saturates at ~0.5A. This load condition
may not represent the user's load but does serve to
illuminate more about the switching characteristics of a
non-linear load.
24 Volt Load Switching Data
Parameter
Calculated
Measured
Voltage Drop CST
4.3V
3.7V
tFALL Figure 5
16S
2S
tRISE Figure 8
1S
tON Figure 6
16S (1580 spec)
38S 4
7.3S
tOFF Figure 7
175S (1580 spec)
189S
The calculated rise time relies on the manufacturer
supplied graphs for CRSS. The actual rise time during
the interval shown in Figure 8 is longer due to the
non-linear nature of the capacitance CRSS. From the
datasheet graphs, the average capacitance is 190pF
over the interval of 5V<VDS<25V. During the initial
turn-off the capacitance is much larger, affecting the
total energy by ~30%. A second-order effect not used
in Equation 2 is due to the gate-source capacitance
CISS. That additional capacitance divided by the
transistor’s conductance and load resistance causes an
additional delay of 5s-10s, so the calculated rise time
is closer to 35s.
The energy in Figure 9 rises to 3.3mJ, and the
switching frequency can be as high as >475Hz which
would make the average power
8
www.ixysic.com
R01
AN-201
INTEGRATED CIRCUITS DIVISION
Discharge from Gate Turning On
Figure 7
Turn-Off Delay
CPC1580 Turn-Off Characteristics
CPC1580 Capacitive Discharge
9
12
20
VIN and VDRAIN (V)
8
VCSTORAGE
15
10
5
VDRAIN
0
-5
-50
0
50
IDRAIN
8
6
5
6
VGATE
4
2
0
-50
50
Time (s)
Time (
Figure 5
Figure 8
Load Current and tFALL
10
8
15
6
10
4
5
0
-5
0
5
MOSFET Voltage (V)
IDRAIN
14
VDRAIN
25
MOSFET Current (A)
MOSFET Voltage (V)
VDRAIN
20
Load Current and tRISE
IDRAIN
12
25
-2
250
150
30
14
30
4
VIN
3
2
1
0
-150
100
10
7
10
8
15
6
10
5
0
0
10
12
20
2
MOSFET Current (A)
VDROP = QG / CST = 3.7V
25
Voltage (V)
14
10
30
4
MOSFET Current (A)
Figure 4
2
0
-100
-50
0
50
Time (s)
Turn-On Delay
Figure 9
CPC1580 Turn-On Characteristics
6
10
4
5
2
R01
3.0
DMOS Power
70
2.5
60
2.0
50
40
1.5
VDRAIN
30
20
1.0
IDRAIN
0.5
10
VIN
0
-5
Amps, Volts, Watts
15
MOSFET Current (A)
8
3.5
Energy
80
10
20
0
90
I DRAIN
VDRAIN
25
Voltage (V)
CPC1580 Switching Losses
12
30
Discharge Power and Energy
0
5
10
Time (s)
15
20
25
0
MOSFET Energy (mJ)
Figure 6
0
-100
-50
0
50
Time (s)
www.ixysic.com
9
AN-201
INTEGRATED CIRCUITS DIVISION
Figure 10 Moderate Inductive Current and tRISE
25
Inductive
Energy in TVS
10
8
6
15
4
10
Inductor leaves saturation
5
2
0
0
-5
-100
-50
0
50
Time (s)
100
150
20
10
9
8
7
6
5
4
3
MOSFET Voltage (V)
25
IDRAIN
20
15
10
5
Magnetics saturate
0
-10
0
10
1.5
10
1.0
5
0.5
0
0
-0.5
-10
0
10
Time (s)
20
30
20
Time (s)
30
CPC1580 Turn-Off Characteristics
100 H / 10 Load
ILOAD Meas
2.5
VDRAIN
2.0
25
ILOAD, Eq. 8
15
1.0
10
0.5
5
0
0
40
As seen in the turn-on characteristic is almost perfectly
inductive where the di/dt forms a non-saturating V/L
curve. The voltage applied remains at 24V.
Figure 13 shows the inductive nature of the turn-off as
seen in the overshoot. In this case Equation 8 was fit
to the time-base and the resistance, inductance, and
capacitance were plugged in. The slope of the line is
steeper than expected, which is what has been
observed in the previous example. Equation 8 was
then modified to include the CISS factor
1.5
20
2
1
The load was modified to avoid saturating the
magnetics allowing comparison of the expected load
current (from Equation 8) versus the measured load
current. The circuit changes were to increase the
resistance to 10.2 Ohms and change the magnetic
inductance to 113H.
10
2.0
100 H Slope
30
MOSFET Current (A)
CPC1580 Turn-On Characteristics
w / 800 H Inductance
VDRAIN
ILOAD
Figure 13 Turn-Off with Modified Load
Figure 11 Inductive Turn-On
30
2.5
15
-5
-2
200
VDRAIN
Load Current (A)
20
12
MOSFET Voltage (V)
VDRAIN
25
MOSFET Voltage (V)
30
MOSFET Voltage (V)
14
IDRAIN
MOSFET Current (A)
35
CPC1580 Turn-On Characteristics
100 H / 10 Load
Load Current (A)
CPC1580 Switch Turn-Off
w / 800 H Inductance
Figure 12 Turn-On with Modified Load
0
ILOAD, Eq. 8
150
170
190
210
Time (s)
230
-0.5
250
(CRSS + CISS/(gFS RLOAD)) and the resultant slope
better approximates the actual slope as expected. It is
worth restating that the slow change at the beginning of
the transition is due to the large non-linearity in
capacitance vs. voltage. While this interval is an
important component of the total energy (~30%) the
calculation is more complicated and not readily
available from the component datasheets. Analysis
described in the references listed will improve the
characteristic to within 10%.
Equation 8 proves to be an accurate model for load
current during the turn-off time, which can be
subsequently used to consume inductive energy during
the turn-off event. The equation can include
second-order terms to more accurately model the
transition region of switching.
www.ixysic.com
R01
AN-201
INTEGRATED CIRCUITS DIVISION
6.2
Case 2: 48V Load Switching
Voltages closer to the peak operating voltage of the
CPC1580 can also be accommodated, but the overvoltage protection becomes more important. Table 1:
shows a sample over-voltage protection component
selection for a 48V/5A design requirement.
The design criteria are more complicated because the
peak voltage at 5A for the TVS component is 77V
which exceeds the voltage rating for the CPC1580 and
MOSFET of 60 volts maximum. Two conditions must be
met for using such a protector: (1) protecting the
CPC1580 from going above it's maximum voltage, and
(2) ensuring the avalanche energy of the MOSFET is
not exceeded. Since the MOSFET breakdown voltage
will be nominally higher than the specification, (or if the
user selects a higher voltage MOSFET), then COVP
should be replaced with a zener diode/TVS to keep the
voltage at pin 7 (VD) to less than 60V but greater than
48V. (Until the parasitic inductance discharges to 1mA
at which the TVS voltage is 59V.)
Figure 14 Case 2 Application Circuit
CPC1580
RLED
VIN
1
CST
8
ROVP
0.01μF/100V
2
680Ω
5.1kΩ
3
4
+VLOAD
LOAD
7
ZOVP
COVP
6
SA48A
0.001μF/100V
5
SUD23N06-31L
-VLOAD
Measured Results
Figure 18 and Figure 19 demonstrate the response
with the inclusion of the inductive load. For the case
shown, the MOSFET energy dissipation exceeds the
stored inductive energy of 160J, so no energy is
transferred to the TVS.
The charge time plays a significant role in the
calculation of the maximum switching frequency for this
case study. However, the charging voltage is very small
so the resulting charge time can be reduced, knowing
R01
that the voltage dropped across ROVP will increase
proportionally. The maximum switching frequency of
the example in Table 1: is FMAX = 0.333 (40s +
600s + (34s | 181s) + 2s)-1 < ~400Hz.
Figure 15 48V Case Study tFALL
50
45
40
35
30
25
20
15
10
5
0
CPC1580 - Case 2 Switch Turn-On
w / Resistive Load (10.2 )
5
IDRAIN
VDRAIN
4
3
tFALL = 2.28 s
2
MOSFET Current (A)
The design for Case 2 was implemented and the
following characteristics observed. Figure 15 shows
the fall time for a resistive load. The calculated fall time
is ~1S. The rise time is shown in Figure 16. The
calculated value is 34S in the linear region shown on
the graph. The peak energy during the transient is
shown in Figure 17. The calculated Peak Energy, from
Equation 6 is 1.36mJ. This value is consistent with the
linear-region switching losses. The additional energy
dissipation is due to the large non-linear capacitance at
the beginning of the transition.
MOSFET Voltage (V)
6.2.1
1
-5
www.ixysic.com
0
5
10
0
Time (s)
11
AN-201
INTEGRATED CIRCUITS DIVISION
Figure 16 48V Case Study tRISE
CPC1580 - Case 2 Switch Turn-Off
w / Resistive Load (10.2 )
30
25
20
3.0
2.5
2.0
tRISE = 38.4 s
15
10
5
0
-50
0
1.5
1.0
0.5
0
100
50
Time (s)
Figure 18 Inductive Turn-On Transition
10
5
0
-10
2.0
ILOAD
1.5
~ 113 H
1.0
0.5
0
10
20
0
MOSFET Voltage (V)
2.5
VDRAIN
25
20
15
Energy
1.2
1.0
0.8
0.6
0.4
0.2
-40
-20
0
20
Time (s)
40
60
80
CPC1580 - Case 2 Turn-Off Characteristics
w / Unsaturated Inductive Load
Load Current (A)
MOSFET Voltage (V)
40
35
30
1.4
VDRAIN
100
0
Figure 19 Inductive Turn-Off Transition
CPC1580 - Case 2 Turn-On Characteristics
w / Modified Inductive Load (113 H / 25 )
50
45
1.6
DMOS Power
MOSFET Energy (mJ)
VDRAIN
50
45
40
35
30
25
20
15
10
IDRAIN
5
0
-100 -80 -60
50
45
40
ILOAD
35
30
25
20
15
10
5
0
-100 -80 -60
Time (s)
VDRAIN
2.5
2.0
1.5
1.0
Load Current (A)
IDRAIN
4.5
4.0
3.5
Amps, Volts, Watts
45
40
35
CPC1580 Switching Losses
5.0
MOSFET Current (A)
50
MOSFET Voltage (V)
Figure 17 48V Case Study Peak Power and Energy
0.5
-40
-20
0
20
Time (s)
40
60
80
0
100
For additional information please visit our website at: www.ixysic.com
IXYS Integrated Circuits Division makes no representations or warranties with respect to the accuracy or completeness of the contents of this publication and reserves the right to make
changes to specifications and product descriptions at any time without notice. Neither circuit patent licenses nor indemnity are expressed or implied. Except as set forth in IXYS Integrated
Circuits Division’s Standard Terms and Conditions of Sale, IXYS Integrated Circuits Division assumes no liability whatsoever, and disclaims any express or implied warranty, relating to its
products including, but not limited to, the implied warranty of merchantability, fitness for a particular purpose, or infringement of any intellectual property right.
The products described in this document are not designed, intended, authorized or warranted for use as components in systems intended for surgical implant into the body, or in other
applications intended to support or sustain life, or where malfunction of IXYS Integrated Circuits Division’s product may result in direct physical harm, injury, or death to a person or severe
property or environmental damage. IXYS Integrated Circuits Division reserves the right to discontinue or make changes to its products at any time without notice.
Specification: AN-201-R01
©Copyright 2014, IXYS Integrated Circuits Division
All rights reserved. Printed in USA.
8/1/2014
12
www.ixysic.com
R01