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 (<4F), 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 >0.01F/100V ZOVP 3 Case 2: 48V/5A Value/Rating SUD23N06-31L >0.01F/100V 3 5% Ceramic Disk Littelfuse TVS-style protector 3 ROVP SA24A 1K COVP 0.001F, 50V 0.001F, 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 (40s + 600s + (40s | 42s) + 0.87s)-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 ~ 1S From Equation 3: Fall Time Calculation tFALL,VD ~ IG_SOURCE 4 (SECONDS) tFALL ~ (24V-5V) 190pF/0.00022 A ~ 16S 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 800H (0.45A)2 = 81J. 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 800H 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 16S 2S tRISE Figure 8 1S tON Figure 6 16S (1580 spec) 38S 4 7.3S tOFF Figure 7 175S (1580 spec) 189S 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 5s-10s, so the calculated rise time is closer to 35s. 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 113H. 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 160J, 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 (40s + 600s + (34s | 181s) + 2s)-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 ~1S. The rise time is shown in Figure 16. The calculated value is 34S 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