Application Note AN-1215

Transcription

1 Application Note Gen2 SIP1A IRAM By Jonah Chen, Pengwei Sun and Anna Grishina Table of Contents 1. Introduction Introductions IRAM Design Concept and Technology 2 2 IRAM Gen2 SIP1A Product Outline Part Number Convention Product Line-Up Package Structure 6 3 Package and Pin Description Outline Drawings Module Pin-out Description 9 4 Internal Circuit and Features UVLO Over Current Protection Fault Output and Auto Clear Function Over Temperature Protection 13 5 Absolute Maximum Ratings 15 6 Bootstrap Circuit Bootstrap Circuit Operation Bootstrap Capacitor Selection Bootstrap Circuit Initial Charging and Bootstrap Diode Recommended Bootstrap Capacitor Value 19 7 Interface circuit General Interface Circuit Example 21 8 Power Loss and Junction Temperature Calculation Electrical Model Thermal Model Electrical and Thermal Calculation IPM Design Tool Functions Design Example 25 9 Packing 28 1

2 1. Introduction 1.1. Introductions With the global emphasis on energy efficiency, there is ever stricter requirement on the efficiency of motor drive circuit. Integrated Power Modules (IPMs) are becoming more popular in the home appliance and industrial motor drive applications, because of higher efficiency, smaller size, easier assembly and shorter development time. Our next generation of SIP1A IRAM is developed with the focus on improving the module efficiency and long term reliability. The combined benefits of advanced Trench IGBT technology and optimized package design have enabled us to achieve higher efficiency and improved reliability, along with minimized module and system cost. The Trench IGBT is able to deliver up to 30% loss reduction compared with NPT IGBT of same die size. In addition, the new IRAM has achieved as much as 50% reduction in IGBT junction temperature ripple, thanks to the superior thermal structure of new Gen2 SIP1A package. This advanced hybrid module is a combination of IR's low V CEON Trench IGBT technology and the industry benchmark 3 phase high voltage, high speed gate driver in a fully isolated thermally enhanced package. A built-in high precision temperature monitor and over-current protection feature, along with the short-circuit rated IGBTs and integrated under-voltage lockout function, deliver high level of protection and fail-safe operation. Using a Single in line package with full transfer mold structure and CTI>600V molding compound minimizes PCB space and resolves isolation problems to heat sink IRAM Design Concept and Technology Trench IGBT Trench IGBTs offer significant improvement in terms of loss reduction, over the last generation of Non-Punch-Through (NPT) IGBTs. For example, Figure 1.1 shows the comparison of V CEON vs. I CE for NPT IRGB8B60K, Trench IRGB4056D and IRGB4060D IGBTs. While the first two IGBTs have the same die size, the last one is about 20% smaller. It is quite clear that the conduction losses can be reduced as much as 30%, for the same die size. Even with the smaller die, it is still possible to achieve 10% loss reduction. Since we have optimized the switching characteristics to be quite similar between Trench and NPT IGBTs, switching loss will largely remain unchanged. As we know, the current rating of IPM are fundamentally determined by the IGBT power losses (P LOSS ) and IGBT junction to case thermal resistance (R THJC ), as showing in the equation below. ΔT JC = P LOSS * R THJC IGBT power loss is a function of motor current and other parameters such as switching frequency. And Junction to case thermal resistance is mainly decided by the IGBT die size, assuming we are using the same module package. The junction to case temperature delta (ΔT JC ) is usually set at 50ºC which is derived from maximum junction temperature (T JMAX ) of 150ºC and maximum case temperature (T CMAX ) of 100ºC. 2

3 25 20 Ice [A] IRGB4056D IRGB4060D IRGB8B60K Figure Vceon [V] at Vge=15V and Tj=150C IGBT V CEON vs. I CE curve of IRGB8B60K, IRGB4060D and IRGB4056D The IGBT technology advancement brings two potential opportunities for the new IPM development. On one side, we can keep using the same size of IGBT die. In this case, the R THJC will remain same, while P LOSS at the same current will become smaller. Therefore, we can increase the current rating while still maintaining the ΔT JC 50ºC. For example, it is feasible to develop modules with current rating of 20A, instead of 15A, with same module package. Therefore, the appliance manufacturer will be able to expand the power range of their motion control board, without pursuing bigger sized modules. On the other side, we can use smaller IGBT die if we want to create modules with same current rating. For example, as shown in Figure 1.1, it is now possible to replace IGB8B60K with IRGB4060D which is about 20% smaller and achieve lower module cost. The R THJC will be bigger in the new module. However, it will be compensated by smaller power losses of Trench IGBT. In the end, we can still meet the requirement of ΔT JC 50ºC. As an additional benefit, the new IPM can use smaller heat sink which also brings down the system cost. Thermal Design The smaller and thinner IGBT die provides a new challenge in the thermal design. Because of its small thermal mass, the IGBT junction temperature tends to swing a lot, especially at low speed operation. In order to improve the transient thermal performance and reduced the junction temperature ripple, we have added copper heat spreaders (HS) with 1mm thickness underneath the all six IGBTs and six diodes. Figure 1.2 shows the junction to case thermal impedance (Z THJC ) curve of IGBT. The red solid curve is the Z THJC curve of module without heat spreader, and the blue dashed curve is for the module with HS. There is a slight difference in the R THJC value (when time is infinite). The reason is that while the added heat spreader constitutes one additional layer in the heat transfer path, it also helps to spread the heat across its bottom surface due to copper s excellent thermal conductivity. Therefore, the layers beneath the heat spreader will have larger effective area for the heat transfer. 3

4 Zth [C/W] Module w ithout HS Module w ith HS Time [s] Figure 1.2 IGBT junction to case thermal impedance for modules with and without heat spreader The big difference lies in the time range from 0.01s to 1s. It can be seen clearly that heat spreader has helped to achieve much lower thermal impedance in this range. Especially at 0.1s time range, which corresponds to low speed module operation condition of f MOD =3Hz, the thermal impedance is reduced by almost 50%. The measurement of IGBT junction temperature shown in Figure 1.3 also verified this advantage Module w ithout HS Module w ith HS LS IGBT Tj [C] Time [s] Figure 1.3 IGBT T J measurement for modules with and without heat spreader While the benefit of heat spreader is significant in lower speed operation, it is less critical when the motor is running at high speed. For example at f MOD =50Hz, the IGBT temperature is mainly determined by the R THJC value, plus the smaller T J ripple determined by Z THJC at 5ms range. It is quite evident from Figure 1.2 that the difference of thermal impedances is quite small at this time range. In Gen2 SIP1A IRAM, we further improved the thermal performance by eliminating the over molded plastics layer found at the back of Gen1 SIP1A modules. The new module has exposed metal backside. Therefore the R THJC of both IGBT and diode are reduced by 15%. Package Design Packaging options include staggered pin-out for maximum creepage distances Both straight or 90 bend options for heat sinks parallel or perpendicular to the circuit board 4

5 Insulated Metal Substrate technology ensures low thermal resistance and less stringent flatness requirements for the heat sink. It also offers significant flexibility in the module layout and internal electrical system. Higher operating case temperature (T CMAX =125 C) compared with 100 C T CMAX for Gen1 SIP1A IRAM. This enables customers to use even smaller heat sink to minimize system cost. Pin to Pin compatible to previous Gen1 SIP1A IRAM by keeping same functionality for easy adoption. Exposed IMS substrate improving the thermal performance. Isolation 2000V RMS min Molding compound with CTI>600V Recognized by UL (File Number: E252584), with T JMAX of 150 C. 5

6 2. IRAM Gen2 SIP1A Product Outline 2.1. Part Number Convention I R A M A 2 Lead Forming (Omit if not used) blank Straight Leads 2 90 Bend Power Stage Topology A Open Emitter Package Code 7 Gen2 SIP1A Voltage Code 6 600V Current Rating 10 PWM Frequency & Tc=25 C Essential Code 256 Three phase inverter IPM Family A Appliance & Industrial (IR) Figure 2.4 Gen2 IRAM Part Number System 2.2. Product Line-Up The Table 2.1 below shows the modules that have been released to production. Additional modules are under development which will extend the current to wider ranges. Part Number Table 2.1 Current (A) Rating Voltage (V) Gen2 SIP1A IRAM Product Line Package Isolation Voltage (V RMS ) Typical Load (W) I Tc = 100ºC (A RMS ) IRAM A(2) W 5 IRAM A(2) Gen2 SIP1A W 7.5 IRAM A(2) W Package Structure Figure 2.2 and Figure 2.3 show the basic structure of IRAM Gen2 SIP1A. It features an exposed Insulated Metal Substrate (IMS) technology which provides enhanced thermal performance and reduced R THJC for both IGBT and diode. In addition, exposed IMS ensures less stringent flatness requirements for the heat sink. It also offers significant flexibility in the module layout and internal electrical system. New Enhanced IMS Substrate features: Higher Reflow Temperature Higher Dielectric Strength Lower Thermal Resistance 6

7 Figure 2.5 IGBT & Diode on heat spreader (6x) Figure 2.6 Exposed IMS 7

8 3. Package and Pin Description 3.1. Outline Drawings Packaging options include staggered pinout for maximum creepage distances and straight (IRAM A) or 90 (IRAM A2) bend options for heat sink parallel or perpendicular to the circuit board. Dimensions in mm Figure 3.7 IRAM A package drawing Figure 3.8 IRAM A2 package drawing 8

9 3.2. Module Pin-out Description Table 3.1 shows the pin arrangement for IRAM Gen2 SIP1A module with pin description. Table 3.2 Pin arrangement for IRAM Gen2 SIP1A Pin Name Description 1 VB3 High Side Floating Supply Voltage 3 2 W,VS3 Output 3 - High Side Floating Supply 3 4 N/A None 5 VB2 High Side Floating Supply Voltage 2 6 V,VS2 Output 2 - High Side Floating Supply 7 8 N/A None 9 VB1 High Side Floating Supply Voltage 1 10 U,VS1 Output 1 - High Side Floating Supply N/A None 13 V + Positive Bus Input Voltage N/A None 16 ITRIP Current Protection Pin 17 VRU Low Side Emitter Connection - Phase 1 18 FLT/EN Fault Output and Enable Pin 19 VRV Low Side Emitter Connection - Phase 2 20 HIN1 Logic Input High Side Gate Driver - 21 VRW Low Side Emitter Connection - Phase 3 22 HIN2 Logic Input High Side Gate Driver - 23 HIN3 Logic Input High Side Gate Driver - 24 LIN1 Logic Input Low Side Gate Driver - 25 LIN2 Logic Input Low Side Gate Driver - 26 LIN3 Logic Input Low Side Gate Driver - 27 VTH Temperature Feedback 28 VCC +15V Main Supply 29 VSS Negative Main Supply 9

10 4. Internal Circuit and Features Figure 4.1 demonstrates the internal circuitry of Gen2 SIP1A module. The 600V IRAM module contains six IGBT dies each with its own discrete gate resistor, six commutation diode dies, one three phase monolithic, level shifting driver chip, three bootstrap diodes with current limiting resistor and an NTC thermistor for over temperature trip. The module has the following features: Integrated gate drivers and bootstrap diodes Temperature monitor Protection shutdown pin Low V CEON Trench IGBT technology Under-voltage lockout for all channels Matched propagation delay for all channels 3.3V Schmitt-triggered input logic Cross-conduction prevention logic Motor power range 0.25~0.75kW / 85~253 Vac Isolation 2000V RMS min and CTI> 600V High operating case temperature, T CMAX =125 C 4.1. UVLO This module provides undervoltage lockout protection on both the V CC (logic and lowside circuitry) power supply and the V BS (high-side circuitry) power supply. Figure 4.2 is used to illustrate this concept; V CC (or V BS ) is plotted over time and as the waveform crosses the UVLO threshold (V CCUV+/- or V BSUV+/- ) the undervoltage protection is enabled or disabled. Upon power-up, should the V CC voltage fail to reach the V CCUV+ threshold, the HVIC will not turn-on. Additionally, if the V CC voltage decreases below the V CCUV- threshold during operation, the undervoltage lockout circuitry will recognize a fault condition and turn off all IGBTs, and the FLT/EN pin will transition to the low state to inform the controller of the fault condition. Upon power-up, should the V BS voltage fail to reach the V BSUV+ threshold, the HVIC will not turn-on. Additionally, if the V BS voltage decreases below the V BSUV- threshold during operation, the undervoltage lockout circuitry will recognize a fault condition, and shutdown the high-side IGBT. However, there will be no FLT/EN low in this case. The UVLO protection ensures that the module operates only when the gate supply voltage is sufficient to fully enhance the power devices. Without this feature, the gates of IGBT could be driven with a low voltage, resulting in the power switch conducting current while the channel impedance is high; this could result in very high conduction losses within the IGBT and could lead to power device failure Over Current Protection The Gen2 SIP1A IRAM is equipped with an I TRIP input pin. Together with external shunt resistor, this functionality can be used to detect over current events in the negative DC bus. The internal HVIC gate driver will continuously monitor the voltage on I TRIP pin. Whenever the I TRIP voltage exceeds the reference voltage (V ITRIP+, 0.49V typical), a fault 10

11 signal will be generated on FLT/EN pin and all six IGBTs will be turned off, as shown in the Figure 4.3 below. V+ (13) Q1 D1 Q2 D2 Q3 D3 VRU (17) VRV (19) Q4 D4 Q5 D5 Q6 D6 VRW (21) R1 VB1 (9) U, VS1 (10) C1 R2 R3 VB2 (5) V, VS2 (6) C2 VB3 (1) W, VS3 (2) C3 R4 R5 R6 D9 D8 D7 23 VS1 22 VB2 21 HO2 20 VS2 19 VB3 18 HO3 17 VS3 LO1 16 R9 24 HO1 25 VB1 LO VCC Driver IC HIN1 (20) HIN2 (22) HIN3 (23) LIN1 (24) LIN2 (25) LIN3 (26) ITRIP (16) FLT/EN (18) 2 HIN1 3 HIN2 4 HIN3 LIN1 5 LIN2 LIN3 6 7 R7 F 8 ITRIP 9 EN RCIN LO3 14 COM 13 VSS 12 VTH (27) VCC (28) VSS (29) C4 C7 R8 C6 C5 Figure 4.9 IRAM Gen2 SIP1A module schematics 11

12 Figure 4.10 UVLO protection Figure 4.11 I TRIP and Fault Timing Waveform. The threshold of over current protection can be determined by V ITRIP+ / R SHUNT, if single bus shunt is used and it is connected directly to I TRIP pin. The delay time of fault reporting and I TRIP shutdown are specified in the datasheet. They are also shown below in Table

13 Table 4.3 Dynamic Electrical Characteristics. Symbol Description Min Typ Max Unit Conditions T FLT T ITRIP I TRIP to Fault propagation delay I TRIP to six switch turn-off propagation delay (see fig ns V IN =0 or V IN =5V, V ITRIP =5V µs I C =5A, V + =300V In the case of short circuit (which is the worst case of over current), the current level will rise very quickly. It is critical to ensure all IGBTs are turned off as soon as possible. Since the IGBTs in Gen2 SIP1A IRAM are short circuit rated, the safe operation of module can be guaranteed by minimizing the delay of external current sensing circuit and making sure its delay plus T ITRIP is less than IGBT short circuit rating. Because IGBT short circuit rating depends a lot on the gate voltage and junction temperature, it is important to consider all possible conditions. Table 4.2 below shows the short circuit rating of IGBT in IRAM A, under typical conditions. Table 4.4 IGBT Short Circuit Ratings. Symbol Description Min Typ Max Unit Conditions SCSOA Short Circuit Safe Operating Area SCSOA Short Circuit Safe Operating Area µs T J =25 C, V + =400V, V GE =+15V to 0V µs T J =100 C, V + =400V, V GE =+15V to 0V 4.3. Fault Output and Auto Clear Function As described in the previous section, in case of over current event, the FLT/EN pin will become low with the turning on of open-drain MOSFET. When over current condition is over, the open-grain MOSFET will be turned off. However, all IGBTs will remain off, until RCIN (shown in Figure 4.1) voltage can reach its positive going threshold. This is called Fault auto clean function, and this time is shown as T FLT-CLR in Figure 4.3. The length of T FLT-CLR is determined by resistance R7 and capacitor C5, and shown in datasheet as Table 4.3. Table 4.5 Fault clearance time Symbol Description Min Typ Max Unit Conditions Post I T TRIP to six switch turnoff clear time (see fig. 2) T C = T C = 25 C FLT-CLR ms 100 C It is critical that PWM generator must be disabled within Fault duration to guarantee shutdown of the system, and overcurrent condition must be cleared before resuming operation Over Temperature Protection All Gen2 SIP1A IRAM modules have internal NTC thermistor to sense the module temperature. Figure 4.4 shows the correlation between NTC temperature (T TH ) and IGBT junction temperature (T J ), which can be used to set the threshold for over temperature protection. 13

14 Figure 4.12 Correlation of IGBT junction temperature and NTC temperature of IRAM A Please note this curve is obtained at rated current condition. For example, Figure 4.4 is for IRAM A, and created at following condition as specified in module datasheet. Sinusoidal Modulation, V + =400V, Iphase=5A RMS, fsw=6khz, fmod=50hz, MI=0.8, PF=0.6 This correlation curve will be different in customer application, if for example the motor current is less than 5A RMS. The general guideline is that the difference between T J and T TH will be smaller, if the module dissipates less heat. In the extreme case of zero current, T J and T TH will be identical. Therefore, the curve in Figure 4.4 will be worst case (highest T J ) because of maximum current. It is also possible to customize the curve to better suit specific condition. One option for approximation is to calculate the power loss in both conditions (rated condition and custom condition) using the online tools (illustrated in section 8). Afterward, (T J - T TH ) difference can be scaled based on ratio of module power losses. 14

15 5. Absolute Maximum Ratings Here are the absolute maximum ratings from module datasheet. Table 5.6 Absolute maximum ratings of IRAM A Symbol Description Min Max Uni t V CES / V RRM IGBT/ FW Diode Blocking Voltage V V+ Positive Bus Input Voltage I T C =25 C RMS Phase Current (Note 1) I T C =100 C RMS Phase Current (Note 1) I PK Maximum Peak Phase Current (Note 2) F P Maximum PWM Carrier Frequency kh z P D Maximum Power dissipation per TC =25 C W V ISO Isolation Voltage (1min) V R T J Operating Junction Temperature (IGBT/Diode/IC) T C Operating Case Temperature Range T STG Storage Temperature Range T Mounting torque Range (M3 screw) Nm I BDF Bootstrap Diode Peak Forward Current A P BR_Peak Bootstrap Resistor Peak Power (Single Pulse) W V S1,2,3 High side floating supply offset voltage V B1,2,3-20 V B1,2, V V B1,2,3 High side floating supply voltage V V CC Low Side and logic fixed supply voltage V V IN Input voltage LIN, HIN, I TRIP, FLT/EN V A C V CES / V RRM IGBT, diode and HVIC driver are rated 600V. In addition, all modules are tested for leakage current at 600V, 100% at module production line. Please note modules are not tested for switching characteristics at 600V DC bus. V+ This is the maximum DC voltage for normal operation with switching action. The module is tested 100% at production line, for switching behavior at V+=450V. I T C =25ºC and T C =100ºC This is the maximum current the module can handle in steady state, due to thermal limitation. Please see section 8 for more detailed description. I PK The maximum current in the pulse condition. F P The maximum PWM switching frequency. P D Maximum Power dissipation per IGBT, This can be calculated as: P D = (T JMAX T C ) / R THJC V ISO 15

16 All Gen2 SIP1A modules are rated 2000V RMS for 1 minute. All modules are also recognized by UL, under File Number E The isolation test is performed 100% at module production line. T J, T C, T STG The maximum temperature ratings of module are 150ºC for junction temperature (IGBT, diode and HVIC) and 125ºC for case temperature, as well as storage temperature. The minimum temperature ratings are -40ºC for all three parameters. I BDF This is the maximum current for the bootstrap diode. Please note the current is limited by the internal bootstrap resistor of 22Ω. P BR_Peak This is the maximum power rating for the bootstrap diode in pulse condition, due to thermal limitation. V S1,2,3 This specifies the V BS voltage range from -0.3V to 20V, for all three phases. V B1,2,3 V B is rated up 600V, for all three phase. V CC The maximum voltage rating for V CC is 20V. It is also suggested to have typical V CC voltage at 15V, with less than +/-10% tolerance, in order to reduce the voltage range on V BS. V+ All input voltage (HIN, LIN, I TRIP, FLT/EN) has maximum voltage rating of 7V. 16

17 6. Bootstrap Circuit 6.1. Bootstrap Circuit Operation The high and low-side driver IC requires a floating voltage supply for each of the three high-side circuits that provide gate pulses to high-side IGBTs. A very convenient way of obtaining such floating voltage supplies is usage of bootstrap circuits. The following Figure 6.1 shows such an implementation for one phase of a three-phase switching inverter drive. The circuit is repeated for each phase. Figure 6.13 Schematic of bootstrap circuit for one phase When the low-side IGBT is on, the bootstrap capacitor C BS charges through the bootstrap diode D BS, resistor R BS and low side switch to almost 15V, since the V s pin of the IC is almost at ground potential. The capacitor C BS is so designed that it retains most of the charge when the low-side device switches off and the V s pin goes to almost the bus potential. Then, the voltage V BS being almost 15 V, the high-side circuit of the driver IC is biased by the capacitor C BS Bootstrap Capacitor Selection Selection of the bootstrap capacitor, diode and resistor is governed by several factors: 1. Voltage V BS has to be maintained at a value higher than the undervoltage lockout level for the IC driver. 2. The capacitor C BS does not charge exactly to 15V when the low-side switch is turned on, depending upon the drop across the bootstrap diode (V FBS ) and low-side switch (V CEON ). 3. When the high side switch is on, the capacitor discharges mainly via the following mechanisms: a. Gate charge Q G for turning the high-side switch on b. Quiescent current I QBS to the high-side circuit in the IC c. Level-shift charge Q LS required by level-shifters in the IC d. Leakage current I DL in the bootstrap diode D BS 17

18 e. Capacitor leakage current I CBS (ignored for non-electrolytic capacitors) f. Bootstrap diode reverse recovery charge Q RRBS Charge lost by the bootstrap capacitor in one switching cycle is given by the following equation: where f SW is the switching frequency and the other parameters are as defined earlier. This charge loss in the bootstrap capacitor as given above results in a drop in the voltage V BS across it. The value of C BS can be designed based on the desired voltage drop in V BS as follows, The drop in V BS can be set as a percentage of the value of V BS before turn-on of the high side switch. The lowest value of V BS in one modulation cycle is given by Note that the above equation gives the worst-case value of the bootstrap voltage with the low side IGBT conducting current in conjunction with high side diode. Current reversal leads to low side diode conduction in conjunction with the high side IGBT, whereupon the equation (3) changes to: Combining equations (1), (2) and (3) and using V BS = 1 % of V BS : For simplified calculation, equation (4) can be approached by I TOT is the total equivalent discharge current of C BS described above, and practically 1mA can be used for a good estimation for C BS. (1) (2) (3) (3a) (4) (4a) 6.3. Bootstrap Circuit Initial Charging and Bootstrap Diode A series resistor R BS of 22Ω is included in the IRAM module. This limits the peak current in the bootstrap circuit during initial charging, which has been known to cause driver latch-up under fast switching conditions. Typically, the low side switch is switched with a constant duty-cycle for charging the bootstrap capacitor initially. The time required for the initial bootstrap capacitor charging, after which input signals can be transferred to the switch gates, is given by: In the above equation, D is the duty cycle of the charging pulses. Note that this discounts effects of discharging processes and hence gives a minimum charging time. (5) 18

19 When high side switch or diode conducts, the bootstrap diode supports the entire bus voltage. Hence for a V system, D BS has to be rated at 600V. The peak current seen by D BS is determined by the series resistor R BS. However since this current spike is quite narrow, it does not seriously affect diode selection. Average current handled by the bootstrap diode is given by the product of the charge supplied to C BS during every switching cycle expressed by equation (1) and the switching frequency f SW. In order to minimize the power loss in the diode and to reduce the size of the bootstrap capacitor, reverse recovery charge in D BS should be as low as possible. For the same reason, reverse leakage current should also be low at the highest operating temperature. Finally, the knee voltage of the diode should be low to minimize the voltage drop across it during charging Recommended Bootstrap Capacitor Value The Gen2 SIP1A IRAM module contains three bootstrap diodes and a series resistor connected internally between the 15V supply V CC and individual V B pins of the three phases. Hence only appropriate bootstrap capacitors need to be connected on the external board. Some layout aspects have to be considered before doing that. Bootstrap capacitors should be connected as close to the V B and V S pins as possible to reduce stray inductance in the connections. Furthermore, it is recommended to use a small high frequency capacitor in parallel to a larger low frequency bootstrap capacitor for local decoupling. Here is an example on the bootstrap capacitor selection. If the switching frequency is f SW = 20kHz, and the allowable discharge voltage V BS =0.1V, from equation (4a), we can calculate the C BS =0.5μF. The capacitance is generally selected to 2-3 times of the calculated value in consideration of dispersion and reliability. Therefore, a 1.5μF bootstrap capacitor is recommended. In the IRAM datasheet, we provide recommended bootstrap capacitor values under different switching frequencies in a chart. From the following chart, at 20kHz switching frequency, 1.5μF is recommended. 19

20 Figure 6.14 Recommended bootstrap capacitor value vs. switching frequency 20

21 7. Interface circuit 7.1. General Interface Circuit Example Notes: Figure 7.15 Typical Application Connections 1. Electrolytic bus capacitors should be mounted as close to the module bus terminals as possible to reduce ringing and EMI problems. Additional high frequency ceramic capacitor mounted close to the module pins will further improve performance. 2. In order to provide good decoupling between V CC -V SS and V B1,2,3 -V S1,2,3 terminals, the capacitors shown connected between these terminals should be located very close to the module pins. Additional high frequency capacitors, typically 0.1µF, are strongly recommended. 3. Value of the bootstrap capacitors depends upon the switching frequency. Their selection is explained in the previous section. Bootstrap capacitor value must be selected to limit the power dissipation of the internal resistor in series with the V CC. 4. After approx. 2ms the fault (FLT) is reset. 5. PWM generator must be disabled within fault duration to guarantee shutdown of the system, overcurrent condition must be cleared before resuming operation. 21

22 8. Power Loss and Junction Temperature Calculation IPM design tool is a software tool to calculate power loss and junction temperature for IRAM, which is available online at: Since junction temperature is a critical factor for long-term reliability of power module operation, this tool provides very important information in addition to the data sheets. It can be used to choose appropriate module based on application needs and to size the heat sink to insure long-term reliability. In order to calculate the power loss, IPM design tool uses the built-in electrical models that describe the conduction and switching characteristics of both IGBT and diode integrated in the power module. Together with the built-in thermal impedance model, this tool is able to predict the maximum temperature inside a power module under various operating conditions, such as switching frequency, modulation frequency and case temperature. Figure 8.1 shows the web interface. Users can select up to three parts for comparison. After modifying the parameters according to the particular application, three kinds of analysis are available to calculate power loss, junction temperature, and maximum RMS motor current etc. The basic structure of this tool is shown in Figure 8.2. The ElectricalCalc function calculated the power loss of the semiconductor based on electrical model and input conditions like switching frequency, bus voltage, etc. The loss information is then passed to ThermalCalc function to calculated junction temperature. Figure 8.16 IPM design tool interface 22

23 ElectricalModel ThermalModel Input Condition ElectricalCalc Loss ThermalCalc Tj Figure 8.17 IPM design tool structure 8.1. Electrical Model The electrical model describes the conduction and switching loss of IGBT/diode with regards to current, under dc or single switching condition. The behavioral model is adopted instead of physical model because of fast simulation and reasonable accuracy in predicting the power losses. The basic equations are shown below. V V E E E CEON F ON OFF RR = V = = V TD T x ( h.1+ h2i ) I y ( m1 + m2i ) = = d1. I + a. I + ad. I d 2 b bd k I n The conduction and switching loss of the IPM are measured at various current levels so that curve fitting method can be used to derive the model parameters. The junction temperature is set to be at 150ºC because this is the worst case in terms of power semiconductor losses Thermal Model The simple thermal resistance model R THJC describes the steady state temperature rise between junction and case. However, when the modulation frequency of the inverter is relatively low, junction temperature will have large ripple beyond the average as described by Rth. The reason is that the power loss is not constant but has a fundamental frequency which is same as modulation. Therefore, IPM design tool uses thermal impedance from junction case to calculate the temperature ripple, such as the one shown in Figure Electrical and Thermal Calculation Under sinusoidal modulation, the power loss has to be calculated in each switching cycle because the device current is changing within the half modulation cycle, as illustrated in Figure 8.3. The upper portion is the high side IGBT current which is used to calculate E ON, E OFF and E CI of IGBT. The lower potion in Figure 8.3 is the low side diode current for E RR and E CD of diode. 23

24 Eon Eoff High Side IGBT Current Low Side Diode Current Eci tsw Err Ecd Figure Loss calculation of sinusoidal modulation Because the loss is not constant over time, its shape depends on current waveform and device parameters. Figure 8.4 illustrate the power loss of IGBT in a typical case. The cyan curve in the upper portion of Figure 8.4 is the power loss, while the purple curve is a simplification in order to use Z THJC curve to calculated temperature ripple. The blue curves are average power loss and junction to case temperature rise, which can be quite different than the real case when modulation frequency is only a few Hertz. Figure 8.19 Junction temperature calculation under sinusoidal modulation 8.4. IPM Design Tool Functions The electrical and thermal models of all released IPM are already incorporated in IPM design tool. When the user selects the part, the associated model will be used for loss and thermal calculation. This tool provides three analysis tools, based on models and calculation method as describe above, in order to help user choose the optimal IPM for their application. Switching Frequency Analysis: calculate the maximum motor current under different switching frequencies. Component Comparison: provide both power loss of IPM and maximum allowable case temperature, which can be used for heat sink selection Power Loss Analysis: calculate power loss under different switching frequencies. 24

25 8.5. Design Example In order to choose the right IPM, the designer needs to collect the information about the intended application. For example, a washing machine application requires maximum 6A RMS phase current at 16kHz switching frequency and dc bus voltage of 320V. Maximum junction temperature is limited to be 150 C. In this case, both IRAM A and IRAM A modules are able to deliver the required output current. However, because power losses and internal R THJC will be different for these two modules, the heat sink required to maintain junction temperature under 150ºC will also be different. IPM design tool can be used to calculate the required heat sink Rth. Figure 8.5 shows the result which includes both power loss of inverter and maximum heat sink temperature for this application, using Component Comparison analysis. Figure 8.20 Component comparison of 6A and 10A modules At 6A RMS, the power losses are 70W for the IRAM A module and 66W for IRAM A. The maximum allowable case temperatures are 91ºC and 106ºC, for 25

26 10A and 15A module respectively. The required heat sink Rth can be calculated as following: R th(s-a) = (T C T A ) / P - R th(c-s) Assuming the ambient temperature of 50ºC and R th(c-s) of 0.1ºC/W, the calculated heat sink Rth are showing in Table 8.1. Table 8.7 Module IRAM A IARM A Heat sink Rth comparison Heat sink Rth 0.49ºC/W 0.75ºC/W As can be seen from the above calculation, the smaller IPM will require a larger size heat sink. Therefore, the final choice should be made based on minimizing total system cost/size, including both the IPM and heat sink. Same method can be used to choose the right IPM for air conditioner application, which usually has 400V dc bus regulated by PFC front-end. Switching frequency will be lower than washer application in order to limit EMI noise. For example if the application requires 10A RMS current at 6kHz switching frequency, IPM design tool can be used to show the tradeoffs between 15A and 20A IPMs. This tool can also be used to analyze the effect of various design parameters such as modulation index, switching frequency, heat sink temperature and power factor etc on the current rating of the power module. This information can help designer to fine tune the system parameters to obtain an optimum solution for the application. For example, one important design parameter is the switching frequency. In this case, IPM design tool can be used to investigate the maximum motor current and power losses of IPM at different switching frequency, as shown in Figure 8.6 and Figure 8.7. User can also select up to three parts in each type of analysis for comparison purpose. It is quite obvious that loss increase and maximum current decrease while increasing switching frequency. 26

27 Figure 8.21 Switching frequency analysis of IRAM A Figure 8.22 Power loss analysis of IRAM A 27

28 9. Packing 28