So basicly IGBT is transitor but controlled like FET? What about thermal runaway?
Hy Fezder,
That is exactly correct; an IGBT is only a MOSFET driving the base of a power BJT in a common case. IGBTs also have a diode connected between the collector and emitter (just like the intrinsic Schotkky diode in a MOSFET). The fabrication process is complicated and hence the relatively high price of IBBTs. Because the two transistors are fabricated at the same time, the characteristics of the MOSFET can be optimized to match the base drive characteristics of the power transistor. There are other advantages too, but they are only secondary.
IGBT have a major shortcoming compared to BJTs and MOSFETS for general purpose applications: IGBT collector saturation voltage (Vsat) cannot be any lower than the band gap voltage of the bipolar power transistor emitter/base, which is typically 600mV at low currents. This coupled with higher cost and the lack of a P type IGBTs means that IGBTs are not widely used.
The three main practical limits of a power transistor, BJT, MOSFET, IGBT, are maximum junction temperature (TJmax), thermal resistance junction to case (ThRJC), and Safe Operating Area (SOA). The maximum power dissipation figure for a transistor is meaningless and is only a marketing ploy (not the case for some specialist transistors like RF types).
Thermal runaway is no more of an issue with IGBTs than a normal power BJT, like a 2N3055 for example. Thermal runaway was a major problem with germanium transistors with a maximum junction temperatures limited to around 100 deg C, but not really with silicon which can stand a junction temperature of either 150 deg C or 175 deg C, depending on manufacturer.
If you stay within the transistor's SOA, defined by the graph of Fig 2 in the attached 2N3055 data sheet for example, you will not damage a transistor or cause thermal runaway. The term thermal runaway is not used much there days; it has effectively been replaced by secondary breakdown. For example the 2N3055 is a 60V, 15A, 115W transistor but it can only stand 800ma with a collector/emitter voltage (VCE) of 60V. And at 15A collector current it can only stand 8V. In fact, even this would not be possible because 8V * 15V= 120W which far exceeds the practical dissipation for a 2N3055).
The Thermal Resistance Junction to Case (ThRJC) allows you to calculate how much power the transistor can dissipate, and not exceed the TJmax, with or without a heat sink, in a given ambient temperature. The ThRJC for the 2N3055 is 1.52 degrees Centgrade Watt (DCW) which is pretty poor: 0.5DCW would be more reasonable for a TO3 can power transistor. In practical terms the maximum safe power dissipation for a 2N3055 would be around 20W. As a general rule of thumb, 30W is the maximum safe dissipation for any TO3 case transistor without liquid cooling.
The IRGP4068D IGBT ThRJC is 0.5DCW which is good for a TO247 plastic case power semiconductor. The TJmax is also good at 175 deg C. The SOA is a bit complicated because it is specified in terms of switching performance but as it is a 600V part, unless you are doing some high voltage work, secondary breakdown will not be a problem. Of course, you must not exceed the IGBT maximum junction temperature of 175 deg C. This means that without a heatsink the IRGP4068D can only dissipate 3.4W and then only providing the case is in free air with a temperature not exceeding 40 deg C.
To make the IRGP4068D conduct collector current, just place a voltage of 4V to 6.5V (Gate Threshold Voltage) between the gate and emitter. At DC the gate current will be low: 100nA maximum. To increase the collector current make the gate voltage more positive, just like for a MOSFET.
2N3055 Data Sheet
http://www.onsemi.com/pub_link/Collateral/2N3055-D.PDF
IRGP4068D-EPbF Data Sheet
http://www.irf.com/product-info/datasheets/data/irgp4068dpbf.pdf