IGBT Vs PWM

Ones hardware, the other is a technique (it is not an apples-to-apples comparison).

IBGT use over other technologies such as mechanical switches or transistors however results in IGBTs being better (better reliability over mechanical switches, lower losses over transistors, etc).

Try a wikipedia search of IGBT and PWM for more information.

Hmm, I'm guessing the relevant terms you're looking for are IGBT, BJT, MOSFET and then PWM, the idea being that the first three are different types of fast electronic switches used to create Pulse-Width Modulation, and IGBTs are preferred.

BJT, or bipolar transistors, were the original fast switching parts used to create PWM, 40 to 50 years ago. Switching frequencies were generally 50kHz or less. Then in the early 80s BJTs were supplanted by the superior properties of power MOSFETs, or Metal-Oxide Semiconductor Field-Effect-Transistors. Shortly after, IGBTs, or Insulated-Gate Bipolar Transistors came along (schematic symbol at left). These combined the best properties of BJTs (low switch saturation voltage) and MOSFETs (zero-DC-current gate drive). At high voltages MOSFETs suffer from a high R_ON (it rises by as much as the square of the breakdown voltage rating). For voltage ratings above about 300 volts IGBTs have a lower switch saturation voltage at high currents than does a MOSFET of the same die size. That's the key word, die size. A 1200-volt 20A IBGT costs less than a similarly-rated MOSFET, which is a big deal.

Furthermore, if you run these two parts for a short time at say 40A, the IGBT's voltage drop will be far superior to the MOSFET's voltage drop. In fact the IGBT's power dissipation rises roughly proportional to the current, while the MOSFET's rises power dissipation rises proportionally to the square of current. What's more, the switches in high-power applications generally run hot, and IGBTs have are much less affected by high temperatures than MOSFETs. That's because of a nice BJT property, V_CE(sat) is independent of temperature. The plot is from International Rectifier's app note AN-983, page 5. Ouch, that makes the MOSFET look really bad. The IPP60R099 in the plot is a 0.1-ohm 650-volt MOSFET priced at $5.60 at Mouser (qty 100), whereas the IRGP4063 is a 600-volt IGBT with V_CE(on) = 1.65V, rated at 48A, priced at $5.37 at Mouser, TO-247 case shown at left.

Notice how the MOSFET was rated with a resistance, but the IGBT was rated with a fixed voltage drop.

[We should add, the MOSFET still has its place over the IGBT, e.g., the IPP60 can switch 400V at 18A in just 5ns, whereas the IRGP4063 takes 45ns for 400V at 48A (and it's called an ultrafast part). Faster switching speeds make a big difference in some applications, but generally not in low-frequency high-voltage off-line switching.]

So this answers the question, Why IGBTs vs MOSFETs for PWM. And the general-wisdom stock-answer answer is, they are better for high-power use at voltages above 300 to 500 volts. They're cheaper and they have lower power loss at high currents.