Yes, you can use IGBTs for this purpose, but as with any other transistor you would have to make sure that the device is properly cooled. You must also make sure, by means of experiments, that your gate control is able to control the device in the way you want to use it. For low frequency operation, the IGBT is not very much different from a MOSFET, except that the voltage drop is different for a certain value of gate voltage. At higher frequencies, the IGBT will not at all respond in the same way as a MOSFET because of its bipolar nature. Generally, the IGBT will act in a "slow" manner at high frequencies.

Can replace But Switching loss (loss at the time of on or off ) is more in IGBT in comparison to MOSFET.

so better use MOSFETs...

Dear Nishant, you seem to be confused: where does 'switching loss' arise in linear regulators ?

In comparison of MOSFET, IGBTs have been the preferred device because, They also have been the device of choice in applications that employ high voltages (> 1000 V), high allowable junction temperatures (> 100°C), and high output powers (> 5 kW).

I again ask you to read my question carefully: can IGBT be used safely in the linear mode ? I am not asking for switching mode operation characteristics or advantages of IGBTs.

It should be possible. IGBT;s have a linear range, similar to that of other transistors - keep within the device temperature v.s. current, and other ratings, and you'll be fine.

First: What do you mean with High Power? I think that is necessary to know the input/output voltage range, power range, wether the efficiency is important or not, etc .. in order to give an answer

IGBT modules are not suitable because of current filamentation. Use Mosfets.

'High Power' means like 28V, 600A output for aircraft system testing and also 100V, 200A for supply to magnetic fields in research labs. Since this is a linear regulator, efficiency is not important. Currently, we are using parallel connected bipolar transistors or MOSFETs, each which are all of smaller current rating while IGBTs are available in larger current rating, hence exploring the option.

At that current demand and with your application, is a switcher out of the question? Operating any kind of bipolar junction in linear mode is going to leave you with heat to contend with and a huge transformer (part of its capacity going to wasted heat). So, if you do use IGBTs (as David above alluded to) you will probably find you must d-rate the given current handling which may not be an improvement over your current supply.

A switcher would tighten up your parameters, provide some reasonable control and improved efficiency (realizing efficiency is not a design concern).

Cheers,

From my point of view, this would be a very unique application of IGBTs which are mostly see working at voltages of at least 600 Vpk. Well, but I understand why it may make sense: Right, you can buy large modules which allows you to get rid of paralleling devices by yourself. Probably the rather fixed VCEsat is also important as it allows you to determine the headroom in voltage you need to achieve - and this for virtually any load. Furthermore, modules may be more suited for the vast power that gets dissipated. At 600 A and say 2 Vce, you get 1.2 kW! I think, I would go for a high frequency switcher and a well designed, multi-order filter instead ..... .

Yes, you can use IGBTs for this purpose, but as with any other transistor you would have to make sure that the device is properly cooled. You must also make sure, by means of experiments, that your gate control is able to control the device in the way you want to use it. For low frequency operation, the IGBT is not very much different from a MOSFET, except that the voltage drop is different for a certain value of gate voltage. At higher frequencies, the IGBT will not at all respond in the same way as a MOSFET because of its bipolar nature. Generally, the IGBT will act in a "slow" manner at high frequencies.

I agree with H.-P.Nee: you have to re-design (probably) an adequate driver and protection circuit for the IGBTs. Although MOSFET and IGBTs exhibit capacitance input, it is also true that IGBTs need negative voltage to switch-off, and if you have voltage variations in collectors side, you will have to keep in mind, since they can modify gate conditions. And of course, you will have to ensure an adequate cooling. What about the dynamic of your circuit?

Thanks to all who have felt that IGBTs can be tried out in DC linear voltage regulators. I am going ahead with a scaled down prototype trial. Is there any published literature available in the form of Application Notes or Research paper ?

Why do you want to use linear regulator?

The most benefit of linear regulator are accurate regulation, small output ripple, very small output noise.

I think for your application this issues are not very important. A 28V, 600A power supply is usually used as starter of aircraft, and accurate regulation or small output ripple in not necessary. Some of the manufactures implant it by diode rectifier.

I would be appreciate you posting your findings from you prototype testing. Thank you,

Dear Ahmad, along with starting of aircraft, its communication equipment are also tested from this supply while on ground, without wasting aviation fuel, Hence the requirement of low noise. Same is the requirement for high stability (less than 50ppm) power supplies for special magnetic fields for research laboratories.

IGBTs are very good for that use due to their high current carrying capacities and voltage rating. You can virtually make any type of regulated power supply in the usable range of voltages.

My intuition tells me that using IGBTs for a linear amplifier would not be your best choice. IGBTs are designed to be used as switchers, i.e. always switching. Where some other transistors are designed for use in a linear amplifier. I would choose one of those transistors.

What is missing from your question is (at least)  : the voltage level, the output current.

From what I know, aircraft DC voltages are 270 V and 540 V (whith specification of overvoltage during transients) and the power should be in the 1kW to 50 kW range.

As H.P. Nee indicated, cooling is the first thing to look at, and you will find that using linear regulation produces a hell of a lot heat.

I suggest that you use a switched pre-regulator (IGBT buck circuit for example for the high voltage handling  >500V ) followed by a floating voltage linear fine regulator  (MOSFET for high current and high speed at low voltages