IGBT or FET module for 7kw Grid tie inverter?

Hello,
Are there any of these type of IGBT modules (or FET modules) for our intended 7kw grid tied inverter which we want to operate at 30khz?
It is for 7kw, and 230VAC mains only.
We will have 4 FETs (or IGBTs)


https://www.americas.fujielectric.com/sites/default/files/1MBI1200U4C-120.pdf

sim (ltspice) of inverter attached

- - - Updated - - -

Even this one has 135milliohm of rdson per fet,...too much

https://www.fairchildsemi.com/datasheets/FP/FPF1C2P5BF07A.pdf

- - - Updated - - -

this one has great rdson but terrible switching loss, and series diodes which aren't needed in a GTI..there's nothing just right

https://www.digikey.co.uk/product-detail/en/APTC60AM18SCG/APTC60AM18SCG-ND/1920299

Have you tried looking for any here? Haven't looked to see if they are any good, just a quick suggestion for places I have gone before for IGBT modules

https://www.mitsubishielectric.com/semiconductors/products/powermod/igbtmod/

https://www.vishay.com/modules/igbt-modules/vces-600/

Thanks, the below igbt has total switching energy of 17mJ. So at 30khz that would be 510W. Do you agree?

https://www.vishay.com/docs/95721/vs-gp100ts60sfpbf.pdf

Other than switching loss, is the Mosfet vs igbt choice just governed depending on whether Vce(on)*Ic(average) for the igbt is less than I^2.RDS(ON) for the mosfet?

Are you really planning to run a hard switching IGBT inverter at 30 kHz?

That Vishay igbt is designed for dc to 1 kHz as is listed on the product summary. I assume this means low conduction losses and high swiching losses.

Thanks, when I worked at an amplifier company, everybody was going on about igbts that can replace mofets, in 300w+ smps, even at 70khz+...thing is, I cant find them. What would you say is generally the max f(sw) for an igbt for a 500w+ smps with vin=380vdc?

I am a beginning electronics hobbyist and I have never designed electronics. However, I'm somewhat familiar with power electronics from my work. I tell this that you can put my writing to proper context.

treez said:

Thanks, when I worked at an amplifier company, everybody was going on about igbts that can replace mofets, in 300w+ smps, even at 70khz+...thing is, I cant find them. What would you say is generally the max f(sw) for an igbt for a 500w+ smps with vin=380vdc?

Click to expand...

I have read some igbt and mosfet datasheets and application notes in order to find out when to choose mosfet and when igbt is better. And this is just to one off test prototype. I found out that voltage rating has a big influence. 1200 V and 1700 V rating use igbt, 650 V and below use mosfet (well, SIC mosfet is available at 1200V and beyond if higher switching frequency is wanted). Low voltage industrial frequency converters (driving 400 V-690 V motor) use 1200 V and 1700 V IGBTs and there is quite a big market for those. I guess that is the reason there are lot of IGBT modules. Mosfet are used on several different topologies and perhaps it is a reason mosfet modules are not so common.

About IGBT replacing mosfet on SMPS. Perhaps it was so that mosfet got better (like coolmos from Infineon) while IGBT didn't improve that much. On SMPS switching frequency affects to magnetics. IGBT might be cheaper than mosfet but IGBT's lower switching frequency means bigger inductor. IGBT works well on frequency converter as inductance is on the motor already as given and there is enough that inductance that lower switching frequency is ok.

I don't know what is max usable f(sw) with igbt when vin=380 VDC. I picked one igbt from mouser with 600 V voltage and 11 A current rating. The IGBT is optimized for 40 kHz and could be ok for SMPS but I don't know if it is better choice than mosfet.

https://www.mouser.fi/ProductDetail/International-Rectifier/IRG4IBC20UDPBF/?qs=%2fha2pyFaduiNZJ%252bu1g3sYrm7bKzBjdMuWFUL5mlyuGZD%252brRVE1xg1g%3d%3d

Unfortunately a big mosfet will always give lower losses for > 50kHz than an igbt, IGBT's are getting better but the tail current and Von mean that 50Hz - 20kHz is where they are best.

We have substituted IGBT's for fets in bootstrap ckts where the Von of the IGBT was ~1.5V at 10A allowing the gate drive boot-strap to get more volts to the gate circuit - the mosfet was dropping 3V - just too much.

However - for rugged inverter circuits - igbt's are preferred due to the 10uS short ckt capability which allows them to be turned off (in 1-2uS) and save their bacon - turning fets off like this kills them.

So - horses for courses...

I picked one igbt from mouser with 600 V voltage and 11 A current rating. The IGBT is optimized for 40 kHz and could be ok for SMPS but I don't know if it is better choice than mosfet.

Click to expand...

Saying this iGBT is "optimized for 40 kHz" somehow abridges the facts. I think you would better say it can be operated up to 40 kHz under circumstances, which is quite a lot for an IGBT. If you look a the load current versus switching frequency diagram, you'll see that 40 kHz is rather the end of the road.

Simply consider that IGBT switching losses are always higher than MOSFET, but IGBT are still a reasonable choice for high power switchers.

FvM said:

Saying this iGBT is "optimized for 40 kHz" somehow abridges the facts.

Click to expand...

You are right. My intention was not to mislead but somehow between reading the datasheet and writing the sentence I forgot what it really said on the datasheet. It says it is ultrafast and optimized for 8-40 kHz hard switching. This already indicates that 40 kHz is high switching frequency for this device. My idea was to show that IGBT for 30 kHz exists but mosfet could be more suitable for SMPS.

Simply consider that IGBT switching losses are always higher than MOSFET, but IGBT are still a reasonable choice for high power switchers.

Click to expand...

Obviously there has to be use for IGBT on something like 600 V rating as they are manufactured. You mention high power and Easy peasy mentioned short circuit ruggedness. Most of the intelligent power modules made for 230 V supply washing machine motors etc. seems to be IGBT. They are often 20 kHz max switching frequency which is to avoid audible noise. Then there are IGBT for induction cooking.

Regardless of which switching device you choose, just remember that the modern devices available nowadays are much improved to what was available 10 years ago. And they still managed to design very good inverters back then.

What I would do, is to write a small comparison table listing the pros and cons of different devices.
List all the important performance parameters.
Do not forget to include the most important one: cost. Which should include the cost of the surrounding components, like drivers, snubbers, etc.

That is how I make my topology decisions, and most importantly, to explain and defend the decision to the money counters.

However - for rugged inverter circuits - igbt's are preferred due to the 10uS short ckt capability which allows them to be turned off (in 1-2uS) and save their bacon - turning fets off like this kills them.

Click to expand...

[post#8 above]
This is a good point, what is the maximum current that a mosfet can turn off without being killed?
Take for example the SCT2160 SiC FET..(rated for 22A continuous)
http://www.wakamatsu-net.com/biz/sct2160ke.pdf

(or any example of your choice)

The total switching energy is dependent on average current switched off (& on), supply rail, and temperature, quality of anti=parallel diodes, and any snubbing present - so the max values on the data sheet may lead you to compute too high a value - hence the graphs.

As to the 2160 SiC mosfet - it it only a 165W device, with 0.7 - 0.9 degC/W junc to case thermals, this tells you the die is very small compared to a similar mosfet (>500W), or IGBT, as such it has virtually no short circuit capability (much like a mosfet) and turning off a large current quickly will cause voltage overshoot on the drain-source due to the di/dt in the wiring inductance, IGBT's are inherently current limiting for a fixed gate voltage - mosfets do not have this and hence are not used in inverters much above 1kW.
If you have a 50A peak o/p on an inverter, you would typically use 100A devices (igbt) for a safety margin.