Stable switching speeds for IGBTs in SMPS design

[thadman]

I'm interested in designing a SMPS for the power supply section of an amplifier.

As frequency increases, the finite time associated with opening/closing the IGBT becomes a larger portion of the total cycle. The majority of the resistance associated with IGBTs is associated with opening/closing. As a result, I believe heat dissipation may increase rapidly with frequency.

This effect counters the benefit of increasing the frequency for full-wave rectification. I believe p<-->p voltage ripple can be defined as 1/2fC, where f=frequency and C=capacitance. If we increase frequency, we can drastically reduce the voltage ripple.

I believe ringing is also associated with higher frequencies.

I know ultrafast IGBTs are available. However I'm not sure how high in frequency they can be taken while remaining stable.

Does STMicroelectronics produce high performing IGBTs?

STMicroelectronics | Error Report

STGW35HF60WD

Is anyone familiar with this model?

Who produces the best IGBTs? I'm specifically looking for IGBTs capable of high voltage and high frequency operation.

Could anybody offer further insight on this?

Thanks,
Thadman

[Sch3mat1c]

Ultrafast IGBTs are almost as fast as MOSFETs. You could use them up to 200kHz, I suppose. The downside is "tail current", which can cause substantial losses. Snubbers or resonant topologies are preferred at HF.

Field Stop technology is a pretty good structure. Fairchild owns this trademark I believe; I don't know if other manufacturers offer the same or similar structures under different names.

Tim

[sawreyrw]

Thadman,

When you operate an IGBT at high frequencies, switching losses will become the dominate factor in total power loss (but don't neglect conduction losses). The spec related to this is energy lost (per cycle) in turn on (Eon) and turn off (Eoff). The power lost is frequency times (Eon +Eoff). The data sheet for the part you mentioned doesn't provide enough information, since it gives the losses at only one operating condition and it is only the typical value. Sometimes vendors give curves of losses vs. Ic; this is what you need. If you plan to push the part, then you need to know worst case data.

Rick

[thadman]

Originally Posted by sawreyrw

Thadman,

When you operate an IGBT at high frequencies, switching losses will become the dominate factor in total power loss (but don't neglect conduction losses). The spec related to this is energy lost (per cycle) in turn on (Eon) and turn off (Eoff). The power lost is frequency times (Eon +Eoff). The data sheet for the part you mentioned doesn't provide enough information, since it gives the losses at only one operating condition and it is only the typical value. Sometimes vendors give curves of losses vs. Ic; this is what you need. If you plan to push the part, then you need to know worst case data.

Rick

HGTG30N60A4D

Are you familiar with this IGBT produced by Fairchild?

According to the data sheet, its used for SMPS. However, I'm not sure if a better model is available. If you are familiar with any, I would greatly appreciate a recommendation.

[ontoaba]

High voltage high frequency is MOSFET, but if you need high current or ultra high voltage is IGBT. In high current and very high voltage, IGBT still have low drop voltage at around 1volt or less. Like this IRG4PC60, this IGBT is reliable.