[Mook Johnson]

http://www.st.com/stonline/books/pdf/docs/9572.pdf

I asked a broader question before and it must have been too much.

Here is a simpler version and more to the point of what I'm after.

Using the above MOSFET in a switching application with the following
specifications.

Configuration: half bridge
Bus Voltage: 500V
peak flat top current 1.5A (start of ramp on step is 1A, Anti-parallel diode
freewheels the inductaive load)
May place an external diode if the internal diode causes too much
dissipation.

switching freq = 10kHz 90% duty cycle

I can figure out the losses due to the Rds_on but I'd like to get an
estimate of the switching losses.

The FETs will be driven with a TC4427 1.5A fet driver with a small ~5 ohm
resistor to damp the resonant tank between the pin an trace inductance and
the gate capacitance.

Can someone help me with this please?

 

[Gerhard Fiedler]

http://www.st.com/stonline/books/pdf/docs/9572.pdf

Configuration: half bridge
Bus Voltage: 500V
peak flat top current 1.5A (start of ramp on step is 1A, Anti-parallel diode
freewheels the inductaive load)
May place an external diode if the internal diode causes too much
dissipation.

switching freq = 10kHz 90% duty cycle

I can figure out the losses due to the Rds_on but I'd like to get an
estimate of the switching losses.

As an approximation, take the switching times (on/off) and assume that the
FET resistance goes linearly from 0 to infinity (or vice versa) in that
time. Integrate over the power; that'll give you the energy that each
switching process leaves in the FET. Multiply by the number of switches per
second, and you get the average switching power loss in the FET.

Gerhard

 

[Donald]

As an approximation, take the switching times (on/off) and assume that the
FET resistance goes linearly from 0 to infinity (or vice versa) in that
time. Integrate over the power; that'll give you the energy that each
switching process leaves in the FET. Multiply by the number of switches per
second, and you get the average switching power loss in the FET.

Gerhard

Please help me understand, why is this information important ??

donald

 

[J.A. Legris]

Please help me understand, why is this information important ??

donald

 

[D from BC]

http://www.st.com/stonline/books/pdf/docs/9572.pdf

I asked a broader question before and it must have been too much.

Here is a simpler version and more to the point of what I'm after.

Using the above MOSFET in a switching application with the following
specifications.

Configuration: half bridge
Bus Voltage: 500V
peak flat top current 1.5A (start of ramp on step is 1A, Anti-parallel diode
freewheels the inductaive load)
May place an external diode if the internal diode causes too much
dissipation.

switching freq = 10kHz 90% duty cycle

I can figure out the losses due to the Rds_on but I'd like to get an
estimate of the switching losses.

The FETs will be driven with a TC4427 1.5A fet driver with a small ~5 ohm
resistor to damp the resonant tank between the pin an trace inductance and
the gate capacitance.

Can someone help me with this please?

I thought I wrote down the formula...but can't find it...anyways.. I
had to Google for it. It's out there..in the wild...somewhere..
D from BC

 

[colin]

http://www.st.com/stonline/books/pdf/docs/9572.pdf

I asked a broader question before and it must have been too much.

Here is a simpler version and more to the point of what I'm after.

Using the above MOSFET in a switching application with the following
specifications.

Configuration: half bridge
Bus Voltage: 500V
peak flat top current 1.5A (start of ramp on step is 1A, Anti-parallel
diode freewheels the inductaive load)
May place an external diode if the internal diode causes too much
dissipation.

switching freq = 10kHz 90% duty cycle

I can figure out the losses due to the Rds_on but I'd like to get an
estimate of the switching losses.

The FETs will be driven with a TC4427 1.5A fet driver with a small ~5 ohm
resistor to damp the resonant tank between the pin an trace inductance and
the gate capacitance.

Can someone help me with this please?

worst case with inductive load, assume the current doesnt change untill the
top freewheel diode starts conducting,
vds rises linearly limited by the gate current charging cdg,
the power at switch off will then be 1/2 V * I * tfall * freq
power at switch on depends on the initial current.

dont forget the gate energy required too, although it should be much less.

Colin =^.^=

 

[Gerhard Fiedler]

Please help me understand, why is this information important ??

Sometimes, when you're switching fast enough, the switching losses are the
ones that determine the necessary heat dissipation. Or the available
dissipation determines the maximum switching frequency.

Gerhard