IGBT capacitance as function of C-E voltage - UK Vintage Radio Repair and Restoration Discussion Forum

Al (astral highway) · 19th Nov 2022, 11:37 pm

Here’s an extract from a datasheet for a fast, TO3 package IGBT. I have some questions, which are theoretical and not application specific.

The device under test has max C-E voltage of 1200, and as shown on Y axis, gate capacitance Cies is close to 5nF.

1/ if we look closely at the slope as it approaches this value on the Y axis, it looks as if it will start to decrease as the X axis (voltage) goes negative.

This is not shown but is this inference true and is the slope a mirror at similar values on the negative X axis? If this is true , what does it mean in practice and who needs to know this?

2/ if we look at Cres (reverse transfer capacitance), why is this only plotted for low voltages compared to the max design value ? Who needs to know this and how is this data used?

Curious to know this.

Cruisin Marine · 19th Nov 2022, 11:46 pm

I guess (looking at it first glance anyway) because above the voltage shown on the chart (30volts) the capacitance will decrease and be insignificant (in most cases).

Radio Wrangler · 20th Nov 2022, 5:13 am

The IGBT is a 2-stage device MOSFET followed by Bipolar.

The gate capacitance is not too dissimilar from a lower power MOSFET. The gate capacitance is distributed across the channel and the channel voltage changes along its length. There are also diode junctions doing reverse biased isolation, so they act as varactors. Device gain also depends on voltage. Mr Miller is at work, too. So the gate capacitances are complicated, voltage dependent and non-linear. That's plenty to be going on with. Compared to just a big single MOSFET of the same ratings, The IGBT input MOSFET is smaller and easier to drive.

The simple model of Cge, Cce and Ccg doesn't fit well, hence the graphs of values versus voltages. Remember that the voltages swing around as the device is used.

If you're using them as fastish switches, it can be easier to think of charge injection than as capacitance charging.

David

Al (astral highway) · 20th Nov 2022, 6:04 pm

‘Charge injection rather than capacitance charging’

That’s a helpful heuristic. Relevant to me as switching in <40nS.

Thank you David. Is that a record for late night ‘all your questions answered by our resident expert!’ or does it count as early morning?

19th Nov 2022, 11:37 pm	#1
Al (astral highway) Dekatron Join Date: Nov 2006 Location: London, UK. Posts: 3,496	IGBT capacitance as function of C-E voltage Here’s an extract from a datasheet for a fast, TO3 package IGBT. I have some questions, which are theoretical and not application specific. The device under test has max C-E voltage of 1200, and as shown on Y axis, gate capacitance Cies is close to 5nF. 1/ if we look closely at the slope as it approaches this value on the Y axis, it looks as if it will start to decrease as the X axis (voltage) goes negative. This is not shown but is this inference true and is the slope a mirror at similar values on the negative X axis? If this is true , what does it mean in practice and who needs to know this? 2/ if we look at Cres (reverse transfer capacitance), why is this only plotted for low voltages compared to the max design value ? Who needs to know this and how is this data used? Curious to know this. Attached Thumbnails __________________ Al

19th Nov 2022, 11:46 pm	#2
Cruisin Marine Heptode Join Date: Jul 2018 Location: Worthing, West Sussex, UK. Posts: 993	Re: IGBT capacitance as function of C-E voltage I guess (looking at it first glance anyway) because above the voltage shown on the chart (30volts) the capacitance will decrease and be insignificant (in most cases). __________________ "Behind every crowd, there's a silver Moonshine"

20th Nov 2022, 5:13 am	#3
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,874	Re: IGBT capacitance as function of C-E voltage The IGBT is a 2-stage device MOSFET followed by Bipolar. The gate capacitance is not too dissimilar from a lower power MOSFET. The gate capacitance is distributed across the channel and the channel voltage changes along its length. There are also diode junctions doing reverse biased isolation, so they act as varactors. Device gain also depends on voltage. Mr Miller is at work, too. So the gate capacitances are complicated, voltage dependent and non-linear. That's plenty to be going on with. Compared to just a big single MOSFET of the same ratings, The IGBT input MOSFET is smaller and easier to drive. The simple model of Cge, Cce and Ccg doesn't fit well, hence the graphs of values versus voltages. Remember that the voltages swing around as the device is used. If you're using them as fastish switches, it can be easier to think of charge injection than as capacitance charging. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

20th Nov 2022, 6:04 pm	#4
Al (astral highway) Dekatron Join Date: Nov 2006 Location: London, UK. Posts: 3,496	Re: IGBT capacitance as function of C-E voltage ‘Charge injection rather than capacitance charging’ That’s a helpful heuristic. Relevant to me as switching in <40nS. Thank you David. Is that a record for late night ‘all your questions answered by our resident expert!’ or does it count as early morning? __________________ Al