Converting Transistor Capacitance curves to charge curves (Help)

[Glenn0010]

Hi All

Im reading the following book called GaN Transistor for efficient power conversion.

He breaks down the gate charge into the following components:


He then goes on to create the following curve. He states that the curve is created from the data sheet curves. However I am unsure of how this curve can be created from the datasheet.


Looking at the datasheets, the only curve that I can think of that would achieve this is the capacitance characteristics curve. However how do you go from that to the image above?

Any help would be appreciated.

Thanks GG

[Terry Bites]

The figure you need is the Reverse Transfer Capacitance from the device datasheet.
See https://assets.nexperia.com/documents/application-note/AN90005.pdf
Note that many GANFETs are composite devices not just a MOSFET made from GaN.

[Glenn0010]

The figure you need is the Reverse Transfer Capacitance from the device datasheet.
See https://assets.nexperia.com/documents/application-note/AN90005.pdf
Note that many GANFETs are composite devices not just a MOSFET made from GaN.

Hi Terry, thanks for your time!

I had a look at the AN you posted. I understand that for Qgd I'd need the reverse transfer capacitance, I Qgs the input capacitance represents most of ther charge (at higher drain source voltages).

However, how do I conver the capactiance to charge? Using Q=CV for the EPC2045 did not give the same results that are posted  in the image that I have attached.

Now I am probabbly interpreting this completely wrong

Taking the capacitance at 100 V Vds

Cgd= 2 pF..... Qgd = 2pF *100 V= 2e-10 C
Cgs = 700 pF............ Qgs = 700 pF * 100 V () =  7e-8 C

Qgd/Qgs = 2.86 e-3....... I am obviously doing someething wrong here as I'm several magnitudes off, of what he's getting.

Cheers

Glenn

[Vojtech]

Input capacitance Cds is charged only by input voltage so only by 5-10V,  and Cgd isn't constant, it depends on Uds.

[Glenn0010]

Input capacitance Cds is charged only by input voltage so only by 5-10V,  and Cgd isn't constant, it depends on Uds.

Hi Vojtech, thanks for replying

By input capacitance I think you mea Cgs right?. So I've also tried this calculation.

For the EPC 2045 taking the capacitances at 100 V Vds
Cgs = 700 pF * 5V = 3.5 nC
Cgd = 2 pF * 100 V =  0.2 nC

0.2 nC / 3.5 nC = 0.057142857. This ratio does not match the ratio shown in that graph at 100 V.

Anyone have an idea on what I'm doing wrong?

Best Regards

GG

[Glenn0010]

Hi All,

So I'm still having a go at this. Attached is a spreadsheet containing the transistor capacitance from the EPC2045. I've done some calculations and have some plots to try to get the charge curves from capacitance curve and hence the miller ratio. I still cannot agree with the result of the book , shown in the first post

Here is the transistor capacitance curves, digitized from datasheet



Using Q = CV, the Cgd was then converted to Qgd, though the curve does not look right. Vgs1 was left constant since it is the charge required to reach the threshold voltage.  There the two of them were then divided to get the miller ratio. What am I doing wrong? The grey curve is on the secondary axis (right)



This is what I'm trying to get to with the grey curve



 if anyone could have a look at the spread sheet attached at what I've done and maybe can find my error?