# Class E amplifier design

#### stingr4y

Joined May 4, 2021
6
Good morning.

I am using an Infineon IMW65R072M1HXKSA1 to build a class E amplifier to operate @ 13.63 Mhz, for operation in the 20-50 watt power range. I am struggling to get the Mosfet to work in class E mode. When I drive the Gate @ 13.63 Mhz , the DC power supply (48 volts) pulls maximum current, even with the load disconnected, implying that the Mosfet is dissipating power across it's internal resistance. As I understand it, voltage and current ought to be 180 deg out of phase. I am driving the mosfet @ approximately 17 volts.

Please see attached the schematic and calculator for approximately the component values I'm using.

Basically I don't know how to keep the voltage and current sufficiently out of phase across the MOSFET such that I can get an efficiency improvement on a class AB type design. Any thoughts would be much appreciated.

#### Attachments

• 158.7 KB Views: 103
• 77 KB Views: 95

#### DickCappels

Joined Aug 21, 2008
10,104
I don't have the datasheet for your MOSFET but if the input capacitance is 2000 pf the capacitive reactance will be about 6Ω. That is going to be 3 amps peak current.

Why so much power?

I don't understand where the 180° difference in phase between voltage and current. Where did you see that?

#### stingr4y

Joined May 4, 2021
6
Thanks for your reply. As I understand it, in class E amplifier operation, voltage and current through the Mosfet are 180 deg out of phase for a (theoretical) 100% efficiency. I'll get some datasheet values for you, cheers

The applicaiton is piezo electric actuator.

#### stingr4y

Joined May 4, 2021
6
I don't have the datasheet for your MOSFET but if the input capacitance is 2000 pf the capacitive reactance will be about 6Ω. That is going to be 3 amps peak current.

Why so much power?

I don't understand where the 180° difference in phase between voltage and current. Where did you see that?

#### Ylli

Joined Nov 13, 2015
1,080
Hmm, I see a 48VDC source connected to a SPST switch (the FET) to near ground. The switch is turned on and off at 13 MHz. When off the current is zero. When on, the current is very high. What do you expect the average current to be? Are you sure you don't want that inductor (L1) in series with the 48 volts?

#### stingr4y

Joined May 4, 2021
6
Hmm, I see a 48VDC source connected to a SPST switch (the FET) to near ground. The switch is turned on and off at 13 MHz. When off the current is zero. When on, the current is very high. What do you expect the average current to be? Are you sure you don't want that inductor (L1) in series with the 48 volts?
thats a mistake in my schematic sorry, the 48v is sitting behind L1.

#### Ylli

Joined Nov 13, 2015
1,080

Actually my mistake, I mentally transposed in and out on this vreg, and figured the +48 was coming from this. But you might want to reconsider where you are tapping the 48 volt bus for the input of this reg. I think you want it on the left side of L1.

#### stingr4y

Joined May 4, 2021
6
View attachment 237606
Actually my mistake, I mentally transposed in and out on this vreg, and figured the +48 was coming from this. But you might want to reconsider where you are tapping the 48 volt bus for the input of this reg. I think you want it on the left side of L1.
yes thats true, I was just thinking that

#### stingr4y

Joined May 4, 2021
6
I don't have the datasheet for your MOSFET but if the input capacitance is 2000 pf the capacitive reactance will be about 6Ω. That is going to be 3 amps peak current.

Why so much power?

I don't understand where the 180° difference in phase between voltage and current. Where did you see that?
Hi Dick,

I've been thinking about this and this may be the core of the problem. I measure 450pf between drain and gate, I guess at 13.6mhz that is basically a low resistance path between drain and gate.

What is the remedy? less capacitance all round? ie. a better mosfet? or are there strategies to manage this?