300V AC MOSFET Switching via PWM

Thread Starter

LmanYan

Joined Jul 2, 2018
2
Hello everyone,
I am currently undergoing a project where I am trying to design a circuit where an alternating 300 volts source will be it's input and it's output will be
part of the input voltage modified by the on-off switching of a certain MOSFET configuration.
Looking through the board I came across this design:

Where the MOSFETs are driven by an IR2111.
Unfortunately, I had no luck in simulating this circuit and what I get is only the negative part of the input voltage and the PWM
seemingly having no effect.

It could be that I just sketched the design wrong, but I would love to hear your guys input about the simulation/idea/configuration.
 

Attachments

ebeowulf17

Joined Aug 12, 2014
3,291
Hello everyone,
I am currently undergoing a project where I am trying to design a circuit where an alternating 300 volts source will be it's input and it's output will be
part of the input voltage modified by the on-off switching of a certain MOSFET configuration.
Looking through the board I came across this design:

Where the MOSFETs are driven by an IR2111.
Unfortunately, I had no luck in simulating this circuit and what I get is only the negative part of the input voltage and the PWM
seemingly having no effect.

It could be that I just sketched the design wrong, but I would love to hear your guys input about the simulation/idea/configuration.
It looks like you've got two MOSFETs in series with each other in such a way that they both need to be on at the same time for any voltage to pass through the load. Since their gates are being driven by signals that are always in opposite states, the MOSFETs are never both on at the same time. That's why your applied PWM signal is having no effect.

What's the purpose of the two MOSFETs? What are you trying to achieve? Could it be done with just one MOSFET?

As for the negative voltages getting through, that's because each MOSFET has a built in "body diode" which conducts when drain source voltage is negative. You can't block AC wth MOSFETs in this way, cause they'll only control current in one direction.
 

ebeowulf17

Joined Aug 12, 2014
3,291
Doh! I see what's going on now. You need the two MOSFETs in order to get the required AC blocking ability, but you need to flip the orientation of one of them, then drive them with the same signal, not inverted signals. Haven't built one of these myself, but I think I get the idea now. I found a good diagram in a TI app note:

18C0752E-3ECE-4B09-87BF-BA052B54F3CF.jpeg

http://www.ti.com/lit/ug/tiduc87a/tiduc87a.pdf
 

Thread Starter

LmanYan

Joined Jul 2, 2018
2
Doh! I see what's going on now. You need the two MOSFETs in order to get the required AC blocking ability, but you need to flip the orientation of one of them, then drive them with the same signal, not inverted signals. Haven't built one of these myself, but I think I get the idea now. I found a good diagram in a TI app note:

View attachment 156255

http://www.ti.com/lit/ug/tiduc87a/tiduc87a.pdf
Thank you for the replay.Yes, after posting this thread I came across this design too, it seems that that's what they were going for in the original thread.

That is indeed the desired simulation result, but the circuit has few differences that I would be so thankful if you could explain:
First of all, I can see that you have flipped one of the MOSFETs and connected the same PWM signal to their gates, and their sources to the ground, all of that makes sense to me, What I am unsure of are these two things:
-The left side of the circuit is a PWM signal fed through a transformer with coupling caps to the MOSFETs in order to simulate the driver(?).
-You connected a resistor with a diode in parallel, in parallel to the ground for protection purposes(?).

And again, thank you both so much for your replies.
 

Bordodynov

Joined May 20, 2015
2,995
The sources of transistors are not grounded (galvanic isolation). The left signal source simulates an amplified pulse-width signal (with an impedance of less than 50 ohm). This may be a shutter driver.
To exclude the bias of the ferrite core, I applied capacitors at the input and output. In this case I took a ferrite torus with a diameter of 5 cm and a section of 5x5 mm ^ 2. It will be more convenient for him to wind a winding of 100 and 200 turns. A resistor and a diode shift the constant component of the control gate voltage to a positive. As a diode, I used a zener diode, which protects against a breakdown gate dielectric.
I calculated for two frequencies of the voltage source - 50 Hz and 10 kHz. This transformer is designed to control a signal with a frequency of at least 1 kHz.
The best option:
2018-07-16_14-32-18.png
 
Top