# opamp controlled inverter questions

#### l0vot

Joined Apr 29, 2013
107
I am trying to use a simple OPAMP multivibrator/oscillator that generates a square wave 17V signal to drive a MOSFET inverter.
I have run into a problem: the OPAMP can only drive one of the two MOSFETs at a time because attempting to drive both simultaneously results in the signal wires from the OPAMP being shorted across the positive rail of the power supply the MOSFETs are supposed to be inverting, this is because i need the MOSFETS do be driven with opposite waves, the easiest way to do this is by wiring the OPAMP output to the gate of one MOSFET and the ground to the drain, and then wire the ground to the gate of the other MOSFET and the OPAMP output to the drain. I tried using diodes to split the wave into it's positive and negative counterparts and drive one MOSFET with one half, and the other with the other half, but since the gate forms a capacitor and the other half of the signal isn't there to rapidly drain that capacitor, then the MOSFETs don't produce a nice square wave like i need them to to function efficiently. I tried adding resistors between the drain and gate, but that only worked at low frequencies. I tried using a little ferrite transformer to isolate the OPAMP, but it draws too much current for the OPAMP to handle.

Is there a way to drive the MOSFETs directly with one OPAMP multivibrator? If only MOSFETs and IGBTs were 4 terminal devices...

#### Ramussons

Joined May 3, 2013
1,264
Can you please put down ... so many words.... to a schematic?

#### l0vot

Joined Apr 29, 2013
107
this isn't a full schematic, but this is basically the power inverter circuit
Can you please put down ... so many words.... to a schematic?
this isn't a full schematic, but it's basically the inverter circuit I'm trying to drive:

the signal source is an OPAMP in a square wave multivibrator configuration I got from here: http://www.electronics-tutorials.ws/opamp/op-amp-multivibrator.html

I'm trying to drive the inverter with the OPAMP multivibrator, the MOSFETS need to be driven opposite of one another otherwise it doesn't work, and they need the full square wave instead of half of it because at the frequencies I'm trying to drive them they don't turn off fast enough without the full wave.

#### crutschow

Joined Mar 14, 2008
29,815
because at the frequencies I'm trying to drive them they don't turn off fast enough without the full wave.
So what is that secret frequency?

#### Ramussons

Joined May 3, 2013
1,264
this isn't a full schematic, but this is basically the power inverter circuit

this isn't a full schematic, but it's basically the inverter circuit I'm trying to drive:View attachment 125518

I'm trying to drive them they don't turn off fast enough without the full wave.
If the MOSFET's dont turn off fast enough, you will need to reduce your frequency of switching. Else, you will have a situation where both the FET's are ON and conducting.

That aside, use 2 more OPAMPS as a Buffer, one with a Inverting output and the other with a Non Inverting output to drive the FET's.

Better would be to use a D Flip Flop and use the Q and Q to drive the FET's

#### l0vot

Joined Apr 29, 2013
107
So what is that secret frequency?
frequency is variable, probably from 30HZ to as high as I can get the MOSFETS to go without the on/off delay becoming a problem, or 50KHz, whichever comes first

#### l0vot

Joined Apr 29, 2013
107
If the MOSFET's dont turn off fast enough, you will need to reduce your frequency of switching. Else, you will have a situation where both the FET's are ON and conducting.

That aside, use 2 more OPAMPS as a Buffer, one with a Inverting output and the other with a Non Inverting output to drive the FET's.

Better would be to use a D Flip Flop and use the Q and Q to drive the FET's
the MOSFET gate basically acts as a capacitor, so it has a pretty long turn off delay as the gate "leaks down", and while it's in between "on" and "off" it's generating a bunch of heat because that's how transistors work, in order to get one to shut off quickly, the gate must be discharged, or reverse charged. One half of the wave saturates the gate as fast as possible to flip it "on", the other half switches the gate "off" as fast as possible, also the reason for using a square wave since the switching time is the shortest for a high voltage square wave (in this case up to +-20V).

I'll try the buffer idea, sort of didn't want to use more OPAMPs, but I do have them on hand, and I don't have flip flops. I may get some flip flops and mess with them since they seem like they might be able to solve a different problem I was going to try and solve some other time. Thanks for being patient with me, this is basically my first project that involves any of these components, besides caps and inductors.

#### Sensacell

Joined Jun 19, 2012
3,001
Nobody uses opamps to drive MOSFET gates in switching applications- they are just too slow.

To get fast switching - use a dedicated gate driver chip, they can drive the gates hard to achieve fast and efficient switching.

#### l0vot

Joined Apr 29, 2013
107
What you are describing is a Royer Oscillator, very common in low cost inverter circuits. Search for that and you will get dozens of schematics to learn from.

https://en.wikipedia.org/wiki/Royer_oscillator

https://images.search.yahoo.com/sea...EEc2VjA3BpdnM-?p=royer+oscillator&fr2=piv-web

ak

I can't seem to find anywhere where the circuit is explained in enough detail for me to design my own out of parts I have on hand, the whole reason I started out trying to use an opamp oscillator to control things is because the opamp oscillator was well documented.

I have some IRF3205s, some inductors, some shotkey diodes, some resistors including 100K POTs, a variable voltage power supply, and a breadboard. I need a square wave oscillator suitable for driving a much larger MOSFET or IGBT based inverter (depending on the voltage) without a bunch of switching losses. The information is around here somewhere, probably on this website in fact, but I can't seem to find it.

#### l0vot

Joined Apr 29, 2013
107
Nevermind, this is clearly not going to work, so I'm going to attack this from a different angle, PWM controller KA3525A is able to do exactly what I want, and it only needs some resistors and a cap to work, I figured this out by examining a cheap inverter, which has an operating frequency close to what I need, and it utilizes two IRF3205s. It will take a bit of messing around to get a solid handle on this chip, but it's versatile, and cheap enough to use for other projects as well.

#### Dodgydave

Joined Jun 22, 2012
10,316
Yes use a deadicated chip like the KA/SG3525 or KA7500( TL494) much easier and stable. If you want super simple try a L6565.