MOSFET Gets Hot in H Bridge

[gauravkothari23]

Hi all.
I tried making a H Bridge Circuit using P Channel and N Channel Mosfet and driving the mosfets using transistors. but the P Channel mosfet gets extremely hot within a second,
can anybody please let me know where the problem is.

 

[baileychic]

You can put voltmeters across

1. drain and source of each mosfet and
2. across collector and emitter of each transistor

in Proteus and see which mosfets and transistors turns On at the same time.

From this you can understand what is wrong.

 

[KlausST]

Hi,

A similar discussion a couple of weeks ago.
https://www.edaboard.com/showthread...0-IC-PROTUES&p=1656458&viewfull=1#post1656458.

Klaus

 

[baileychic]

What are you trying to drive from the H-Bridge? Motor? If yes, what is the voltage and current rating of the motor?

 

[wwfeldman]

the source of the N channel FETs is at 0V
the gate of the N channel FETs is at 12V or 0V, depending on the state of the transistor
therefore the N channel FET is ON or OFF depending on the transistor, depending on the drive signal

the source of the P channel FETs is at 24V
the gate of the P channel FETs is at 12V or 0V, depending on the state of the transistor
therefore the P channel FET is ON all the time

when Q1 is ON: Q6 conducts through the load, as it should AND Q2 conducts with no load (no current limit) through Q1.

When Q1 turns OFF, and Q5 turns ON, Q1 and Q6 change roles.
so the P channel FETs are ON all the time, with no limiting device (nominally) 50% of the time

connecting R1 and R5 to 24V instead of 12V may help

there may still be a common conduction time in the timing between when Q2 and Q1 are switching
and likewise Q6 and Q5, but it will be during the transitions and not all the time

 

[gauravkothari23]

the source of the N channel FETs is at 0V
the gate of the N channel FETs is at 12V or 0V, depending on the state of the transistor
therefore the N channel FET is ON or OFF depending on the transistor, depending on the drive signal

the source of the P channel FETs is at 24V
the gate of the P channel FETs is at 12V or 0V, depending on the state of the transistor
therefore the P channel FET is ON all the time

when Q1 is ON: Q6 conducts through the load, as it should AND Q2 conducts with no load (no current limit) through Q1.

When Q1 turns OFF, and Q5 turns ON, Q1 and Q6 change roles.
so the P channel FETs are ON all the time, with no limiting device (nominally) 50% of the time

connecting R1 and R5 to 24V instead of 12V may help

there may still be a common conduction time in the timing between when Q2 and Q1 are switching
and likewise Q6 and Q5, but it will be during the transitions and not all the time

How can a P channel mosfet stays ON all the time.
to completely ON the Mosfet i need atleast 0V and to OFF it i needs atleast 10V at gate.

The reason behind using the H bridge circuit i drive the Inductive load at high frequency of approx 300 Khz, and then convert it to AC.
THe IC's which are readily available in the market works for not more than 25 Khz.

 

[baileychic]

If Gate voltage is lower than Drain then it will turn On.

For example Drain is 24V and Gate is 12V then MOSFET will turn On. There is a difference of 12V (24V - 12V = 12V).

and maybe that is sufficient to turn On your chosen Mosfet.

Ate you driving a transformer from H-Bridge to make something like a 24V to 220V Inverter?

You need Full-Bridge Circuit and not H-Bridge. Diagonally opposite Mosfets in the bridge has to turn On at a time.

You need FET Drivers to drive the Mosfets.

 

[gauravkothari23]

If Gate voltage is lower than Drain then it will turn On.

For example Drain is 24V and Gate is 12V then MOSFET will turn On. There is a difference of 12V (24V - 12V = 12V).

and maybe that is sufficient to turn On your chosen Mosfet.

Ate you driving a transformer from H-Bridge to make something like a 24V to 220V Inverter?

You need Full-Bridge Circuit and not H-Bridge. Diagonally opposite Mosfets in the bridge has to turn On at a time.

You need FET Drivers to drive the Mosfets.

got your point. Thanks
i just need 24V AC output, but only the thing is i need high frequency of appox 300 Khz.
will L298 Work for it, or can you suggest any other IC's or circuit

 

[baileychic]

Refer datasheet for device operating frequency.

24V from 24V means no transformer in your design as load but what is the inductive load? Can it operate at 300 kHz?

 

[gauravkothari23]

i have came across one of the article
https://tahmidmc.blogspot.com/2013/05/debugging-bridge-tips-to-successfully.html
will this help

- - - Updated - - -

i need to drive a inductive load

- - - Updated - - -

Refer datasheet for device operating frequency.

24V from 24V means no transformer in your design as load but what is the inductive load? Can it operate at 300 kHz?

300 Khz is not fixed, but approx

 

[baileychic]

You need SPWM to get Sine wave AC output from Bridge.

Can be done but what max load current you want?

300 KHz frequency will need a very fast processor for the timer interrupt code which updates the PWM duty to generate SPWM.

I guess PIC or AVR 8-bit can't do it.

 

[gauravkothari23]

max load current is around 6 Amps.

- - - Updated - - -

software is not an issue for me. will try my best for the processor to give output of 300 Khz.

 

[baileychic]

Is it AC sine wave output that you want? If yes, is it 300 kHz sine wave output?

In Inverters the PWM of SPWM will be say 7 kHz to sometimes 300 kHz but the output sine wave will be 50 or 60 Hz.

 

[gauravkothari23]

I need a 300 Khz AC square wave output.

 

[baileychic]

Here it is a quick circuit that I made. Not sure if 8-bit PIC can generate 300 kHz PWM. Maybe you need dsPIC to generate a 300 kHz PWM (complementary) to drive the FET drivers.

I have a few more questions.

Can you tell me more about the load? What is the actual load? because you might not even need a H-Bridge to drive the load.

If you just need a 24V DC square wave that varies between 0V and 24V ther you just need a single mosfet to drive the inductive load.

Or do you want a square wave varying between +24V and -24V?

 

[gauravkothari23]

Here it is a quick circuit that I made. Not sure if 8-bit PIC can generate 300 kHz PWM. Maybe you need dsPIC to generate a 300 kHz PWM (complementary) to drive the FET drivers.

I have a few more questions.

Can you tell me more about the load? What is the actual load? because you might not even need a H-Bridge to drive the load.

If you just need a 24V DC square wave that varies between 0V and 24V ther you just need a single mosfet to drive the inductive load.

Or do you want a square wave varying between +24V and -24V?

Its a pure inductive load ( Insulated copper winding on copper pipe)with max current consumption of 6 Amps.
I need +24V and -24V

 

[baileychic]

From your circuit in post #1 you use 24V DC and 0V as GND and so I thought the output signal varies between 0V and 24V.

Edit:

My circuit works for your requirement if the GND connection in the Full-Bridge circuit is replaced with -24.0V power supply.

Build the circuit on PCB with proper track widths for the Load Connections and test it.

I only tested with 300 Hz and 3 kHz in Proteus because PWM frequency above it overloads the simulation and simulation runs very slow.

You need a oscilloscope to check the output signal with first a resistive load to confirm output signal quality.

- - - Updated - - -

The output voltage is not going up to +24V. It is somewhere near 18V. Not sure why.

 

[FvM]

The output voltage is not going up to +24V. It is somewhere near 18V. Not sure why.

Unsuitable high gate series resistors. Use 10 to 50 ohms for smooth turn-on.

 

[baileychic]

That was a mistake. I changed them to 10R but the output voltage goes to max 9V now.

HOx and LOx voltgaes are fine and they are 36V and 12V respectively.

HOx = 24V + 12V = 36V and LOx = 12V same as VCC of IR2112.

What is causing the output high voltage to drop by 4V?

- - - Updated - - -

No, I had measured HOx and LOx voltages when Bridge was connected between +24V and GND.

When Bridge is connected between +24V and -24V the HOx voltage is 44V and LOx voltage is 12V.

 

[KlausST]

Hi,

I recommend to increase the value of C1, C4 to be at least the value of C2, C5.

Imagine: every time when the bootstrap capacitors C2, C5 are charged, the energy needs to come from C1, C4.
Thus the voltage of C1, C2 will immediately drop to the value of C2, C5 .... then the rest of the enery needs to come from the supply, but this is slow, because of stray inductance .. causing relatively high impedance.

Especially at high duty cycles the bootstrap capacitors are discharged for a long time, but they are charged within a short time.
Expect high voltage ripple at C1/C4, and the bootstrap capacitors maybe are not charged fully. This also may explain your low output voltage at high side.

Klaus