Generating uni-polar PWM signal using IR2113

[Ng Jing Xi]

Hi,

Currently, I am stuck with generating a uni-polar PWM signal to drive the IGBTs using IR2113.
As in, I connected the schematic attached below but the inputs to the IR2113 waveform seems to be quite weird.
Is there anything wrong with the schematic design?

 

[alec_t]

The TL072 opamp is for audio frequencies and has a rated gain-bandwidth product of only 3MHz. Its gain at 80kHz will be inadequate. Also, the schematic shows it as an integrator but there are no R or C components to support it ??

 

[ronsimpson]

Usually I would use a voltage comparator not an op-amp.
Watch out many voltage comparator need a pull up resistor on the output.

 

[Ng Jing Xi]

The TL072 opamp is for audio frequencies and has a rated gain-bandwidth product of only 3MHz. Its gain at 80kHz will be inadequate. Also, the schematic shows it as an integrator but there are no R or C components to support it ??

I am sorry, actually I used TL072 as a comparator. What do you mean by gain at 80kHz is inadequate?

 

[Ng Jing Xi]

Usually I would use a voltage comparator not an op-amp.
Watch out many voltage comparator need a pull up resistor on the output.

But why is the main difference between voltage comparator and an op-amp

 

[ronsimpson]

opamp is built to amplify while a voltage comparator compares two voltages as indicates which is higher. Exactly what you are doing.

As to the question on 80khz gain:
A 3mhz opamp really does not work well at 3mhz. It maximum gain is 1. At 300khz is max gain in 10. At 30khz its gain is less than 100.

 

[alec_t]

.... and at 80kHz (your carrier frequency) the max gain would be only ~ 37, so the TL072 output would look more like a ramp voltage than the sharp vertical edge desirable from a comparator. The ramp voltage would result in your IGBT turning on/off relatively slowly and hence dissipating a lot of power and getting very HOT.

 

[Ng Jing Xi]

Hi,

Actually I tried using LM339 with pull-up resistors, voltage comparator, for the generating of PWM signal instead of TL072. However, the waveform that I got is quite weird. [ Ramp up on the rising edge, straight down on the falling edge, Refer to LM339(1) & LM339(2) ]

Therefore, I changed back to TL072 which gave me the same results as previous week. [Refer to TL072(1) & TL072 (2)]
After I connect my TL072 to the IR2113, I measured the waveform at the Hin and Lin Pin of IR2113.
Which is the part I am currently stuck at. [Refer to Hin.jpg]

 

[audioguru]

No schematic.
If the comparator has a high value pullup resistor and the circuit was built on a breadboard with stray capacitance all over the place then the rising edge of an 80kHz rectangle wave will be nice and slow as you show.

Your Hin signal shows that the 50Hz "sinewave" is clipping like crazy maybe because the IR2113 has digital inputs that cannot be used for an analog sinewave.
Your other circuit produces the PWM signal to feed the IR2113 driver.

 

[Ng Jing Xi]

No schematic.
If the comparator has a high value pullup resistor and the circuit was built on a breadboard with stray capacitance all over the place then the rising edge of an 80kHz rectangle wave will be nice and slow as you show.

Your Hin signal shows that the 50Hz "sinewave" is clipping like crazy maybe because the IR2113 has digital inputs that cannot be used for an analog sinewave.
Your other circuit produces the PWM signal to feed the IR2113 driver.

Hi, this is the exact schematic design I connected on my breadboard. The output of my triangular wave and sine wave are prefectly fine but the driver circuit is giving me a headache.

I tried to measure the waveform at the driver input and the comparator (U2 & U3). On some of the period, this is the waveform I captured. [Refer to Hin & Lin from Comparator.jpg]
However, on some of the other period, this is the waveform I captured. [Refer to clipping.jpg & clipping (2).jpg]

 

[audioguru]

Your input voltages to the IR2113 might be destroying it. The Hin and Lin are LOGIC voltages that must NEVER go below ground (negative).
Your "comparator" outputs go to +10.5V and -10.5V. The -10.5V must not go negative.

 

[Ng Jing Xi]

Your input voltages to the IR2113 might be destroying it. The Hin and Lin are LOGIC voltages that must NEVER go below ground (negative).
Your "comparator" outputs go to +10.5V and -10.5V. The -10.5V must not go negative.

But isnt it supposed to be like this? If not how can I rectify this problem?

 

[audioguru]

A REAL comparator is usually powered from +Vcc and ground (with no negative supply), because its output is usually +Vcc and ground (a logic signal). Then it can directly feed the logic inputs on the IR2113.

You can still use an opamp as a "comparator" and feed its "logic" output to the IR2113 if you use a single supply for the "comparator" and use an input offset voltage for it.
Simply use coupling capacitors to feed the sine and triangle waves to the inputs of the comparator and bias the inputs at half the positive supply voltage.
There might be a problem when the supply is turned on where the inputs of the "comparator" go negative for a moment until the coupling capacitors charge.
This problem can be avoided by powering all the opamps from a single positive supply and biasing them at half the supply voltage.

 

[Ng Jing Xi]

A REAL comparator is usually powered from +Vcc and ground (with no negative supply), because its output is usually +Vcc and ground (a logic signal). Then it can directly feed the logic inputs on the IR2113.

You can still use an opamp as a "comparator" and feed its "logic" output to the IR2113 if you use a single supply for the "comparator" and use an input offset voltage for it.
Simply use coupling capacitors to feed the sine and triangle waves to the inputs of the comparator and bias the inputs at half the positive supply voltage.
There might be a problem when the supply is turned on where the inputs of the "comparator" go negative for a moment until the coupling capacitors charge.
This problem can be avoided by powering all the opamps from a single positive supply and biasing them at half the supply voltage.

Seems like I fixed the problem about the Hin and Lin. However, I am not getting any output when I measured Ho and Vs.

 

[audioguru]

Seems like I fixed the problem about the Hin and Lin.

How? What did you do?

However, I am not getting any output when I measured Ho and Vs.

Then maybe your IR2113 is destroyed by the input voltages that went negative.

 

[Ng Jing Xi]

How? What did you do?

Then maybe your IR2113 is destroyed by the input voltages that went negative.

Actually wait, I didnt fixed it.
If I were to use LM339, How do I do it? Is this schematic able to produce the PWM signal that is going to drive the IR2113?

Or if I were to use back TL072, what modification do I need to be able to feed the signal into IR2113 and to prevent my IR2113 from being destroyed?

This is the drafted schematic for Lm339.

This is quite urgent! :x

 

[audioguru]

Your entire schematic is WRONG.
Every IC has +12V which is good but they also have -12V which might destroy the IR2113.

Change all -12V to GND. Bias the + inputs of the opamps and comparators to half the supply voltage (+6V).
The 10k output pullup resistors on the comparators have a value too high. The LM339 has a pretty good saturation with an output current of 6mA so use an output resistor value of 12V/6mA= 2k ohms.

You have correct pin numbers for one of the comparators in the LM339 but wrong pin numbers for the other three comparators.

 

[Ng Jing Xi]

Your entire schematic is WRONG.
Every IC has +12V which is good but they also have -12V which might destroy the IR2113.

Change all -12V to GND. Bias the + inputs of the opamps and comparators to half the supply voltage (+6V).
The 10k output pullup resistors on the comparators have a value too high. The LM339 has a pretty good saturation with an output current of 6mA so use an output resistor value of 12V/6mA= 2k ohms.

You have correct pin numbers for one of the comparators in the LM339 but wrong pin numbers for the other three comparators.

If I were to change the other comparators from -12V to GND, my waveform will be clipped at the negative cycle. [E.g Sine wave]
How do we bias the inputs of the opamps to half the supply voltage?

You can ignore the pin no. for LM339 cause I am just using a Ideal op-amp model.

 

[alec_t]

How do we bias the inputs of the opamps to half the supply voltage?

Like this:-

 

[audioguru]

The frequency and unclipped output level of the triangle wave will be changed since the total supply voltage is half what it was before.