Hello,I am driving a three phase bridge rectifier which control the variation of speed of an asynchronous motor(Induction three phase motor) of 5.5KW the PWM signal used to drive the bridge is generated by a PIC18F46K22 where each output pin is giving 0.2V to 2.087V (with the Analog input of the potentiometer when adjusted );the driver I am using is IR21362 which requres 5V of input in order to give a good controlling signal of the output,could you please let me know how that voltage of the PIC can be increased such that I can find the +5V to input to the driver(IR21362) so that the IGBTs can operate effectively?

The IGBTs I am using are IRG4PC50FD

Please help.

 

Is PIC18F46K22 5 volt system?

I don't know how you are looking at the PWM pins to see 0.2-2.087 volts, but... There is a trick that some people use PMW for. As you know PWM signal is a square wave where the length of high and low section can be adjusted. However, some people use PMW not as a square wave, but as a dc signal, it like having an extra DAC to convert digital values into analog values. If you look at PMW signal on oscilloscope, you see square wave. However, if you look at PWM signal using dc voltmeter, the dc voltmeter does not see the square wave, it sees sort of "average" value of the wave which is sort of dc signal. Neat trick if you are all out of DAC to use.

Which brings us to your problem. If your controllers don't see 5 volts, but see 2.087 and therefore do not activate because they are not receiving proper TTL signals, a couple of things come in mind:
1. PWM rise time. Since you are dealing with square wave, the rise should be fast so you can actually see the "square" nature of the signal.
2. Oscillations on the horizontal sections of the square wave.

Bottom line. I think it would be a good idea to look at your PWM signal on the oscilloscope to see how square it actually is.

 

A scope would also show whether your measured "2.087" is the average or peak signal level.
If you're running the PIC on a 5V supply then peak output levels of only 2.087 would suggest the outputs are too heavily loaded.

 

It is normal as stated to see less than pic supply on a PWM output, pretty much verified by the seeing .2v to 2.087 over the rpm range, also depends on what PWM % that represents, you would see it at 100% PWM but you cannot go to 100% PWM using a charge pump type IC as you would get no drive to the 'top' IGBT.
Max.

 

A scope would also show whether your measured "2.087" is the average or peak signal level.
If you're running the PIC on a 5V supply then peak output levels of only 2.087 would suggest the outputs are too heavily loaded.

Put a buffer (using op amp) between the PWM pin and IR21362?

 

Put a buffer (using op amp) between the PWM pin and IR21362?

With the buffer using LM324 solve the problem?
Please help

 

It is normal as stated to see less than pic supply on a PWM output, pretty much verified by the seeing .2v to 2.087 over the rpm range, also depends on what PWM % that represents, you would see it at 100% PWM but you cannot go to 100% PWM using a charge pump type IC as you would get no drive to the 'top' IGBT.
Max.

Dear MaxHeadRoom,what are you suggesting could be used to make the IR21362 operational?find all my project here,in the circuit as there is no IR21362 in proteus I replaced it with 3 IR2104 where one of them input one signal and output two signals to control a half bridge.

http://www.edaboard.com/thread355384-2.html#post1521798

on post #23

Please help

 

Is PIC18F46K22 5 volt system?

I don't know how you are looking at the PWM pins to see 0.2-2.087 volts, but... There is a trick that some people use PMW for. As you know PWM signal is a square wave where the length of high and low section can be adjusted. However, some people use PMW not as a square wave, but as a dc signal, it like having an extra DAC to convert digital values into analog values. If you look at PMW signal on oscilloscope, you see square wave. However, if you look at PWM signal using dc voltmeter, the dc voltmeter does not see the square wave, it sees sort of "average" value of the wave which is sort of dc signal. Neat trick if you are all out of DAC to use.

Which brings us to your problem. If your controllers don't see 5 volts, but see 2.087 and therefore do not activate because they are not receiving proper TTL signals, a couple of things come in mind:
1. PWM rise time. Since you are dealing with square wave, the rise should be fast so you can actually see the "square" nature of the signal.
2. Oscillations on the horizontal sections of the square wave.

Bottom line. I think it would be a good idea to look at your PWM signal on the oscilloscope to see how square it actually is.

Dear Shteii01,how could I set that reference of 2.087V,will the system give good result once done?

see my codes and circuit here http://www.edaboard.com/thread355384-2.html

post #21
Please help

 

To ensure your bootstrap Cap is correct see Isolation Bootstrap Calculator | Silicon Labs
Max.