1
\$\begingroup\$

I'm trying to build a half H-bridge using an IR2104 driver. I want to use it for driving "high power" things such as a brushless motor (100V, 40A), or high power LED (200W).

But it doesn't work as expected: When I increase the "power" voltage (50V on the schematic) to more than 10V, the voltage at the output of the half H-bridge (Uout) doesn't increase to more than 10V. I measured the voltage between the gate and the source of the IGBT Q1, there is nearly no voltage difference (0.3V). Q1 and Q2 are FGH40N60. The Arduino is powered by an external 12V power supply, so the Vin pin of the Arduino also outputs 12V.

On the input side of the IR2104, I have an Arduino Uno which sets IN (pin 5 on the Arduino) to 0V (LOW) and pulses -SD (pin 3 on the Arduino) 5 microseconds HIGH and 1 microsecond LOW in the main loop.

Can somebody help me to increase the output voltage to be nearer to the power supply voltage?

The code on the Arduino is:

void setup() {
  pinMode(3,OUTPUT);
  pinMode(5,OUTPUT);
  digitalWrite(3,HIGH);
  digitalWrite(5,HIGH);
  delay(500);
}

void loop() {
  digitalWrite(5,HIGH);
  delayMicroseconds(5);
  digitalWrite(5,LOW);
  delayMicroseconds(1);
}

And the schematic:

schematic

\$\endgroup\$
1
  • 1
    \$\begingroup\$ There are many, many things that could be checked. Can you start by providing a scope trace of IN, Lo, Ho and Vs, and ideally VB? \$\endgroup\$ Dec 18, 2020 at 23:00

2 Answers 2

1
\$\begingroup\$

Do you understand how the IR2104's bootstrap diode and capacitor are supposed to work?

C1 is intended to be charged up to 12V and then floated up to be between Q1's gate and source so a 12V source-referenced voltage is available to turn on Q1.

The diode prevents the C1's charge from dumping back into the supply when it is floated up and it's negative terminal is at a voltage higher than GND.

But C1 has to be charged from somewhere, and it has to be charged periodically. Are you aware that Q2 has to turn on frequently and periodically in order to refresh the charge on C1? Only when Q2 conducts does current flow from 12V to GND through D1 and C1 to GND in order to charge up C1.

At startup, Q2 must be turned on before Q1 in order to pre-charge C1. Then during operation Q2 has to periodically turn on to continue to refresh C1 so that Q1 can turn on. Usually Q2 turns on frequently enough during regular operation (like commutating a motor, switching an inverter, or synchronous rectification) of the half-bridge that no special switching of Q2 needs to be done.

\$\endgroup\$
0
\$\begingroup\$

Thanks for the reply.

DKNguyen: No i didn't understand the "mecanics" of this circuit before you explain me (wanted to try before going further). Thanks, I checked and Q2 seems to turn on sometimes (at least the gate is activated) sometimes.

Heath Raftery: I checked what you said. Here is the IN in yellow and SD in green: enter image description here

Here is the HO in yellow and the LO in green: enter image description here

Here is the VB in yellow and the LO in green: enter image description here

I put the "high power" voltage at 20V for the measure provided above. As we can see, VB seems to saturate at 12v instead of going to 20v(high power voltage)+12v(voltage provided to the vcc of the ir2104 and used to charge C1). I don't understand why...

I also checked the voltage at bornes of C1 (mass of the scope at VS and probe at VB) and it seems to be charged at 12v continously. enter image description here

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.