Push-pull circuit

cxq742536574

Push-pull circuit [Copy link]

 

Why is the lower arm PNP transistor of the H-bridge push-pull circuit often damaged, and the three legs of the transistor are directly short-circuited?

chunyang

Add a freewheeling diode.

chunyang

Note that the model of the bridge arm transistor should be determined according to the load to meet the withstand voltage and current carrying capacity, and appropriate margin should be left. At the same time, a freewheeling diode should be added to the inductive load. The original poster did not explain the load situation, so it is naturally impossible to confirm whether the transistor selection is reasonable.

maychang

The relay winding is a strong inductive load, and the four transistors in the bridge circuit should be connected with anti-parallel freewheeling diodes.

PowerAnts

maychang posted on 2022-1-18 19:37 The relay winding is a strong inductive load, and the four transistors in the bridge circuit should be connected in anti-parallel with the freewheeling diode.

maychang

PowerAnts published on 2022-1-19 10:05

The current direction in the figure is that the transistors on the upper right and lower left are turned on at the previous moment, but are turned off at this moment. The current direction generated by the self-induced electromotive force on the inductive load. The self-induced electromotive force at both ends of the load is positive on the left and negative on the right.

Consider the transistor on the lower left. The emitter of the transistor is positive to the collector, and there is a positive base current. Therefore, the transistor should be turned on, that is, it has a certain collector current (which may be slightly less than the freewheeling current on the load), rather than the collector current being zero.

A similar analysis can be made for the upper right transistor.

PowerAnts

maychang posted on 2022-1-19 11:41 The direction of the current in the figure is that the upper right and lower left transistors are turned on in the previous moment, and they are all turned off at this moment. The current direction generated by the self-inductance electromotive force on the inductive load...

Yes, the circuit itself has a voltage clamping function, the clamping voltage is VCC+2*0.7V, -2*0.7V

maychang

PowerAnts posted on 2022-1-19 12:46 Yes, the circuit itself has a voltage clamping function, and the clamping voltage is VCC+2*0.7V, -2*0.7V

If the diodes are connected in anti-parallel to each transistor in the bridge circuit in Figure 5, the result will be different. The energy stored in the load will be fed back to the DC power supply.

PowerAnts

maychang posted on 2022-1-19 12:59 If the transistors in the bridge circuit in the figure on the 5th floor are all connected in anti-parallel with diodes, the result will be different. The energy stored in the load will be fed back to the DC power supply.

By adding 4 diode clamps, the energy stored in the load is indeed fed back to the DC power supply. What are the parameters?

The coil current of a matchbox-sized relay is 30 to 50 mA, the inductance is 300 to 400 mH, and the energy storage is less than 1 mJ. VCC probably does not lack this energy.

In the same working state, the conditions of the BJTs at the diagonal of the H bridge are similar. Only the P tube is burned. Focus on checking the source of supply to see if the N tube is a "big chip" and the P tube is a "small chip". If not, these 4 BJTs can be replaced with SOT-89 packages. The price is about 1 cent more expensive, saving the cost of patching 4 diodes.

PowerAnts

Moreover, when the control signals on both sides are 0, the two upper tubes of the H bridge are pulled high by R3 and R7. When one side is driven and turned off, the coil is always demagnetized through one of the upper tubes and the diode connected to the upper tube on the other side. The voltage on both sides of the coil is close to VCC, and the energy storage cannot be recovered.

PowerAnts

The current correction for the 5th floor is as follows: the thin line indicates a smaller base current, and the thick line indicates a larger collector current. It is not clear that 4 diodes must be connected in parallel.

PowerAnts

Supplement: This drive circuit provides bidirectional current and must be a single-coil magnetic latching relay. Only 100-200mS pulse current is required for opening and closing.

Fred_1977

I just passed by and saw that the analysis above is very accurate. He must be an old hand in this field.

maychang

PowerAnts published on 2022-1-19 15:34 Moreover, when the control signals on both sides are 0, the two upper tubes of the H bridge are pulled high by R3 and R7. When one side is driven and turned off, ...

If the control signals on both sides are both low level (or both high level), and both upper tubes (or lower tubes) of the H-bridge are turned on, then the two upper tubes (or lower tubes) should also be connected with diodes in parallel. At this time, the energy stored in the winding is dissipated by circuit loss (tube voltage drop, etc.) and does not return to the power supply.

cxq742536574

chunyang posted on 2022-1-18 17:31 Note that the model of the bridge arm transistor should be determined according to the load to meet the withstand voltage and current carrying capacity, and appropriate margin should be reserved. At the same time, the inductive load needs to add a freewheeling diode. Floor...

Relay drives the motor, the current is about 10A

cxq742536574

maychang posted on 2022-1-18 19:37 The relay winding is a strong inductive load, and the four transistors in the bridge circuit should be connected in anti-parallel with the freewheeling diode.

I have added a tube, but the damage phenomenon cannot be reproduced. I don't know how to verify whether the addition is effective.

cxq742536574

PowerAnts posted on 2022-1-19 12:46 Yes, the circuit itself has a voltage clamping function, and the clamping voltage is VCC+2*0.7V, -2*0.7V

Thank you for your analysis

chunyang

cxq742536574 posted on 2022-1-24 09:50 Relay drives the motor, the current is about 10A

What should be examined here is the relay coil parameters, not the load.

chunyang

cxq742536574 posted on 2022-1-24 09:52 The tube has been added, but the damage phenomenon cannot be reproduced. I don’t know how to verify whether the addition is effective?

This is just the conclusion drawn from the analysis. If the analysis is thorough, there is no need to verify it. Time will give the answer.

cxq742536574

PowerAnts posted on 2022-1-19 12:46 Yes, the circuit itself has a voltage clamping function, and the clamping voltage is VCC+2*0.7V, -2*0.7V

The clamping voltage you analyzed refers to the two ends of the coil, shouldn't it be VCC-0.7V, 0.7V?