The problem that the MOS tube cannot be turned on normally in the half-bridge drive (LM5105) circuit

NPU_LB

The problem that the MOS tube cannot be turned on normally in the half-bridge drive (LM5105) circuit [Copy link]

 

 

Hello, senior teachers:

I used LM5105 to make a half-bridge inverter circuit. The schematic diagram of the circuit is as follows. I want to output a 14V square wave voltage between ports COIL_A and COIL_B to drive the load.

Figure 1 Schematic diagram of half-bridge inverter

I found two problems during the test. The first problem is that there is a driving voltage between the GS poles of the two tubes, but the tubes cannot be turned on normally. I used an oscilloscope to measure the GS and DS voltages of the upper and lower tubes. The specific waveforms are as follows:

Figure 2 Upper tube GS voltage

Figure 3: DS voltage of upper tube

Figure 4: Lower tube GS voltage

Figure 5 DS voltage of lower tube

If it works normally, there should be a square wave with a peak value of 28V between the upper tube DS, but judging from the waveform, the upper tube is not fully turned on. I would like to ask the teachers which part this may be the problem.

The second question is: the voltage at the midpoint of capacitors C17 and C20 is not half of the bus voltage. The bus voltage is fully divided on C17, and the voltage of COIL_B to ground is 0. This is not consistent with the theory. COIL_B should be the reference voltage for the half-bridge output. If there is no voltage here, even if the MOS tube is turned on normally, the half-bridge circuit will not work properly. Teacher, is it because of the capacitor selection problem or is this voltage division scheme not actually feasible?

My knowledge is limited and I have been trying hard to come up with a solution. I hope the teacher can give me some guidance.

maychang

[The second question is: the voltage at the midpoint of capacitors C17 and C20 is not half of the bus voltage. The bus voltage is fully distributed on C17, and the voltage of COIL_B to ground is 0. ]

Under what circumstances is this measured? I mean: is there an input pulse signal at the PWM when measuring?

maychang

When measuring the GS voltage and DS voltage of the upper tube, where in the circuit is your oscilloscope ground clip clamped?

NPU_LB

maychang posted on 2023-9-22 10:00 When measuring the upper tube GS voltage and the upper tube DS voltage, where is your oscilloscope ground clip clamped in the circuit?

Hello, I used a differential probe to measure, so the probe was directly connected to the GS and DS of the tube.

NPU_LB

maychang published on 2023-9-22 09:58 [The second question is: the voltage at the midpoint of capacitors C17 and C20 is not half of the bus voltage. The bus voltage is fully distributed on C17. COIL_B is to ground...

I measured it when there was a PWM input signal, and the voltage was divided by the upper capacitor.

maychang

NPU_LB posted on 2023-9-22 10:47 I measured it when there was a PWM input signal, and the voltage was divided by the upper capacitor

The DC power supply voltage falls on the upper capacitor, which is the result of the upper tube not being fully turned on. However, the reason why the upper tube is not fully turned on cannot be seen from the oscilloscope waveform. The waveform between the upper tube GS is basically normal.

There is no problem with the first circuit.

Since the waveforms between the upper tube GS are generally normal, it can be suspected that the upper tube is damaged.

NPU_LB

maychang posted on 2023-9-22 10:55 The DC power supply voltage falls on the upper capacitor, which is the result of the upper tube not being fully turned on. However, the reason why the upper tube is not fully turned on cannot be seen from the oscilloscope waveform...

Thank you, teacher, for your answer. The maximum voltage allowed between the GS of this MOS tube is ±20V. From the measured waveform, the ringing peak of the square wave exceeds 20V. The tube is indeed likely to be damaged. I now want to increase the gate drive resistance to reduce the ringing, and then replace a new tube to try. In addition, you said that the voltage falls on the upper capacitor, which is the reason why the upper tube is not fully turned on. I don’t quite understand this. What I understand is that no matter how the circuit works, the voltage division on the two capacitors should be basically unchanged, because the two capacitors are currently connected to the bus. Can you help explain it again?

maychang

NPU_LB posted on 2023-9-22 11:24 Thank you, teacher, for your answer. The maximum voltage allowed between the GS of this MOS tube is ±20V. From the measured waveform, the ringing peak of the square wave...

[You said that the voltage falling on the upper capacitor is the reason why the upper tube is not fully turned on. I don’t quite understand this.]

The reason why the upper tube is not fully turned on is that the voltage falls on the upper capacitor, not that the voltage falls on the upper capacitor, which is the reason why the upper tube is not turned on.

The upper tube is not turned on, but the lower tube is turned on. When the lower tube is turned on, the current flowing through the lower tube can only be supplied by the upper capacitor. In other words, after the lower tube is turned on, the DC power supply charges the upper capacitor through the lower tube and the load, and the lower capacitor discharges at the same time. However, when the upper tube is not turned on, the DC power supply cannot charge the lower capacitor through the upper tube and the load.

maychang

Suddenly I saw a resistor R14 connected between the drain of the upper tube and the positive of the DC power supply, and the resistance value is quite large. I missed it just now. What does it mean to connect this resistor here? With such a large resistance, how can the upper tube be turned on? There is not much current passing through the upper tube, which is the same as if the upper tube is not turned on.

NPU_LB

maychang posted on 2023-9-22 11:58 [You said that the voltage falls on the upper capacitor, which is the reason why the upper tube is not fully turned on. I don’t quite understand this] The upper tube is not fully turned on because the voltage falls on the upper capacitor...

I understand, teacher. Thank you. I will sort out my thoughts again. I am currently testing it with no load.

NPU_LB

maychang posted on 2023-9-22 12:02 Suddenly I saw a resistor R14 connected between the drain of the upper tube and the positive of the DC power supply, and the resistance value is quite large. I just missed it. What is this resistor connected here...

Teacher, this 20M resistor has been short-circuited before the test. It was disconnected in order to debug other circuits first.

maychang

NPU_LB posted on 2023-9-22 12:14 I understand, teacher. Thank you. I will sort out my thoughts again. I am currently testing it without any load.

[I am currently testing it without load]

Note the reply on the 9th floor. R14 is here, and there is no current even if the upper tube is turned on.

不亦心

It is getting harder and harder to find pleasing schematic diagrams. The LZ's diagram is better than most diagrams in the forum, but I still don't want to read on. Alas.

soso

Bu Yixin posted on 2023-9-22 13:04 It is getting harder and harder to see pleasing schematic diagrams. LZ's diagram is better than most diagrams in the forum, but I still don't have the desire to read on. Alas

What are the requirements for a pleasing schematic diagram?

NPU_LB

maychang posted on 2023-9-22 12:18 [Currently I am testing it with no load] Pay attention to the reply on the 9th floor. R14 is here, and there is no current even if the upper tube is turned on.

You are right. I short-circuited R14 with tin during debugging. Sorry for not explaining it in advance.

NPU_LB

Bu Yixin posted on 2023-9-22 13:04 It is getting harder and harder to see pleasing schematic diagrams. LZ's diagram is better than most diagrams in the forum, but I still don't have the desire to read on. Alas

I try to draw as neatly as possible, firstly so that I can see clearly, and secondly because I have a bit of obsessive compulsive disorder.

NPU_LB

The problem has been solved after replacing two MOS tubes. Thank you, seniors.