Oscilloscope observation of Class D power amplifier

jkairup

Oscilloscope observation of Class D power amplifier [Copy link]

 

I currently have two Class D amplifier circuits with the same 1kHz sine wave input signal output as shown in the figure

Why can only half of the sine wave be observed in Figure 1?

Is it because there is no filtering circuit in Figure 2 that the output is a typical square wave graph of Class D?

Question: How can we observe a complete sine wave? How can a Class D amplifier observe a sine wave?

Thanks

maychang

Please post the electrical schematic and mark the points you measured.

Otherwise, no one will know where in the circuit you are measuring the waveform.

jkairup

maychang posted on 2021-6-17 19:47 Please post the electrical schematic and mark the points you measured. Otherwise, no one will know where in the circuit you measured the waveform.

The load is a 4 ohm speaker with one end of the output terminal suspended and one end of the load connected to the output positive terminal and the other end grounded.

jkairup

jkairup posted on 2021-6-17 20:03 The load is a 4 ohm speaker. One end of the output is suspended. One end of the load is connected to the output positive and the other end is grounded

The probe measures the load RL across the

maychang

jkairup posted on 2021-6-17 20:03 The load is a 4 ohm speaker. One end of the output is suspended. One end of the load is connected to the output positive and the other end is grounded

The chip seems to have two output pins, marked OUT+ and OUT-. Apparently this is a bridge output.

What about the power pins of the chip? Is a single power supply or dual power supplies used?

chunyang

jkairup posted on 2021-6-17 20:03 The load is a 4 ohm speaker. One end of the output is suspended. One end of the load is connected to the output positive and the other end is grounded

Your circuit is not a Class D circuit and cannot work under Class D output.

chunyang

The original poster should strictly follow the circuit application in the device manual. If you want to use it yourself, you should fully understand the relevant circuit principles and parameter ranges before trying it. As far as the original poster's circuit is concerned, some amplifiers compatible with Class D and Class AB can be connected in that way, but that connection is conditional and can only work in Class AB state. From the waveform, the negative half cycle is cut off, indicating that the circuit connection is incorrect. This situation in traditional Class AB amplifiers often means that the negative power supply voltage is incorrect, so Maychang mentioned dual power supplies specifically. However, today's low-power Class D amplifiers, including Class AB compatible Class D amplifiers, are powered by a single power supply. At this time, clipping can only mean that the circuit is wrong.

jkairup

chunyang posted on 2021-6-17 20:38 The original poster should strictly follow the circuit application contained in the device manual, and try it on the basis of fully understanding the relevant circuit principles and parameter ranges. As for the original poster...

This is the case. This power supply is a single power supply.

I have tried before, clamping OUT+ and OUT- at both ends of the load RL respectively, and a sine wave was produced in someone else's laboratory. However, with exactly the same board and connection method, I can only produce the upper half here. I plan to take a signal generator and an oscilloscope to someone else's place to try again next time.

Now the problem is that the requirement is to connect only one end of OUT+ or (OUT-) to RL through a polarized capacitor, and the other end of the load RL is connected to GND. And it is a single power supply. How to experiment with this? At present, my experiment is as shown in the post. Only the upper half

jkairup

maychang posted on 2021-6-17 20:11 The chip seems to have two output pins, marked OUT+ and OUT-. Obviously this is a bridge output. What about the power pin of the chip? It uses...

This is the case. This power supply is a single power supply.

I have tried before, clamping OUT+ and OUT- at both ends of the load RL respectively, and a sine wave was produced in someone else's laboratory. However, with exactly the same board and connection method, I can only produce the upper half here. I plan to take a signal generator and an oscilloscope to someone else's place to try again next time.

Now the problem is that the requirement is to connect only one end of OUT+ or (OUT-) to RL through a polarized capacitor, and the other end of the load RL is connected to GND. And it is a single power supply. How to experiment with this? At present, my experiment is as shown in the post. Only the upper half

chunyang

jkairup posted on 2021-6-17 22:34 It is like this. This power supply is a single power supply. I have tried OUT+ and OUT- before. They are clamped at both ends of the load RL. In other people's laboratories, it is...

It is recommended that you read and understand the device manual carefully first, as the circuit cannot be connected as you want.

maychang

jkairup posted on 2021-6-17 22:35 It is like this. This power supply is a single power supply. I have tried OUT+ and OUT- before. They are clamped at both ends of the load RL. In other people's laboratories, it is...

"I have tried before, clamping OUT+ and OUT- at both ends of the load RL, and a sine wave can be produced in someone else's laboratory."

This is the correct drain mode. This class D power amplifier chip uses a single power supply and the output stage is full-bridge operation.

"But I can only get the top half of the board to work with the same connection method."

The Class D power amplifier chip may be partially damaged, or you may have connected it incorrectly.

maychang

jkairup posted on 2021-6-17 22:35 It is like this. This power supply is a single power supply. I have tried OUT+ and OUT- before. They are clamped at both ends of the load RL. In other people's laboratories, it is...

"The problem now is that the requirement is to connect only one end of OUT+ or (OUT-) to RL through a polarity capacitor, and the other end of the load RL to GND."

It is recommended to read the chip manual carefully, as the chip may not allow the load to be connected in this way.

1nnocent

The load is connected between OUT+ and OUT- respectively. The waveform is measured using two probes to measure OUT+ and OUT- respectively. Then, the calculation function of the oscilloscope is used to subtract the signal between the two channels.

1nnocent

jkairup posted on 2021-6-17 22:35 It is like this. This power supply is a single power supply. I have tried OUT+ and OUT- before. They are clamped at both ends of the load RL. In other people's laboratories, it is...

If others can measure it, it means your measurement method is wrong. Use two probes to measure OUT+ and OUT- respectively (the premise is that RL needs to be connected to both ends of OUT+ and OUT-), and then subtract the waveforms of the two channels, which is the waveform at both ends of RL.

jkairup

Hello, thank you for your answer. According to what you said, the probes are connected to both ends of the load RL at OUT+ and OUT- respectively. The sine wave signal obtained by the oscilloscope subtraction function is shown in the figure below.

Question: 1: This method uses the oscilloscope's built-in calculation function. Is the real signal a sine wave in actual use? If so, how do you measure its distortion? Because if the output signal is directly connected to a distortion meter, there will be no sine wave signal.

How do you view this subtraction signal?

2: I have used this board and the same connection method in other people's labs before (OUT+ and OUT- are connected to both ends of RL respectively , but only one probe is used, the probe positive is connected to OUT+ (one end of RL), and the probe ground is connected to the other end of RL, and a complete sine wave is measured. Because only one probe is used, there is no signal subtraction ). Why is this?

maychang

jkairup posted on 2021-6-18 11:48 Hello, thank you for your answer. According to what you said, the probes are connected to both ends of the load RL at OUT+ and OUT- respectively, and the results are obtained by the subtraction function of the oscilloscope...

"Why is this?"

This means that your Class D amplifier chip does not allow the load to be grounded at one end and connected to the output pin at the other end. The load can only be connected to the OUT+ and OUT- pins respectively.

jkairup

Is there any Class D amplifier chip that supports grounding one end of the load? I haven't found any

Please recommend

chunyang

jkairup posted on 2021-6-18 15:36 Is there any Class D amplifier chip that supports grounding one end of the load? I haven't found any here. Please recommend

As mentioned before, there are Class D amplifier chips that are compatible with Class AB mode, but if you connect it like you did, it can only work under Class AB.

jkairup

chunyang posted on 2021-6-18 17:11 As mentioned before, there are Class D amplifier chips that are compatible with Class AB mode, but the connection you made can only work in Class AB.

So can this method of grounding one end of the load only be used with OTL circuits?

chunyang

jkairup posted on 2021-6-19 21:24 So can this method of grounding one end of the load only be used with an OTL circuit?

That may not be the case. The load of OCL is also connected between the output and the ground. However, this is not what you should care about. There are multiple standards for distinguishing the types of audio amplifiers. What you emphasized before was Class D amplifiers, which are not in the same category as OTL. Just like a person, when you describe his weight, you should use "kilograms" and never "meters", although "meters" can also be a parameter of a person.

As for the OTL circuit, it can be either class AB or class D, but most of the current class D amplifier chips have BTL output. As for whether the one you are using can be connected in OTL mode, read the device manual carefully.