What is "deglitch logic" in class-D amplifier?

deglitch

I found the function block named "deglitch logic" in some class-D amplifier datasheets. Can anyone explain the use of this block?

Thanks a lot!!

deglitch logic

rockycheng said:

I found the function block named "deglitch logic" in some class-D amplifier datasheets. Can anyone explain the use of this block?

Thanks a lot!!

Click to expand...

One of the major issues in the design of Class-D amplifiers is the prevention of shoot-through current as the transistors in the H-bridge structure cannot turn on or off instantaneously. For a short time both transistors in the stack are on, providing
a low impedance path between the power supply and ground.This results in very large current spikes. This not only drastically reduces the power efficiency of the amplifier, but can also lead to output device failure.


Deglitch logic prevents large currents from passing through the H-Bridge by implementing a time delay between output modulations.

deglitch circuit

I know the short-circuit current during the transient of switching the output transistors could be eliminated by introducing the dead-time. But I feel that "deglitch logic" is not for that purpose.

de-glitch

rockycheng said:

I know the short-circuit current during the transient of switching the output transistors could be eliminated by introducing the dead-time. But I feel that "deglitch logic" is not for that purpose.

Click to expand...

Hi rockycheng,
Hope the attached article helps. Pls refer page.2, "deglitch logic" part.

shoot through current prevention

Thanks Chang! I've read that student project report. Actually I think that their understanding and statement on it are not very accurate. So don't be mislead by that school report.

If you search around, you'll find that most of the term "deglitch" in the field of class-D comes from the datasheets of TI. I debt the deglitch block in their products are not for avoiding shoot through 'cause that block is controlled by the internal oscillator (pls check TPA2000D4). As far as I know, you don't need that control signal to introduct dead time.

In my opinion, the deglitch logic could be simply implemented by a schmitt triger to clean up the PWM singal. But why using a control signal?

deglitch time

rockycheng said:

In my opinion, the deglitch logic could be simply implemented by a schmitt triger to clean up the PWM singal.

Click to expand...

Hi,rockycheng
Have you noticed the comparator before the deglitch bloch have some hysteresis characteristics. Then we still need a schmitt triger?

d logic amp

Actually the hysteresis in the comparator is used to generate some delay time between the two comparators. There still could be glitches on the PWM signals.

deglitch

rockycheng said:

Thanks Chang! I've read that student project report. Actually I think that their understanding and statement on it are not very accurate. So don't be mislead by that school report.

If you search around, you'll find that most of the term "deglitch" in the field of class-D comes from the datasheets of TI. I debt the deglitch block in their products are not for avoiding shoot through 'cause that block is controlled by the internal oscillator (pls check TPA2000D4). As far as I know, you don't need that control signal to introduct dead time.

In my opinion, the deglitch logic could be simply implemented by a schmitt triger to clean up the PWM singal. But why using a control signal?

Click to expand...

Hi, Rockycheng,
I have communicated with our TI engineer.

They confirmed that it is indeed to avoide the shoothrough current. Below is the explainations from them:
...............
The purpose of the deglitch circuit in the block diagram is to prevent large currents from passing through the H-bridge in the functional diagram. One of the major issues in the design of Class-D amplifiers is the prevention of shoot-through current as the transistors in the H-bridge structure cannot turn on or off instantaneously. During cross-conduction, both transistors in the stack are on, and provide a low impedance path between the power supply and ground. This results in very large current spikes. This not only drastically reduces the power efficiency of the amplifier, but can also lead to output device failure.

Deglitch logic implements a time delay between output modulations and prevents this problem.
...............

Hope it helps.

When both transistors in the stack are on, this results in very large current spikes between the power supply and ground. You can easily see this glitch in the power supply. That's why it is named "deglitch", I guess.

comparator delay classd

Thank you Chang!

ti class d amplifier timer function

got it