Class D Measurements

[adailton]

Hi,

I want to measure my class D amplifier and I made THD analysis and transient analysis. (THD= 0.001%)
But I didn't make AC analysis, LtSpice gave me -300 dB and absurd phase graphs. Is it normal?

Second question : Are there any analysis for class D?(psrr, snr etc.)

Edit: I couldn't find frequency responce analysis for Ltspice, help!!

[Aucosticraft]

Check this link.Cheap simple class D amp circuit to build.

[Eugene Dvoskin]

You would not be able to get results from AC analysis directly.
Things that you can do:
1. Create average model and analyze it in AC analysis.
2. Make transient analysis with different frequencies and analyze input/output by checking waveforms or using .MEASURE postprocessing command.

Creating realistic average model could be quite complicated. While this method is widely used in simulation of switching power supplies, I would not recommend it for audio.
I'm preparing THD analyzer for LTSPICE, that will be capable of plotting harmonics versus frequency and harmonics versus amplitude. Your question brought up a reason for including frequency and phase response into it. It will be ready in a few days and I'll post the link.

[adailton]

Originally Posted by Eugene Dvoskin

You would not be able to get results from AC analysis directly.
Things that you can do:
1. Create average model and analyze it in AC analysis.
2. Make transient analysis with different frequencies and analyze input/output by checking waveforms or using .MEASURE postprocessing command.

Creating realistic average model could be quite complicated. While this method is widely used in simulation of switching power supplies, I would not recommend it for audio.
I'm preparing THD analyzer for LTSPICE, that will be capable of plotting harmonics versus frequency and harmonics versus amplitude. Your question brought up a reason for including frequency and phase response into it. It will be ready in a few days and I'll post the link.

Best thing is waiting for you Average model isn't simulation models? Because, I made simulation and transient results is good. I don't understand why AC analysis doesn't work.

[Pafi]

I don't understand why AC analysis doesn't work.

AC analysis uses a linear model. A comparator is absolutely not linear. The whole ClassD is based on nonlinear operation.

[Pafi]

This is an example for linearisation (making a linear approximation) of a ClassD amplifier:

http://audiodiyers.hu/download/file.php?id=435

[Eugene Dvoskin]

Originally Posted by Eugene Dvoskin

You would not be able to get results from AC analysis directly.
Things that you can do:
1. Create average model and analyze it in AC analysis.
2. Make transient analysis with different frequencies and analyze input/output by checking waveforms or using .MEASURE postprocessing command.

Creating realistic average model could be quite complicated. While this method is widely used in simulation of switching power supplies, I would not recommend it for audio.
I'm preparing THD analyzer for LTSPICE, that will be capable of plotting harmonics versus frequency and harmonics versus amplitude. Your question brought up a reason for including frequency and phase response into it. It will be ready in a few days and I'll post the link.

Link to my THD analyzer. It is also capable of simulation Gain/Phase and impedance during transient simulation.
If somebody is interested, discussion is here.