After the crossover in each band there is a signal delay module.
This module is necessary for adjusting the synchronicity of the three speakers at the listening point. As a rule, the signal from the LF speaker lags in time behind the signal from the HF and MF speaker. Therefore, if you want to achieve synchronous operation of the speakers at the listening point, you need to introduce a time delay in the MF and HF channel. As they say - to align the acoustic centers of the speakers.
If a phase-linear crossover is used for frequency division, you can adjust the acoustic centers relatively easily by measuring the amplitude phase response of the speaker system at the listening point while observing the step response of the measured amplitude phase response.
After we have adjusted the frequencies and the slope of the crossover, we measure the amplitude phase response at the listening point and observe such a step response, as we see from the picture, three maxima from three speakers are at different distances from each other on the time scale, this means that the signal from three speakers arrives at the listening point at different times.
By introducing a delay in the midrange and high-frequency bands, we see in the measurements how the maxima approach each other.
And at a certain value of delays, we get a step response that has one common maximum.
Also, the inversion of the speaker is clearly visible on the step response, if the speaker is connected in inverse polarity, then its maximum on the step response will be turned down.
The low-frequency delay module is installed for possible correction of the low-frequency channel delay between two speakers during fine-tuning of the joint operation of two three-band speakers.
This module is necessary for adjusting the synchronicity of the three speakers at the listening point. As a rule, the signal from the LF speaker lags in time behind the signal from the HF and MF speaker. Therefore, if you want to achieve synchronous operation of the speakers at the listening point, you need to introduce a time delay in the MF and HF channel. As they say - to align the acoustic centers of the speakers.
If a phase-linear crossover is used for frequency division, you can adjust the acoustic centers relatively easily by measuring the amplitude phase response of the speaker system at the listening point while observing the step response of the measured amplitude phase response.
After we have adjusted the frequencies and the slope of the crossover, we measure the amplitude phase response at the listening point and observe such a step response, as we see from the picture, three maxima from three speakers are at different distances from each other on the time scale, this means that the signal from three speakers arrives at the listening point at different times.
By introducing a delay in the midrange and high-frequency bands, we see in the measurements how the maxima approach each other.
And at a certain value of delays, we get a step response that has one common maximum.
Also, the inversion of the speaker is clearly visible on the step response, if the speaker is connected in inverse polarity, then its maximum on the step response will be turned down.
The low-frequency delay module is installed for possible correction of the low-frequency channel delay between two speakers during fine-tuning of the joint operation of two three-band speakers.
Next, mute blocks are installed, which allow you to quickly turn off the signal passing through the band.
Next, there are modules for adjusting the signal level in the band.
If we need to coordinate the SPL level of the speakers in each band, this is done by entering the required level of signal level reduction in this module. In this project, you can only use a decrease in the signal level, since increasing the signal level will lead to clipping of the signal at maximum input volume.
In order not to overload the DAC input with a high output signal from the ADAU1452, you need to monitor the maximum signal level that the ADAU1452 can produce with a maximum input signal. To prevent accidental overload of the DAC input when setting up the project, a signal level limiter is installed after the amplification modules.
When entering a corrective frequency response, it is not always convenient to obtain a corrective frequency response without amplification. An example of such a case.
From the graphs it is clear that at an input signal frequency of 40 Hz with a level of -10 db (Fs) a signal cut off in level will appear at the DAC input; to eliminate clipping it is enough to reduce the signal level by -10 dB in all three bands using the gain block.
If you want to have a stable result with the entire range of the amplifier's output power, you need to monitor the level of the maximum output signal from the ADAU1452 that goes to the DAC input. If you feed a signal of more than 0 dB to the DAC (removing the level limiter block), then the overload protection will start to work and the speakers may produce sound that should not be there. The amplifier itself will not burn out, but you can hear very unpleasant sound combinations.
Next, there are modules for adjusting the signal level in the band.
If we need to coordinate the SPL level of the speakers in each band, this is done by entering the required level of signal level reduction in this module. In this project, you can only use a decrease in the signal level, since increasing the signal level will lead to clipping of the signal at maximum input volume.
In order not to overload the DAC input with a high output signal from the ADAU1452, you need to monitor the maximum signal level that the ADAU1452 can produce with a maximum input signal. To prevent accidental overload of the DAC input when setting up the project, a signal level limiter is installed after the amplification modules.
When entering a corrective frequency response, it is not always convenient to obtain a corrective frequency response without amplification. An example of such a case.
From the graphs it is clear that at an input signal frequency of 40 Hz with a level of -10 db (Fs) a signal cut off in level will appear at the DAC input; to eliminate clipping it is enough to reduce the signal level by -10 dB in all three bands using the gain block.
If you want to have a stable result with the entire range of the amplifier's output power, you need to monitor the level of the maximum output signal from the ADAU1452 that goes to the DAC input. If you feed a signal of more than 0 dB to the DAC (removing the level limiter block), then the overload protection will start to work and the speakers may produce sound that should not be there. The amplifier itself will not burn out, but you can hear very unpleasant sound combinations.
After the level indicator there is a DC filter and after it the signal goes to the ASRC to convert the signal into the optimal format for the DAC.
In fact, the input of the output ASRC is the end of the structural diagram in the Sigma Studio, the rest of the work is done by the hardware part of the ADAU1452 and the DAC chip.
The outputs on the amplifier board are numbered according to the ASRC channel number. The signal that is fed to input 6 of the ASRC will be present at the amplifier output under the name Ch6.
If desired, you can specify a different output number for the low-frequency band. For example, you can assign a low-frequency band to pin 5. But for this to be possible, you must first assign a different number to pin five, so that number five appears in the list of available free channels.
It should be noted that in this example the left channel of the stereo signal is processed. To process the right channel, you need to switch the conductor to the right channel at this point.
This is the main functional purpose of the blocks of the starting project for an active three-band speaker system with a FIR crossover.
In fact, the input of the output ASRC is the end of the structural diagram in the Sigma Studio, the rest of the work is done by the hardware part of the ADAU1452 and the DAC chip.
The outputs on the amplifier board are numbered according to the ASRC channel number. The signal that is fed to input 6 of the ASRC will be present at the amplifier output under the name Ch6.
If desired, you can specify a different output number for the low-frequency band. For example, you can assign a low-frequency band to pin 5. But for this to be possible, you must first assign a different number to pin five, so that number five appears in the list of available free channels.
It should be noted that in this example the left channel of the stereo signal is processed. To process the right channel, you need to switch the conductor to the right channel at this point.
This is the main functional purpose of the blocks of the starting project for an active three-band speaker system with a FIR crossover.
In the project for a three-way speaker system called I2S 3Way IIR 48kHz, a library crossover is used to divide the bands. All other blocks of the project are the same as the I2S 3Way FIR 48kHz project.
The library crossover allows you to very easily configure the crossover parameters, and it is interactive - all changes in the crossover settings are automatically transferred to the ADAU1452 and are heard in the speaker system if a music signal is fed to the board input.
It should also be taken into account that when using an IIR crossover, the ability to align the acoustic centers of the speakers by step response becomes impossible. When using an IIR crossover, you should monitor the in-phase of the bands; it is impossible to achieve synchronicity in IIR filters. If an IIR filter is needed in the crossover that takes into account the features of the speaker radiation phase, then for each band you will have to use an IIR filter in which you can specify its quality factor and the amount of signal gain. The required IIR crossover filter settings are found using third-party CAD software and speaker measurements.
The library crossover allows you to very easily configure the crossover parameters, and it is interactive - all changes in the crossover settings are automatically transferred to the ADAU1452 and are heard in the speaker system if a music signal is fed to the board input.
It should also be taken into account that when using an IIR crossover, the ability to align the acoustic centers of the speakers by step response becomes impossible. When using an IIR crossover, you should monitor the in-phase of the bands; it is impossible to achieve synchronicity in IIR filters. If an IIR filter is needed in the crossover that takes into account the features of the speaker radiation phase, then for each band you will have to use an IIR filter in which you can specify its quality factor and the amount of signal gain. The required IIR crossover filter settings are found using third-party CAD software and speaker measurements.
In the project called I2S 3Way subtractive 48kHz, an IIR subtractive filter is implemented using a 4th order Butterworth IIR filter.
The principle of operation of such a filter is relatively simple, after the signal has passed through the low-pass filter, this filtered signal is subtracted from the input common signal, for subtraction, an inversion of the filtered signal and a delay unit are required that compensates for the delay of the low-pass filter.
The resulting frequency response can be observed in the probe window after pressing the stimulus button.
The subtractive filter allows you to get a phase-linear crossover characteristic when using IIR filtering. This allows you to use step response when adjusting the acoustic centers of the speakers and reduce the time delay for the crossover compared to the FIR crossover.
The principle of operation of such a filter is relatively simple, after the signal has passed through the low-pass filter, this filtered signal is subtracted from the input common signal, for subtraction, an inversion of the filtered signal and a delay unit are required that compensates for the delay of the low-pass filter.
The resulting frequency response can be observed in the probe window after pressing the stimulus button.
The subtractive filter allows you to get a phase-linear crossover characteristic when using IIR filtering. This allows you to use step response when adjusting the acoustic centers of the speakers and reduce the time delay for the crossover compared to the FIR crossover.
The project called Tube is intended for those who want to get distortions similar to a tube amplifier in their passive stereo system.
In this project, the level of nonlinear distortions on the left and right channels of the amplifier is adjusted from 0.005% to 15% using an external potentiometer.
The design of the project introduces distortions according to this principle:
- When changing frequency, the set level of distortion is almost the same for any frequency;
- As the signal level decreases, distortion increases;
- Even harmonics prevail over odd ones.
In this project, the level of nonlinear distortions on the left and right channels of the amplifier is adjusted from 0.005% to 15% using an external potentiometer.
The design of the project introduces distortions according to this principle:
- When changing frequency, the set level of distortion is almost the same for any frequency;
- As the signal level decreases, distortion increases;
- Even harmonics prevail over odd ones.
The project called Test is designed to obtain speaker characteristics using a microphone and measurement software.
As can be seen from the diagram, the left channel signal is output to output 4 of the board, the signal from the right channel is output to outputs 5 and 6.
As can be seen from the diagram, the left channel signal is output to output 4 of the board, the signal from the right channel is output to outputs 5 and 6.
You specified these parameters in post #2 :
Performance parameter PWM DAC:
Performance parameter PWM DAC:
- Output power: 100 W per channel, with a 32 V supply and R load = 4 Ohms.
- Speaker impedance: 4 - 8 Ohm (2Ohm – max 50W)
The output LPF choke holds 6.5A, with a load of 2 Ohm, the peak power at which the choke will be saturated will be approximately equal to 6.5A * 6.5A * 2R = 84W. The average power will be approximately 84/2 = 42W. According to measurements, the current protection in the TAS starts to operate at an average power of 50W with a load of 2 Ohm.
Simply put, in this case, in this implementation, the limitation is in the saturation current of the choke, and not in the TAS.
Since this board was developed for home active speakers, which usually use 4 Ohm and 8 Ohm speakers, then 2 Ohm, which are usually used in cars, were not a priority.
If I set it to 12A inductance, the dimensions of the board will increase and the price too. I chose smaller dimensions and lower cost.
By the way.
Just today I took measurements from a 2-channel DSP Amp at 200W per channel on TAS5162. It has high-current chokes and these boards can already handle 200W at 2 Ohm.
Below are the measurement results for a 4 Ohm load.
I'm thinking of creating a new branch for these boards, their power is neither here nor there, it's hard to say whether there will be interest in them.
Thanks for the explanation. In my case I have a 3-way with two 4 ohm woofers. So I'm limited to either a 8 ohm (series) or 2 ohm (parallel) load. It seems to me the series configuration is the better way to go.