This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TAS5828M: How to control 2 Amps using Class-H control

Part Number: TAS5828M

Hi team,

My customer is considering two TAS5828M for their BT speaker. One for Tweeter and One for Woofer.

They will use Class-H control. And as far as I know, Class-H senses input signal and generate feedback signal to the external DCDC.

Q1. If you see in customer's configuration as below, they will share I2S input and PVDD power rail. In this case, which amp's HPFB should good to be used? For Woofer, Bass range will be boosted a lot.

If Hybrid pro algorithm only cares input signal, it seems to be fine to select any of them like below. But if Hybrid-Pro algorithm automatically considers EQ and DRCs too, than Isn't it going to be clipping on Woofer amp?

Q2.Does Hybrid-Pro algorithm consider/foresee boosted EQ gain/DRC in order to prevent clipping at output? Or just senses input signal only?

Customer's configuration:

Thanks,

Jay

  • Hi Jay.

    Q2.Does Hybrid-Pro algorithm consider/foresee boosted EQ gain/DRC in order to prevent clipping at output? Or just senses input signal only?

    A: You can take below process flow as a reference.

    As can be seen from the flow chart, the input of the Hybrid-Pro algorithm is the audio signal processed by the internal DSP. In other words, the Hybrid-Pro algorithm takes into account various gains of the internal DSP.

    Q1. If you see in customer's configuration as below, they will share I2S input and PVDD power rail. In this case, which amp's HPFB should good to be used?

    A: Customers need to consider the worst case to avoid clipping of amps that do not use the Hybrid-Pro algorithm. Customers can add an additional comparator or other methods to automatically select the worst output of the two amps and send it to the boost circuit.

    BR.

    Wei Qiu.

  • Hi Wei, I got additional questions to check with you.

    Customers need to consider the worst case to avoid clipping of amps that do not use the Hybrid-Pro algorithm. Customers can add an additional comparator or other methods to automatically select the worst output of the two amps and send it to the boost circuit.

    Q1.Is there a reference design for use of  additional comparator? If no, please let me know which comparator should be used instead.

    Q2. And in this case, input of comparator should be HPFB of two amps, right? Then the two amps should use separate Hybrid-pro algorithms. Am I correct?

    Q3. How other customers's hybrid-pro configuration normally look like if they use 2 Class-H audio-amps? Do they use 2 power rails? or share one PVDD like my customer's config.?

    Thanks,

    Jay

  • Hi Jay

       We can consider enable Open Drain mode to support this kind of using. The Open Drain mode, our GPIO port will only gives out low level and Hi-Z state, won't gives out high level. So two Amp's class-H port can connect together and won't affect each other. The final class-H output will be the longest low level of this two Amp, to create the highest PVDD voltage for both of the Amp. Use under this way, remember to give additional pull up resistor to DVDD for the Class-H GPIO port.

  • Hi Shadow,

    Thanks for your reply.

    Is this example block what you've meant for?

    I'm sorry, I'm a bit confused. In my understanding, Amp's Class-H port generates PWM signal depending on the audio input. And the average voltage for the each amps will be somewhere between 0 and DVDD respectively. If so, Isn't the final class-H output voltage summed and ended up giving undesirable feedback to DCDC? For example, if bass boosted signal comes in, the woofer amp will generate class-H feedback near Low level(Let's say 0.5V). And Tweeter amp will generate near DVDD voltage(let's say 2V). I'm wondering what's happening in this case.

    Thanks,

    Jay

  • Hi Jay

       The example you raised, is the normal mode of our GPIO port. What I said is to enable "Open Drain" mode.

       Seems you are not familiar with the "Open Drain" circuit. I try to explain it here, maybe you can also google it for more information. With open drain, there's no pull up inside of our IC, there's only a MOS act as a switch. So there's only two state our GPIO can gives out under this mode, one is pulled to GND, so it's zero. Another is MOS turn off, the output is Hi-Z state, not high voltage level. For our class-H working, it's still PWM working for this MOS, but of course it can't gives out any high level voltage, it's only keep changing from GND to Hi-Z state.

       You can imagine two device's all set into this "Open Drain" mode, and connect the output port together. Is there any conflict would happen? Maybe not, with only GND or Hi-Z state, two device won't affect each other. Then you can add additional pull up resistor for this connection. Only two IC both give Hi-Z state, the final output will be high level DVDD, any of the device gives out Zero, will make the final output become Zero. The final output will be a new PWM, generate by two device together. 

  • Hi Shadow,

    Thanks for your kind explanation. Actually I understood Open-Drain characteristics but just have been a bit confused. 

    It seems should be okay for both of amps not affecting each other like you said. Since the High level PVDD is decided by Zero output. Lastly, Could you confirm if it is right to follow the FB/RC filter values generated on GUI even though customer is using two amps? I believe it should be fine. 

    Thanks,

    Jay

  • Hi Jay

       I'm afraid enable Open Drain will make the Resistor network different from normal using, especially when you add additional pull up resistor, this value will also affect the final results.

       To calculate the Resistor value is not difficult, our Class-H GPIO output logic is simple. After you set the Min/Max voltage value into the PPC3, when the amp need Min voltage, it will output 100% duty cycle. And when amp needs Max voltage, duty cycle comes to 0%. When in the middle, the duty cycle will change with equal proportions. Using this logic, and work with Buck/Boost device's datasheet, should be able to calculate the resistor network that suit your need.

       If you find any problem, please e-mail to Jess or Qiu Wei for help, we'd look at the detailed design of our customer. They will all back to office tomorrow.