• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 393 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

PCM1690: DC-Coupled OPA-Output Filter with PFFB on Akita Amplifier Family

Lukas Jani
Prodigy 160 points
Part Number: PCM1690
Other Parts Discussed in Thread: OPA1662, TPA3251, PCM4104

Hello all, 

i'd need a little reassurance if what i'm trying to design will actually work:

A PCM1690 differential DAC is used for conversion from a DSP. Using an OPA1662 (please ignore the Part in my LTSpice simulation below, in the actual circuit an OPA1662 will be used instead of the LT1124), the suggested DC-Coupled Differential to Single Ended Conversion Filter from the Datasheet is implemented with a little higher Gain to compensate the PFFB-Gain losses.

Afterwards the PFFB & Input Components for the TPA3251/Akita-Amps are placed. To use the TPA-Amps in normal BTL Mode, CH1 and CH2 of the DAC connected to Input A and B, the signal of CH2 will be inverted in the DSP, hence a normal differential signal is seen from the actual amp. 

The reason i do not want to 1:1-connect the positive and negative output of one DAC channel is that i want to configure SE or BTL Mode in the dsp. If my understanding of the DAC is correct, it should be easily possible as described above. (given the Mode-Pins on the Amps are connected correctly)

Furthermode, by using the OPAmps without and Bias (0V Centered, +-12V supply), it must be possible to eliminate the AC-Coupling Caps directly on the output of the DAC itself, the OPAMP will simply cancel out the +2.5V DC-Bias on each channel and the DC-Offset going to the Amp is 0.

I'd be glad if someone could confirm my statements here or correct me if i made any mistakes. Simulation Modell can also be provided if helpful.

Thanks alot,

best regards!

  • 0
    TI__Mastermind 22550 points

    Hello Lukas,

    the model that you have above seems okay to me and it is fine and should work. I also want to remind you we have SPICE/TINA models for most ( if not all of) our amplifiers in their product folders.

    The PCM1690 has eight channels . Note that VOUT1+ and VOUT1- together compromise the signal of Channel one- which could be converted to single ended ,  similarly VOUT2+ and VOUT2- together  make the signal in  channel 2 and can  be converted to single ended . Now , if I need a stereo signal,  for example, then I can use these two single ended  signals as my 2 channel stereo. Within this concept you can use these signals. I tried to have a pictorial description of the above in this drawing. 

    You can also look at the application diagrams for PCM4104 ( a 4-channel DAC) in its datasheet such as figure 21 and 22 for other possible configuration of filtering SE or DIff .

    Regards,

    Arash

  • 0 in reply to Arash Loloee
    Prodigy 160 points

    Hello Arash, thanks for the answer.

    I'm quite familiar with LTSpice, that is why i keep using it. Most circuits can be abstracted to a very well working LTspice Modell if the designer actually understands the circuit.

    I think i do understand your drawing, thanks the elaboration.

    To summarize my intent:

    • CH1 of the DAC (+ & -) are converted to Single Ended, then fed to the AMP at Input A
    • CH2 of the DAC (+ & -) are converted to Single Ended, then fed to the AMP at Input B
    • The Amp is used in BTL Mode
    • The Audio Signal on CH1 and CH2 is the same, CH2 Signal is inverted to then - after SE-Conversion of both Channels - again have differential signal

    --> this effectively makes it only a 4CH DAC with differential signals, or an 8CH DAC with SE Signals. The Amplifier System consisting of two TPA3251 Chips can therefore be used as 4CH BTL or 8CH SE without resoldering or hardware changes to the audio path. 

    This only is possible if between CH1 and CH2, the time-delay/fitter of the signal is nearly zero. (or at least several magnitudes lower than 1/20khz). That would work right?

    Best Regards,

    Lukas

  • +1 in reply to Lukas Jani
    TI__Mastermind 22550 points

    Hello Luke,

    Thanks for reply, originally I thought you want to use  one output of each channel and send those  to input of A & B of the amp as diff signal. Thanks for clearing it up.

     Although the data sheet has only the delay within a channel speced and we do not have any data of delay mismatch b/w channels, I don't expect that to be significant. So your plan should be ok.

    Regards,

    Arash

  • 0 in reply to Arash Loloee
    Prodigy 160 points

    Hey Arash, thanks for your assessment. I'll probably just try it out and measure a few samples if there is anything to see :)