• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 30 subscribers
  • Views 1145 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPA6130A2 input configuration, SN drops 20dB when I direct connect?

Gordon Rankin
Prodigy 70 points
Other Parts Discussed in Thread: TPA6130A2, TPA6120A2

Gang,

I have a Wolfson differential dac running at 4.7V's on the analog. I have I2C control working fine and all the volume etc working fine.

I have connected the TPA6130A2 using a separate 3.3V supply and cap coupling (0.47uF on all 4 inputs) and the results on the analyzer are better than expected. The sound on the other hand sucks.

So I took that design and direct coupled the TPA6130A2 to the Wolfson dac and put a 4V regulator in place of the 3.3V for the TPA and the sound was great but the signal to noise ratio dived taking 20dB off what I had before.

So today I dropped the 4V separate regulator and used the same output of the 4.7V for the Wolfson and the results were still really bad.

I have 1vac rms going in on all the tests.

THD for direct is about 0.22 @ 1KHz 50%, volume = 1.01

THD for cap coupled is 0.05 @ 1KHz 50%, volume = 1.01

SNR for direct coupled I would put at like maybe 90dB

SNR for cap coupled I would put at about 115dB

Any ideas? I am running out of them.

Thanks in advance,

Gordon

  • 0
    TI__Genius 16645 points

    Hello Gordon:

    Can you let me know the di-electric type of the 0.47uF caps you are using? Ceramic? X7R, Y5V

    Also, can you send me a .pdf of the schematic that you are using?

    Is the output of your Wolfson DAC ground referenced?

    Rgds,

    Fred

  • 0 in reply to Fred Shipley
    Prodigy 70 points

    Fred,

    Thanks for the reply!

    I am using Panasonic ECP film caps for the 0.47uF. These are 16v units, also sourced by CD, data sheets available at Digikey.

    The Analog supply for the dac is run off a National LD3985-4.7 for a 4.7V rail as this design is based on the TAS1020B and powered by VBUS ~=5V. This puts the Vmid of the Wolfson at about 2.25V in testing and that is where the outputs sit.

    The schematic is the direct coupled version, but just replace the 4 wires for the input with the 0.47uF/1206 caps and you have the original design. In the original I was using an MCP1700-3.3V regulator for the TPA6130A. This is a bit better noise wise as compared to the MCP1702. I have also put in a LD3985-4.7 thinking maybe having them identical would make things better but really not.

    Let's see VDD = 3.3v from TAS1020B.

    Thanks

    Gordon

  • 0 in reply to Gordon Rankin
    TI__Genius 16645 points

    Hey Gordon:

    You might be having an issue with out of band noise from the DAC.

    Have you tried a first order RC filter between the DAC and the TPA6130A2?

    Rgds,

    Fred

  • 0 in reply to Fred Shipley
    Prodigy 70 points

    Fred,

    It would not make for a 20-25dB change between cap coupled and direct. The Wolfon is pretty much dead quiet all the way out past my Prism dScope, right at the pin. I did try resistors, ferrites and inductors.

    Really nothing seemed to work except to cap couple the two.

    I really think the TPA6130 falls on it's face if it's input is not at zero volts dc.

    Thanks

    Gordon

  • 0 in reply to Gordon Rankin
    Prodigy 70 points

    Fred,

    Maybe a better idea would be to send you a picture of what I am talking about. Here we are sending the standard digital JTEST for jitter testing. This is an asynchronous Class 1 USB DAC with dual really low jitter Crystek (0.5ps) oscillators going to the Wolfson dac chip and the TAS1020B MCLKI pin for generation of the I2S frames coming out of the TAS1020B.

    The RED is the TPA6130A with the cap coupling, it sits just below the 120dB marker. The WHITE is the direct connect using the 4V regulator for the TPA. The Green is using the National 4.7V regulator, the same kind that is used on the dac chip for the analog output.

    Thanks

    Gordon

  • 0 in reply to Gordon Rankin
    TI__Genius 16645 points

    Hello Gordon:

    I checked the design files and the DC from the DAC plus the input signal may be biasing some of transistors in the input of the device.

    This would account for the degradation with the inputs direct coupled.

    For best performance in your design, I suggest that you use input capacitors.

    If your applicaton requires maximum performance, I suggest that you consider the TPA6120A2.

    Sorry that the news is not better.

    Rgds,

    Fred

  • 0 in reply to Fred Shipley
    Prodigy 70 points

    Fred,

    Thanks, but no room, no volume control... so I guess I have to find some way to make this work or try the National 48822. I have the eval board for that but nobody there could tell me how to use it and the software crashed all the time.

    The engineers did say it should work with differential biased inputs. But who knows....

    Thanks

    Gordon