• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 29 subscribers
  • Views 437 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.

Original question:

EVM PCM4222 DC on output

PCM4222: DC on PCM4222 output and THD+N not up to spec

Thomas Fripon
Prodigy 80 points
Part Number: PCM4222
Other Parts Discussed in Thread: OPA227, , DIT4192, OPA1612,

I am currently designing a prototype analog-to-digital converter with a PCM4222 and DIT4192. The analog front end uses OPA1612 input buffers. As recommended in the datasheet, I'm using OPA227 opamps as VREF drivers. In some areas, the overall design is similar to TI's EVM board, in others it's not. Clocking is done using an FPGA. I'm getting 121dB of dynamic range (noise floor with no input signal, 20-22.000 Hz BW) within the audio band which is according to spec but there is a lot of DC on the output which also was reported in this post by another poster.

With the internal HP filter enabled it hovers around -95~105dB on both channels but when the filter is off, I'm getting -69dB on the left and -76dB on the right channel. I even tried to temporarily adjust the DC offset using a trimmer but I was not able to get rid of it. Quite the opposite actually - it got only worse.

Is there something I can look into to remove the DC from the signal?

Also, the datasheet states -108dB THD+N @ -1dBFS - but no matter what I do, the best I can get at -1dBFS are some measly -92dB. These are mostly higher order harmonics. When measuring the actual analog signal after the input buffers (+2dBu at each VIN+ and VIN-), there's no sign of harmonics at all - so it must be the PCM4222 input stage or conversion adding those.

Any ideas?

  • 0 in reply to Thomas Fripon
    TI__Mastermind 23050 points

    Hey Thomas,

    If you are not AC coupling the input then there will be some DC offset from the input amplifier stage in addition to the offset from the modulator. You say you are using OPA1612, I'm assuming that you are using 2 of these in a differential amplifier configuration, right?

    As you have observed, the measurement bandwidth of 22Hz - 20kHz is critical for recreating datasheet specs. Generally audio ADCs are not trimmed for precision DC performance since most systems tend to be AC coupled. If you need to further reduce the DC offset then you could implement a digital filter downstream in your processor to better remove this.  

    I'm not sure what conditions are resulting in -92dB THD+N, but if you recreate the datasheet test conditions you should not see anything worse than -101dB with a -1dB input. This does require you to band-limit the signal to 22Hz to 20kHz to see this performance. I tested the same condition on the PCM4222EVM and observed a THD+N of -105dB. If you are band-limiting the measurement and still do not see datasheet performance than you may need to revisit your layout and double check that you have sufficient bypass capacitance on the power and reference rails.

    Best,

    Zak

  • 0 in reply to Zak Kaye
    Prodigy 80 points

    Hi Zak,

    thanks for the reply.

    Yes, there is one OPA1612 for each input feeding the differential ADC inputs and they are AC coupled.

    Fun fact: I have implemented the same reset switches as they are on the EVM board schematic and when I press and hold only the ADC reset, the signal itself vanishes from the spectral analyzer but the DC offset stays as long as the button is hold. As soon as the button is released the offset also seems to reset and eventually settles at a new value each time I repeat this procedure.

    I guess, I need to try a digital filter downstream as you suggested.

    With regards to bypass capacitance, my design adheres closely to the recommendations from the datasheet. To further test the THD performance I removed some of the bypass caps near the PCM4222 but that didn't change a thing...

  • 0 in reply to Thomas Fripon
    TI__Mastermind 23050 points

    Hey Thomas,

    Thanks for the update, I'm sorry you're still not seeing the expected performance. Can you confirm that your bandwidth was limited to the 22Hz to 20kHz range for the THD+N measurements? Also, are you calculating the -1dBFS input level from the full scale spec or measuring it from the ADC output to confirm it is at -1dBFS? Noise and harmonics tend to increase as you approach the full scale range and the inputs begin to approach saturation so if you are actually driving a slightly larger signal on the input this may be why you are seeing a little more distortion. 

    Best,

    Zak