• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 30 subscribers
  • Views 499 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.

PCM1808: Current capabilities of VREF pin

Martin Klang
Prodigy 80 points
Part Number: PCM1808

Hello,

it is not clear to me from studying the datasheet if the VREF pin on the PCM1808 can be used as an output, to externally bias the input signal.

I understand that the inputs are internally biased to VREF and that the preferred solution is to AC couple the signals to allow them to self centre.

However, in my application I would like to attenuate by 6dB with a voltage divider, which then needs to be referenced to VREF. It's a single supply design, so even if I put the voltage divider before an AC coupling capacitor I will still need to bias it to 2.5V / VREF.

So I'd like to know: how much current can VREF source/sink? What is its impedance? If I shouldn't use it for bias, can you recommend some other input attenuation method?

  • 0
    TI__Mastermind 23050 points

    Hi Martin,

    The Vref pin is not intended as an output, this pin is provided to allow you to properly decouple the reference voltage for performance. I would not recommend connecting this pin to anything else.

    PCM1808 has a 60kOhm input impedance, so one option would be to place a 60kOhm resistor in series with your input and then the ADC inputs will be -6dB from your input signal. The input impedance will vary a bit with process, typically up to +/-30%. This will introduce some uncertainty in your gain, but this can be calibrated out if it is an issue, we see many people take this approach.

    Best,

    Zak

  • 0 in reply to Zak Kaye
    Prodigy 80 points

    Thank you Zak, your quick answer is much appreciated.

    It's a nice, simple solution but +/-30% seems like a lot. If I've got it right (gain = 60k+-30% / ( 60k+-30% + 60k) ) it would mean that the input might vary from dBFS = 5.3Vpp to 7.3Vpp, something like 2.75dB gain uncertainty.

    I'm not sure how we could calibrate that out. Any suggestions?

  • 0 in reply to Martin Klang
    TI__Mastermind 23050 points

    Hey Martin,

    Keep in mind the PCM1808 has single-ended inputs and the input range is only 3Vpp, so with 6dB attenuation you should never be inputting larger than 6Vpp, and taking the worst case into account with the input impedance variation you probably don't want to input larger than 5.3Vpp to avoid clipping.

    It may not be necessary to calibrate depending on how you are using the device. Audio devices are often not designed for absolute accuracy since this error in gain could be reasonably accommodated with a change in volume. But if you wanted a 1Vrms signal for example to output roughly the same code between multiple devices you would need to measure this reference level and apply some digital gain/attenuation. The gain error of the PCM1808 is 3-6% FSR without this impedance variation, so you would probably need to do this anyways if you need channels to be tightly matched. Most of the time it is not necessary to match the channels as long as you avoid clipping.

    If you wanted to avoid increasing the gain error, you could use an inverting op amp attenuator with the non-inverting pin biased to the appropriate voltage to set your midpoint.

    Hope this helps!

    Best,

    Zak