• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 46 subscribers
  • Views 316 views
  • Users 0 members are here
Support feedback
Options
Options

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.

ADS1292: Is ripple on digital supply coupling to output data?

Julian Sandrock
Prodigy 20 points
Part Number: ADS1292
Other Parts Discussed in Thread: TPS62162

We use the ADS1292 in our design and supply it with 3.3V generated from the TPS62162 DC/DC converter (for low power reasons we can only use a DC/DC).

Because our overall current consumption at 3.3V is 2 mA max, the TPS is using low fequency PFM instead of high frequency PWM. This results in a supply ripple of something like 6 mV with 8 kHz (sawtooth ripple due to inductor current rise and fall), see attached screenshot.

My first instinct was to simply filter the analog supply of the ADS with an RC lowpass (e.g. 47 Ohm + 22uF). This results in a decently low noise analog supply.

But in this articel by TI employee Bryan Lizon a practical example is shown (in figure 5) how noise at the digital supply can also couple straight into the output data.

This results in my question: Do I need to filter the ripple on the digital supply voltage of the ADS1292 as well? Will the sawtooth ripple with 6 mVpp on DVDD be visible in the output? I think I can't just use the same RC filter for analog and digital supply, because the "large" 47 Ohm resistor would result in additional digital noise beeing coupled to the analog supply. The better solution might be to use an additional LDO like the TPS7A2030 to generate 3V from the 3.3V supply with surpressed ripple and supply both AVDD as well as DVDD with this 3V?

But if your estimation would be that 6 mV of ripple at the digital supply are no problem for the ADS1292 I would like to avoid the additional LDO.

  • 0
    TI__Mastermind 19815 points

    Hi,

    I will try to get back to you 2/28.

  • 0 in reply to ChienChun Yang
    TI__Mastermind 19815 points

    Hi,

    Directly use/connect DCDC Switching power supply to ADS1292 is not recommended nor to any precision ADC device.

    --------------------------------------------------------------------------------------------------------------------

    Is the "RC lowpass (e.g. 47 Ohm + 22uF)"  passive or active filter?

    Did you add more noise decoupling capacitor at the output of the DCDC switching power supply?

    We are not DCDC switching power supply department, so maybe you could contact the DCDC switching power supply department for further advice to reduce the noises or ripples.

    --------------------------------------------------

    Is your Switching Mode DCDC smaller than the one used on the evaluation kit/board(EVM)?

    TPS73033DBVR

    https://www.digikey.com/en/products/detail/texas-instruments/TPS73033DBVR/705344

    Is it a battery powered product?  Do you want to try battery + LDO?

    You may experiment/test this setup by using a bench top power supply to provide a battery equivalent voltage(e.g. 3.3V) and then connect to a LDO.

    Or, you can try test with battery directly without LDO, but make sure to example the power ripple and signal integrity by some experiments/tests/data collection&analysis&V&V before making final decision.

    Thanks

  • 0 in reply to ChienChun Yang
    Prodigy 20 points

    Hey, thanks for getting back.

    Yes, I'm aware that a DCDC is not great for supplying a precision ADC, but from an efficiency standpoint we not really have an alternative.

    The device is supplied by 2-S Li cells, so input voltage ranges from ~6.5V to ~8.4V. We can't source more than 0.8mA to supply both the ADS as well as the host controller for readout and filtering of the signal with 3.3V. So an LDO is not really an option.

    Adding more decoupling to the DCDC output is also no going to help a lot in this case, as output current is so low. This means the DCDC is not using a high frequency PWM, but a fixed puls length PDM with a pulse length of something like 200ns. Only when the output voltage sank low enough again (it seems like that is reached at about -5mV) the TPS triggers the next pulse. So even with a higher output cap the ripple would probably not be reduced that much.

    That's why my idea was to supply the µC directly from the DCDC output and filter only the ADS supply with a simple passiv RC low pass. Because analog supply of the ADS is only 300uA the 47 Ohm drop is no problem. The question just remains, wether this 5 to 10 mV sawtooth ripple is a problem on the DVDD pin of the ADS. I always thought that the digital supply of an ADC is uncritical, so I never had any thought about ripple on the DVDD. But somehow the digital supply seems to be able to couple into the output codes?

    Do you think 5mV on DVDD will degrade performance?

  • 0 in reply to Julian Sandrock
    TI__Mastermind 19815 points

    "input voltage ranges from ~6.5V to ~8.4V. " <--- too high for ADS1292

    datasheet page 6

    ----------------------------------------------------------

    You have to have a steady analog power supply of at least 2.7V~5V and with sufficient current supply -

    The amount of current depending on how many functions and features the product needs/enables. 200mA should be a safe and more than enough estimate.

    It's suggest customer use EVM to do some preliminary tests&evaluations to estimate the current consumption based on the features the product needs.

    -------------------------------------------------------------------------------------

    Do you think 5mV on DVDD will degrade performance?

    Customer will need to do some experiments&tests&data collection&analysis&V&V before making decisions.

    Thanks

  • 0 in reply to ChienChun Yang
    Prodigy 20 points

    Ok I think we are talking past each other. Yes, 6.5V is to high for the ADS. That's where the DCDC comes into play to generate 3.3V for host controller and ADS ;)

    I hoped you had some data on how much impact a small ripple on digital supply has with the ADS1292, but I guess we have to do our own evaluation.

  • 0 in reply to Julian Sandrock
    TI__Mastermind 19815 points

    Hi,

    we don't have data for " data on how much impact a small ripple on digital supply has with the ADS1292".

    Customer will need to do some experiments&tests&data collection&analysis&V&V before making decisions.

    Thanks