• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 48 subscribers
  • Views 290 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.

ADS7953 input abnormal ripple

Ted Zhang
Intellectual 2280 points

Part Number: ADS7953
Other Parts Discussed in Thread: ADS7953

HI Team,

My customer found that with ADS7953,  when input voltage near “0V” there will be a ripple on the input as below picture. However for other channels non"0V" input the ripple goes away. Will get more details however just want to check if any first thought on the potential cause from team?

Thanks. 

  • 0
    TI__Intellectual 2280 points

    HI team, 

    Attached more detail info ADS7953 Questions.docx

  • 0 in reply to Ted Zhang
    TI__Mastermind 25070 points

    Hello Ted,

    Is this on a PCB, or does this set up include wired connections? 

    I find it odd that Channel 0,1 and 4,5 showed this, but channel 2,3 which are in between the two pairs did not. Are the sources between Channel 0,1 and 4,5 some how coupling with each other somewhere in the set up?

    The customer mentioned removing the GPIO functionality, but none of the affected channels are connected to the mux. 

    Ask the customer to change one aspect at the time to isolate and find the source. 

    When the input at ch0 is low, does it show the ripple when the ADC is not sampling? meaning is the ripple only present when the ADC is sampling?

    Regards

    Cynthia

  • 0 in reply to Cynthia
    TI__Intellectual 2280 points

    HI Cynthia,

    Thanks for feedback. It's test on the PCB not wired connection. 

    The CH0,1,4,5 which show this because no input voltage present on the input channels, they try to add some voltage then issue recovered. 

    And customer try to remove the connection between the input and ADC channel, which means ADC is directly pull low with a pull down resistor, and it can still see the issue. As mentioned in the file, that if the channel is not sampling, then the ripple recover as well.

    See case 3 that they only sample the CH0, you can see that CH1,4,5 ripple all recovered.

    Any further comments?

  • 0 in reply to Ted Zhang
    TI__Mastermind 25070 points

    Ted,

    Having such affects on the input voltage while ADC is sampling can happen, in particular if the input source does not have a driving buffer. This is due to the switch capacitor circuit that the ADC uses.

    If the input settles to the expected voltage within the acquisition time, then this should be fine. The customer can add a capacitor to the analog input to show a decrease in the ripple. The capacitor will help to charge the ADC switch capacitor circuit. 

    On a side note, make sure no inputs are left floating, this will make them susceptible to noise 

    Regards

    Cynthia

  • 0 in reply to Cynthia
    TI__Intellectual 2280 points

    HI Cynthia,

    We found the root cause is that there are other channels input over the limits out of the Vref range and after we ensure it within limits and the issue recovered. Any comment on this?

  • 0 in reply to Ted Zhang
    TI__Mastermind 25070 points

    Hello Ted, 

    This discussion has been taken offline. To close this thread and provide a comment. 

    This device has a mux inside, all the inputs are connected to the same mux, though only one is connected to the ADC at a time.

    When one analog input is over AVDD, protection diodes will turn on inside the device. This will then drive all the input voltages high.

    If a channel is left floating, for example, the voltage at that channel pin will follow the diode voltage. If a channel has an op amp connected to it, it will see a slight increase in voltage.

    The analog input voltage should not be greater than the device power supply, this is violation of the abs max. If this is a possible fault condition, then a protection scheme should be implemented

    Regards

    Cynthia