• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 841 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.

ADS7952, First sample is offset with 30mV

Mattias Ericson
Prodigy 30 points
Other Parts Discussed in Thread: ADS7952, OPA192

We are using an ADS7952 to measure a few voltages. We have noticed a behavior where the input signal is offset with 30mV in the first sample. Consecutive measurements are correct.


We have buffered the MUX with the following circuit.

We are measuring the same channel multiple times (using auto 1 mode). Ch4 (green) shows the CS signal and Ch1 (yellow) shows the output from the MXO pin. The input signal to the channel is 0V during the whole series of samples.

The first sample shows about 30mV offset. The other samples shows the correct value, but the MXO pin bounce back to the 30 mV offset for about 80 us after the sampling.


If we stop reading values from the ADC, the MXO output will return to show this offset after about 200ms. Meaning the next time we sample, we will have the offset again.

 In this example, we have only sampled one channel, but we see the same problem if we sample different signals using the mux. The first value will always be offset and incorrect.


We are running the ADS7952 with 5V and a reference of 2.5V. Inputs are set to use a 5V range.

What is causing this offset?

/Mattias

  • 0
    Prodigy 30 points

    Images where missing in the last post.

  • 0 in reply to Mattias Ericson
    TI__Guru 54566 points

    Hello Mattias,

    Can you please provide an schematic that shows the circuitry or sensors driving the Chx inputs, REF input and also the supply connections/bypass capacitors? What is the output impedance of the circuits (or sensor) driving the CHx inputs?

    The MUXO has a 5pF capacitance associated with the pin (Figure 59); in addition to the input capacitance of the LT6220. This capacitance will tend to store charge of the previously scanned channel. If the impedance of the circuit or sensor driving the CHx inputs is high; it is possible that settling issues occur depending on the allowed acquisition time and the specific sequence used.   The datasheet recommends restricting the Rsource CHx impedance to less than 500 Ohms for 12-bit settling at 1-MSPS.  

    Please let us know the max ADC sampling rate that will be used in the application and the SCLK frequency; and the program register settings for Auto-1 Mode used.  

    Thanks and Regards.

    Luis

  • 0 in reply to Luis Chioye
    Prodigy 30 points

    Hi Luis,

    For these measurements we have grounded the input pin, so the impedance is basically 0 ohm.

    Here is the full schematic. The sensors all have low impedance outputs but in this case we have used a channel that we simply have grounded.

    The sampling rate is about 300 samples/second the SCLK is running at 100kHz.

    The following data sequence is used during the setup and conversion.

    // Setup
    0x8000
    0x0020
    0x2C40
    // Conversion
    0x2000
    The text above was written before we found the solution. I have left it in this post as a reference.
    As a test, we ordered the opamp used in the datasheet. This opamp has much lower input bias current. We can't reproduce the same offset using this opamp. So using OPA192 instead of LT6220 solves our problem.
    /Mattias