• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 46 subscribers
  • Views 806 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.

ADC12DJ2700: Calibration Modes

Hans
Genius 12865 points
Part Number: ADC12DJ2700

Hello Team,

The ADC has several calibration modes. One important calibration procedure seems to be the offset calibration. The internal occurring DC-offsets (in dual channel mode) should be eliminated with this calibration procedure, am I right?

Can you provide a detailed description on how to use this calibration procedure during the ADC initialization.

The ADC datasheet describes, that there should be no Signal near or at DC. In our device, we can not be sure, that there is no DC-signal at the input of the RX-path.
- Is it really necessary to disable the input signal during the offset calibration?
- Do we have to disable the input signal during the other calibration procedures, too?
- Is it possible to disable the input signals in the ADC?


Thanks and Best Regards,

Hans

  • 0
    TI__Mastermind 44680 points

    Hi Hans

    If used properly the offset calibration mechanism will remove DC-offsets. This is true for both dual-channel and single-channel modes of operation.

    During the linearity and gain calibration the ADC is disconnected from the input signal, so that portion of the calibration is not sensitive to applied input signals.

    During the offset calibration, the entire signal path including the ADC inputs are active. If input signals are applied and have certain characteristics (signals at or near DC or aliased signals that will fall near DC) the offset calibration results will be degraded. For this reason we recommend that offset calibration only be enabled when the input signal is known or can be disconnected. If the device needs to be re-calibrated in the future, and the signal cannot be disconnected then only the linearity/gain calibration should be performed. The results of the prior offset calibration will be maintained.

    See sections 7.4.6, 7.4.7 and 7.4.8 of the ADC12DJ2700 datasheet for more information.

    I hope this is helpful.

    Best regards,

    Jim B

  • 0 in reply to Jim Brinkhurst1
    TI__Genius 12865 points

    Hi Jim,

    regarding offset calibration:

    Does "a known input signal" mean you can connect the inputs (+ and -) to ground?

    Thanks and Best Regards,

    Hans

  • 0 in reply to Hans
    TI__Mastermind 44680 points

    Hi Hans

    Terminating the VINx+/- inputs to GND using either 50 ohm or lower resistance will work.

    It would also be OK to terminate the input of any signal conditioning circuitry upstream of the ADC. If a balun is used, either terminating or disconnecting the signal input to the balun would be sufficient. If an amplifier is used, terminating or disconnecting the input to the amplifier (in a manner that maintains the normal DC offsets at the ADC inputs) would be OK.

    Regards,

    Jim B

  • 0 in reply to Jim Brinkhurst1
    TI__Genius 12865 points
    Hello Jim,
    would it make sense to provide a defined signal an the input of the receiver and calibrate out the offset of the whole signal chain in the digital domain (FPGA)?
    Will this work?
    Will this impact the performance if the offset is not calibrated by the ADC?

    Thanks and Best Regards,
    Hans
  • 0 in reply to Hans
    TI__Mastermind 44680 points
    Hi Hans
    That could be an effective solution as long as the correction algorithm can apply different offset values to the data output by the 2 (dual input modes) or 4 (single input modes) interleaved sub-converters in the ADC. If the algorithm can do that it will also be sensitive to input signals that are very near to DC, Fs/4 (single input modes) and Fs/2 and multiples of those frequencies.
    If the signal can be constrained to avoid those frequencies then I would recommend first using the ADC built-in offset calibration mechanism, and then considering whether anything additional needs to be done in post-processing.
    Best regards,
    Jim B