• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 47 subscribers
  • Views 1434 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.

INA226: Comparison with INA220 on resolution in same conversion time

Luke Sun
Intellectual 1830 points
Part Number: INA226
Other Parts Discussed in Thread: INA220

Hi team,

My customer is using INA226 to replace INA220. Here's one question:

Previously when using INA220, they set conversion time to 532us (table 4 on INA220 datasheet), which has 12 bit resolution & 1 sample point.

Now in INA226 datasheet table 8, they want to set conversion time to 588 us.

They want to understand what kind of resolution on INA226 when conversion time set to 588us. Is it also 12 bit resolution & 1 sample point?

Thanks!

  • 0
    TI__Mastermind 26335 points

    Hello Luke,

    Thanks for reaching out on the forum.  The INA226 actually has different architecture than the INA220.  The ADC resolution for the INA226 cannot be modified like the INA220 and will always be 16 bits.  However, the conversion time will impact how precise the lower bits are as there will be more noise.  Figure 20 in the datasheet (posted below), illustrates the impact of the conversion on the noise level.  The conversion time corresponds to 1 sample that can then be averaged by sample count specified in the configuration register.

  • 0 in reply to Patrick Simmons
    TI__Intellectual 1830 points

    Hi Patrick,

    Could you give me an apple-to-apple comparison between INA220 & INA226 when the conversion time is both set to ~550us?

    How much RMS or pk-pk noise difference between these two devices?

  • 0 in reply to Patrick Simmons
    TI__Intellectual 1830 points
    Hi Patrick,
    Any update?
  • 0 in reply to Luke Sun
    TI__Mastermind 26335 points
    Hello Luke,

    We do not have the noise measurement data for the INA220 as it an older device and our datasheets typically evolve to include more characterization data with each new generation of devices. However we do have the shunt voltage LSB resolution. The lowest LSB voltage attainable for the INA220 is 10uV while the lowest for the INA226 is 2.5uV. From the graph I posted above you can approximate the pk-pk noise of ~550us conversion time by multiplying the 1.1ms conversion time pk-pk noise by 1.4, thereby giving you a pk-pk noise of ~6.3uV. The quantization noise of the INA220 alone is at 10uV.

    Now imagine we want to measure a shunt with 100uV across it, if we super-impose the noise that value may be measured as any value between 96.85 and 103.15uV. For a discrete scale with 2.5uV steps where any voltage measured beneath a step is assigned to the previous step, the voltage could be measured between 95uV and 102.5uV, while for 10uV steps the measurement could be 90uV or 100uV assuming the noise level is equal or even smaller.