• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 257 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.

INA253: Questions about the noise spec

Joseph McGowan
Prodigy 20 points
Part Number: INA253

Hello,

I would like to know how the noise spec was measured for this product. On the datasheet, it says that the noise density for this should be 40 nV/√Hz. If I understand this correctly, then white noise on a scope with a 100 MHz bandwidth shouldn't be larger than 0.4 mV peak to peak. However, when I use the product, matching the conditions on the datasheet as well as I can, I see noise which is an order of magnitude larger. So, I'd like to know some more details about how you did this measurement and how I could remeasure that 40 nV/√Hz number here in the lab. In particular I'd like to know what device you used to measure and its input impedance, or what resistance you measured across if you did that. The support team suggested I should ask here since they were unable to answer this question.

Thanks!

  • 0
    TI__Mastermind 27380 points

    Hi Joseph,

    There are a few points to keep in mind –

    First the noise density is input referred. When calculating output noise, we need to take into account device gain.

    Second, the calculated value is RMS. To get to peak to peak, multiply the result by 6.

    Third, the 100MHz bandwidth is decades above the bandwidth of the device. The measurement will not correlated with calculation because the noise is flat only up to about 1MHz, above which the spectral density peaks due to noise shaping. Therefore you want to limit the measurement bandwidth to 1MHz and below when using the 40nV/√Hz number.

    For more helpful tips on noise measurement, please refer to this video.

    Regards, Guang

  • 0 in reply to Guang Zhou
    Prodigy 20 points

    Thanks for the response; I have a couple of follow-up questions.

    First; when you say "take into account device gain", you are referring to the amplification of the voltage across the shunt to the output voltage, yes? In other words, roughly a factor of 100.

    Second, wouldn't using an overestimate of the bandwidth of the measurement result in an overestimate of noise as well? That is, if I the actual noise is falling off above 1 MHz, wouldn't that mean that the measured noise would be *less* than what I expect by the calculation I did previously?

    Thanks, 

    Joseph

  • 0 in reply to Joseph McGowan
    TI__Mastermind 27380 points

    Hi Joseph,

    The noise voltage gain is 50, 100 and 200 respectively for the three gain versions. If you’re referring to A2 then yes it is 100. After calculating the input referred RMS noise, time it by the gain to get the output noise.

    You’re correct, if the noise spectral density falls below 40nV/√Hz, the calculation would be an overestimate. However in reality it peaks up at higher frequencies after 1MHz even though there is a dip just before the peaking. Again this is due to noise shaping pushing low frequency noise to high frequency band.

    Regards, Guang