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.

INA745B: Gain error in rangeof 0 to 1 Amps

Part Number: INA745B
Other Parts Discussed in Thread: INA745A, INA700

Tool/software:

Hi All,

Our group has been extensively using INA226s with an external shunt resistor for many years now, we plan to move to INA745a/b for our upcoming project.

Our concern with ina745A/b is it's ability to measure current accurately below 1Amps, in the image below taken from the ina745x's datasheet the gain error increases exponentially below 1Amps.

Could you please provide us the test data for current between 0-1Amps in steps of 50mA and it's associated gain error. Your timely input is much appreciated.

Thanks,

Darshan

  • Hello Appreciated Engineer,

    I am looking this over and will respond shortly.

    Sincerely,

    Peter

  • Hello Darshan,

    The current sense gain error from -25A to +25A is specified to +/-0.75%. So the variation in gain for <1A will still remain within specification. Also please note that this plot has calibrated out offset error.

    This variation in gain error is really just a variation in linearity. However, this change in linearity is dependent upon layout and how much heat dissipation is given to the device. The more thermal conduction provided to the device, the flatter this curve will be. Overall, this curve will always be there due the inherent flow of heat from integrated shunt through package.

    Measuring <1A will probably require an offset calibration. Are you considering this?

    Also, measuring <1A will probably require averaging out noise as much as possible. So what is the minimum allowable throughput? Or what is minimum required BW? 

    Do you need to measure bus voltage (VBUS)?

    Do you have any required over-current detection times?

    Sincerely,

    Peter

  • Hi Peter,

    Thanks for your reply!

    Please find my answers/questions below - 

    Measuring <1A will probably require an offset calibration. Are you considering this? - What does the calibration process entail? Does it mean that every ina700/745 in the field has to be calibrated manually? Can this be automated? 

    Also, measuring <1A will probably require averaging out noise as much as possible. So what is the minimum allowable throughput? Or what is minimum required BW? - Mostly it will be low bandwidth quiescent and leakage current measurements.

    Do you need to measure bus voltage (VBUS)? Yes, that is preferable!

    Do you have any required over-current detection times? No, not at least for now!

    Thanks,

    Darshan

  • Hey Darshan,

    1-point offset calibrations calibrate out the offset of a single device at a particular temperature, common-mode voltage, and supply voltage. It can be done once for each device.

    Easiest way is to record the device’s average current reading when the load current is 0. Take this value and then record into host memory and subtract it from subsequent measurements.

    Automating this would mean programming the procedure into system firmware once system has powered up.

    Low bandwidth is helpful to get most accurate measurements. There are a lot of combination on how to choose conversion and averaging option, so it is best to narrow down what minimum signal throughout is. 

    Sincerely,

    Peter

  • Hey Darshan, 

    Here is the data as requested. Statistical analysis was performed on 3000 measurement at each current level.

    Read out from PMBUS (A) Gaps (A) Tolerance
    I Load true (A) Max I Min I Ave I eMaxI eMinI eIave eMaxI eMinI eIave
    0.0665 0.084 0.0612 0.0713904 0.0175 -0.0053 0.0048904 26.32% -7.97% 7.35%
    0.1602 0.1764 0.1536 0.1645616 0.0162 -0.0066 0.0043616 10.11% -4.12% 2.72%
    0.2892 0.306 0.282 0.2939832 0.0168 -0.0072 0.0047832 5.81% -2.49% 1.65%
    0.35383 0.3708 0.348 0.359408 0.01697 -0.00583 0.005578 4.80% -1.65% 1.58%
    0.4535 0.4752 0.4464 0.4585244 0.0217 -0.0071 0.0050244 4.79% -1.57% 1.11%
    0.5779 0.5928 0.57 0.5820676 0.0149 -0.0079 0.0041676 2.58% -1.37% 0.72%
    0.67262 0.6912 0.666 0.678128 0.01858 -0.00662 0.005508 2.76% -0.98% 0.82%
    0.74235 0.7572 0.732 0.7459176 0.01485 -0.01035 0.0035676 2.00% -1.39% 0.48%
    0.86695 0.8868 0.858 0.8696616 0.01985 -0.00895 0.0027116 2.29% -1.03% 0.31%
    0.96666 0.984 0.9564 0.9701684 0.01734 -0.01026 0.0035084 1.79% -1.06% 0.36%
    1.94667 1.962 1.9308 1.9472228 0.01533 -0.01587 0.0005528 0.79% -0.82% 0.03%
    2.9803 2.9928 2.9568 2.9762388 0.0125 -0.0235 -0.0040612 0.42% -0.79% -0.14%
    3.9523 3.966 3.9276 3.9461192 0.0137 -0.0247 -0.0061808 0.35% -0.62% -0.16%
    4.95381 4.968 4.9188 4.9438916 0.01419 -0.03501 -0.0099184 0.29% -0.71% -0.20%
    5.95915 5.9796 5.9196 5.9466812 0.02045 -0.03955 -0.0124688 0.34% -0.66% -0.21%
    6.95711 6.9744 6.9084 6.9424528 0.01729 -0.04871 -0.0146572 0.25% -0.70% -0.21%
    7.92977 7.9452 7.884 7.9162324 0.01543 -0.04577 -0.0135376 0.19% -0.58% -0.17%
    8.93218 8.9496 8.8848 8.9164932 0.01742 -0.04738 -0.0156868 0.20% -0.53% -0.18%
    9.9342 9.9528 9.8856 9.921732 0.0186 -0.0486 -0.012468 0.19% -0.49% -0.13%

    Sincerely,

    Peter

  • Hi Peter,

    I appreciate your effort in putting this data together, this is very helpful! We are proceeding to use INA745s and INA700s in our designs!

    Thanks,

    Darshan