Post Go back to editing

CN0540 + CN0532 LSB to g conversion

Category: Hardware

Hi, according to the AD7768-1 datasheet (with a 4.096 V reference as is the case in CN0540) 1 LSB size is equal to 488.28 nV.

Also, according to the CN0532 and ADXL1002 datasheet, this sensor produces 40 mV/g.
 
That would mean that 81920 LSB at the output of AD7768-1 equals 1 g measured by the CN0532, assuming that 1 V at the output signal of CN0532 is equal to 1 V at the input of AD7768-1. 
But when using CN0540 + CN0532, I roughly measure that 1 g equals 63200 LSB (I've checked by fixing the accelerometer positive axis downward, flat and upward).
I'm using the CN0540 (AD7768-1) default settings.
Could somebody please explain to me the reason for that difference? I'm missing something in the middle between CN0530 and CN0540 when making that calculation?
Regards,


Typing error modified
[edited by: FelipeCook at 1:29 PM (GMT -5) on 17 Nov 2023]
Parents
  • Hi  ,

    With the CN0532's ADXL1002 sensitivity of 40 mV/g, 1 g = 40 mV.

    The CN0540 has a total gain = 0.8 from sensor to ADC.

    40 mV * 0.8 = 32 mV.

    You are correct that the AD7768-1 itself has an LSB = 488 nV.

    32 mV * 1 LSB / 488 nV = 65574 LSB. Your measurement of 63200 LSB is close, and there might be some offset remaining 65574 - 63200 = 2374 LSB = 1.16 mV referred to ADC output. With the CN0540's DC bias compensation algorithm, ideally, the remaining offset is less than 1 LSB of the DAC referred to the ADC output ("Total LSB DAC").

    However, there can be some other sources of error such as nonlinearities.

    All-in-all, I think most of the mismatch from your expected and actual measurements is due to the system gain. I hope this sufficiently answer your question.

    Regards,
    Malcolm

Reply
  • Hi  ,

    With the CN0532's ADXL1002 sensitivity of 40 mV/g, 1 g = 40 mV.

    The CN0540 has a total gain = 0.8 from sensor to ADC.

    40 mV * 0.8 = 32 mV.

    You are correct that the AD7768-1 itself has an LSB = 488 nV.

    32 mV * 1 LSB / 488 nV = 65574 LSB. Your measurement of 63200 LSB is close, and there might be some offset remaining 65574 - 63200 = 2374 LSB = 1.16 mV referred to ADC output. With the CN0540's DC bias compensation algorithm, ideally, the remaining offset is less than 1 LSB of the DAC referred to the ADC output ("Total LSB DAC").

    However, there can be some other sources of error such as nonlinearities.

    All-in-all, I think most of the mismatch from your expected and actual measurements is due to the system gain. I hope this sufficiently answer your question.

    Regards,
    Malcolm

Children