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.

ADS124S08: Is there any degraded performace when a external reference is used instead of internal reference of ADS124S08?

Part Number: ADS124S08
Other Parts Discussed in Thread: ADS1235

 Hello guys,

 One of my customers is evaluating ADS124S08 for their new weight scale products.

 At this moment, they have the following question. Could you please tell me your advice?

Q. Their weight sensor output level is very small. It is about 1V as the maximum even though PGA is set x128.

 So they is going to try to use an external reference IC which has 1.024V or 1.25V output to increase effective resolution for the PGA output (about 1V).

 Is there any problem or any degraded performance when 1.024V or 1.25V external reference is used instead of 2.5V internal reference?

 Your reply would be much appreciated.

 Best regards,

 Kazuya.    

  • Hi Kazuya-san,

    thanks a lot for your question.

    A smaller external reference voltage could be used in this scenario but I am afraid it will probably not yield any better resolution. Let me try to explain that.
    The true noise performance of the ADC is determined by its input-referred noise as e.g. specified in Table 1 in the ADS124S08 datasheet. The uVrms number tells you what the smallest signal is that the ADC can resolve at the configured PGA gain and data rate. Decreasing the reference voltage will not change this input-referred noise number in a meaningful way.

    However decreasing the reference voltage will change the LSB size of the ADC. The LSB size is defined as:
    1 LSB = (2 x VREF) / (Gain x 2^24)
    That means the LSB size decreases with a decreasing reference voltage. Now because the input-referred noise does not change when decreasing the reference voltage it will mean that more bits will be 'noisy'. So your effective resolution calculated in bits as per equation 1 in the datasheet will actually decrease. But that doesn't mean your ADC performance has become any worse because the input referred- noise (which is the important number) is still the same.
    Talking in terms of effective resolution (in bits) can therefore always be very misleading.

    What I would actually strongly recommend for weigh scale measurements is to implement a so-called ratiometric measurement. That means the excitation voltage of the bridge is used as the reference voltage for the ADC. That way any change or noise in excitation voltage will be cancelled out and does not effect the accuracy of the weight measurement. You can refer to the application section in the ADS1235 datasheet as an example.

    If you use a fixed reference voltage for the ADC which is independent from the excitation voltage of the bridge, then any change in excitation voltage will also change the output signal of the bridge. Means if the excitation voltage increases, the bridge output signal will increase even though the weight did not change. You are therefore measuring a larger weight.
    In a ratiometric measurement implementation any change in excitation voltage is cancelled out.

    Regards,

  •  Hello Joachim,

     Thank you very much for your detail explanation.

     I will read ADS1235 datasheet application section and learn it.

     I will ask you in a few days if I have any question about the ratiometric measurement.

     Thank you again and best regaregards,

     Kazuya.