• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 46 subscribers
  • Views 671 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.

ADS1241 (do I need clean supply or just Vref)

Signalflow
Prodigy 210 points
Other Parts Discussed in Thread: ADS1241

I am using an ADS1241 ADC to read some thermocouple in a noisy environment.  Since the ADS1241 is directly connected to the outside world, I am using a ADUM3154 to isolate the SPI signals from the microcontroller.  But I would also like to isolate the power and ground of the ADS1241 from the microcontroller (since the power and ground of the ADC will probably get lots of transients on it from the thermocouples (some grounded to chassis, some not).  So I plan on using an isolated DC-DC converter (PDM2-S5-S5-S) to generate the supply voltages and ground for the ADC.  However, I measure about 400mV-pp out of the DC-DC converter (and it also varies +/-0.2 Vdc depending on the load.  So it may shift between +5.2 downt o +4.8V depending on load.

My question is will this affect the ADC accuracy if I hook the DC-DC converter output up to the ADC DVdd/Dgnd and AVdd/Agnd?  My understanding is that the ADC bases the output only by comparing analog input voltage to the reference voltage.  I have an LM4128-2.5 voltage reference so the reference should be clean.

Thanks.

  • 0
    TI__Guru* 97585 points
    Signalflow,


    I would generally use an LDO to clean up the supply. I don't have any specific information on the noise seen because of using a DC-DC converter but the listed power supply rejection at DC is typically 90dB (with a min of 75dB). It's not perfect, but it gives an idea of how much noise you'd see based on the noise in the supply. Really, 400mVpp is pretty significant and it wouldn't surprise me if the PSR could be even lower depending on the nature of the supply noise.

    Another issue might be the choice of reference. Since the ADC is measuring the input compared to the reference, the noise of the reference is reflected in the ADC measurement. As an example. If you're measuring an input of almost 2.5V that has no noise, but the reference is 2.5V with noise. The ADC measurement will see all of that noise. However, I'd note that the reference noise has almost no contribution when the input is 0V (comparing 0V to an reference, even with noise, still basically looks like 0V).

    With the LM4128-2.5, the noise is 275uV. Compared to the full reference value, this would be about almost 1000 counts of noise from the ADC measurement at positive full scale. If this is of concern, I'd look at using a lower noise reference such as the REF5025 and even filter it further it you need lower noise.


    Joseph Wu