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.

ADS1219: ref monitor performance

Part Number: ADS1219
Other Parts Discussed in Thread: ADS122C04

I’m considering an upgrade to your ADS122C04 part, for access to higher PGA gain settings.  I’d like to use the function of reading the external reference voltage VREFP, through the “system monitor” function with input mux set to 1100.

 

However I’m concerned about this statement in the data sheet (highlighted):

 

8.3.9 System Monitor

The device provides some means for monitoring the analog power supply and the external voltage reference. To

select a monitoring voltage, the internal multiplexer (MUX[3:0]) must be configured accordingly in the

configuration register. The device automatically bypasses the PGA and sets the gain to 1, irrespective of the

configuration register settings when the monitoring feature is used. The system monitor function only provides a

coarse result and is not meant to be a precision measurement.

When measuring the analog power supply (MUX[3:0] = 1101), the resulting conversion is approximately (AVDD –

AVSS) / 4. The device uses the internal 2.048-V reference for the measurement regardless of what reference

source is selected in the configuration register (VREF[1:0]).

When monitoring the external reference voltage source (MUX[3:0] = 1100), the result is approximately (V(REFP) –

V(REFN)) / 4. The device automatically uses the internal reference for the measurement.

 

I understand that the resolution will be limited somewhat due to the /4 ratio;  but is there another consideration that will make this measurement of the (VREFP - VREFN) less precise than any other input?  Specifically, other than the approximation of the /4 ratio (initial gain tolerance), is this measurement going to have higher noise, or higher temperature drift, than the other inputs?

 

This part should help our design with an overall reduction in component count, if we can expect stability of at least < 50 μVRMS on the reading of the VREFP  input.  A wider tolerance on the gain (ADC counts per volt) is not a concern, as long as it is stable.

 

I’d appreciate any reassurance / cautions your applications team can provide.

 

  • Hi Ajayt,

    The reference input voltage measurement uses a voltage divider which is prone to lot to lot and device to device variation.  This voltage divider will also drift.  So the short answer is yes, the measurement will have noise and drift that differs from the normal input path with respect to an analog input into the mux.  The purpose of the reference voltage input measurement is not to get a precise voltage measurement, but rather to be used in cases such as sensor diagnostics.  For example, a 3-wire RTD using a low-side reference utilizes 2 current sources for lead cancellation in the measurement path.  If the one of the current sources is missing due to a broken lead, the measurement will be 1/2 of the normal reference measurement as only 1/2 of the current is exciting the reference resistor.  For this case you do not need a precise measurement of the reference.

    As far as stability of the reference voltage measurement, there will be drift with temperature and there are no performance guarantees.  If the measurement is held at a stable temperature (room temperature for example), the result should be relatively stable.

    Best regards,

    Bob B