ADS1248

Prashant,


What you describe is possible, but there are a lot of errors that may need to be calibrated out. I'll describe some of the errors you may see in the measurement, and explain what sort of setup problems may occur.

First, there shouldn't be much problem with measuring the RTD. You will have to account for the error of the RTD itself, and then the error in the reference resistor (which is a direct addition to the gain error). The max gain error of the ADS1248 is 0.02%, so you are already near the max error that you want from the system (which you have said that is 0.025%). Since the measurement is ratiometric, you don't rely on the absolute accuracy of the reference. The sum of these errors may be more than 0.025% but much of this can be calibrated, but it's easy to see where this error comes from.

I would read one of the TI designs that we use as an example. If you go to this link, you can get the write-up:

www.ti.com/.../slau520a.pdf

Read through it carefully. It describes how to make precise measurements and accounts for the errors that you may see with an RTD system. It is written for a 3-wire system, but a lot of the analysis is applicable. Again, the calibration for this measurement is difficult. By understanding these errors, you'll be able to identify which errors you will be able to (or want to) calibrate out.

The second measurement of the 4-20mA loop current may also be difficult in getting the desired accuracy. For this you rely on the accuracy of the resistor used to shunt the current. You also rely on the accuracy of the reference voltage. The ADS1248 internal reference has an absolute error of 10mV, or 0.5%. This means you can't use the internal reference, unless you've measured it and are able to calibrate that out.

Additionally, if you are using a unipolar supply (0 to 5V for example), you cannot simply shunt the current to ground. The input range of the ADC does not extend AVSS, in this case GND. Since the input is based on the front end of an instrumentation amplifier, the input signal needs to be in the common-mode input range of this front end. The common-mode input range becomes more restrictive as the device is placed in higher gain. For more information about the common-mode input range, please see the datasheet.


Joseph Wu

Prashant,


In general, the ADS1248 is close to the performance that you want. The gain error is extremely low. This will still require a precision reference that can help you get to the final gain error specification that you require. The primary problem is that the measurement of the 4-20 mA current is not as easy because AVSS (when connected to ground) is not in the common-mode input range of the device.

One possibility is to use an ADS1220 for the 4-20 mA loop. For this ADC, the input range will extend to AVSS when the PGA is disconnected. This device has a slightly higher gain error, but I'm not sure if you have a different accuracy specification for this measurement.

May we contact you through your email? We may have other recommendations that may fit your needs.


Joseph Wu