What's up TI,
We have used ADS1248 for many years with 1000 ohm 2 mV/V strain gauges, 2 mA constant current (1 mA x 2 in parallel) and 3.6VDC rails. Internal reference, 1 ksps. We have tens of thousands in the field with very stable a/d outputs. Recently we required this part to transduce a higher impedance bridge, 10K ohms, so we used lower current source (100 uA x 2) which gave similar outputs and common mode levels with all else being the same. In this high impedance situation we find quite a bit of 1/f noise with a/d values drifting back and forth by as much as 10% of the dynamic range of the part. Because of the low frequency of undulations it is difficult to average out.
As an experiment we bought 10K 0.01% precision resistors and built a bridge on the PCB just a few mm from the ADS1248 inputs. This should have removed any cabling noise or thermocouple effects but the 1/f noise remained. Does the ADS1248 have a 1/f noise problem when transducing a high impedance bridge with current sources set at low outputs?
MikeN
