Hi Joseph, We have spent the last ten days trying to figure out where the issue is originated from (i.e. within or around the ADS1247 ?). Tonight, I can confirm that the issue is well originated from within the ADS1247 itself. Here is the schematic of one channel (the PLC has 10 identical channels): And a small wrap-up: The ADS1247 are powered with +/-2.6V; the IDAC current is 500uA; the RRef resistor is 2.0k 0.1% Power-up, configuration of the ADS1247s as RTD 4-wire (SELFOCAL done), the RTDs are plugged in (100R 0.01% on each of the 10 RTD channels - the PLC has 10x ADS1247 , one per AIN channel) Result is OK, let's say 100.97R measured (as per ratiometric equation) before any calibration has been performed. We leave the PLC up & running, just unplug the terminals with the 10x 100R RTD simualtion resistors. The ADS1247 start measuring anything, each channel displays something different: clear, the IEXC loop is open, the IDAC desperately tries to generate its 500uA (leading to 2,5V roughly on the IEXC pin) We plug the terminal with the 10x 100R again (in a real calibration case, it would be the 250R resistors set) The ADS1247 resumes measuring but some RTD AIN channels (i.e. some of the 10x ADS1247 ) now display an offset; in the above mentioned case, instead of 100.97R, we now have 101.29R (or a completely crazy value, sometimes). Should I change the MUX configuration to another input and then back to the RTD config OR reset & re-SELFOCAL the ADS1247 OR reboot the PLC (power off-on) THEN the initial RTD value of 100.97R is back... We can reproduce that at will. We have measured all voltages (power supplies and RTD signals along the line), the IEXC current : NO DIFFERENCE between the two situations ! The noise on the power rails is strictly the same in both situations - so no influence of the power supply. The ADG467 overvoltage protectors have no impact, as they are purely resistive. It really looks like that having the IDAC desperately trying to let its IEXC current flow through the open loop sets an offset in the part, that can be reset through a reconfiguration&SELFOCAL. Oh, an important point is that: the more RTD channels being active and unplugged at the same time, the higher the occurance and probability of that issue: i.e. wery difficult to achieve while unplugging one single channel, quite guaranteed to happen when unplugging all 10 channels. Kind regards Alain
↧