Hi,
I have a question concerning the XTR105. Right now I am using one XTR105 for data acquisition of the temperature via RTD / pt100 (3-wire setup). But I like to extend my system to 8 sensors. Unfortunately the XTR105 is for such a purpose too costly. Therefore I thought to multiplex the RTD in order to use just one XTR105.
However I dont know which wire of the RTD I have to multiplex. Maybe someone can give me a clue regarding this problem.
Thanks in advice
EDIT:
After some considerations for my first approach is that I would use two multiplexers 8:1 to switch through Rline1 and Rline2 of each RTD with the result that just Rline3 is connected to the XTR105, but no loop. RTD is therefore disconnected from the circuit. And the Ron of the MUX can be neglected because I switch the RTD Line and the Refline
Hello Dennis,
You have an interesting system. I did some quick math and if you MUX the RTD and REF lines between the 8 different RTDs, the additional switch resistance will cancel out leaving the basic principles of a 3-wire design in place.
However, I see a couple of potential issues with your system design concepts. First, how are you intending to communicate with the MUX? To maintain the 2-wire design concepts the IRET voltage on the XTR105 must float based on the output current and can not be connected to any system ground. If you connect IRET to the loop-supply GND then the XTR105 will absolutely not function properly and may break. Therefore you have two options to communicate with your MUX.
1.) Have a local processor which is powered by the XTR105 VREG and sinks all of it's supply current into the IRET pin of the XTR105. This processor would then be sitting at the same potential as the MUX and XTR105 and there will not be any grounding or communication voltage level issues. However, in this type of setup I see no way that the receiver of the loop-current will have any idea which of the 8 RTDs the loop-current being transmitted represents. This is because unless you have some type of communication between the loop-supply and the processor local to the XTR105, there will be no way to know which RTD the transmitted output current represents because there is no synchronization or communication between the loop-supply side and the remote XTR105 side.
2.) Communication distance will be an issue with this second method. Since I2C can not travel very far, you will be limited to a distance of roughly 2 - 6 feet between the XTR105 and the loop supply. You could use some type of further travelling communication interface such as RS-485 or CAN, but you will still need a local isolator / level translator where one side of the isolator is biased between the XTR105 VREG and IRET signals and the other side biased to the loop-supply GND. Therefore a processor sitting on the loop-supply side of the environment will control the MUX. These communication signals will be isolated and shifted to the proper voltage potential levels for the XTR105 and MUX. This would be a little odd though for a typical 2-wire loop because usually the purpose of designing one of these 2-wire systems is to have only 2 wires that are located up to a very substantial distance from the loop-supply, the additional communication lines will defeat this purpose.
I'll leave it at that for now. Please fill in some additional details about your system and we can try to help further.
Regards,Collin WellsPrecision Linear Applications
Thank you for this precise answer, Collin.
After some reflections I recognized that this won't work because of the Ron of the MUX (as you stated as well). Therefore I have to look for some other ICs to measure my RTDs in the system.
Hello,
I'm still not sure if your system requires a 2-wire 4-20mA loop or not. But if it does not, you may want to consider a product such as the ADS1248 or LMP90100. They are Analog-to-Digital converters that are capable of interfacing with several RTDs and are a full system on chips including the current sources.
Best Regards,Collin WellsPrecision Linear Applications