4-20mA signal generator

Hi Everyone,

This is my first post, so in advance thanks.

I am building a hand held portable 4-20mA signal generator for testing inputs on PLCs Etc.
I am using an AD694 current transmitter to generate the signal fed from an 18V supply (2X9V batts).
I want to use an arduino to control an LED display showing mA output. However I have run into a little problem.
I intended to use a 250 Ohm resistor as a shunt to create 1-5V across it and use this voltage to calculate the output mA, however I realised that I couldn't do this as I am using a common ground for my AD694 and my arduino. Doing it this way would just close the current loop. Is there a way of creating an isolation between the two or has anyone got an other idea for this project.



One way perhaps would be to use an optocoupler LED in place of a resistor shunt. Manufacturers claim that current transfer ratio should be quite linear relative to the forward current. If the temperature is stable then it should be even possible to achieve a reasonably good calibration.

I think the short answer is there isn't a short answer to this problem (at least that I know of).

I'm presently working on a project which uses a XTR116 current driving chip.

I think the XTR116 requires fewer additional parts than the XTR111 (which I has previously worked with).

Will you always be supplying the power for the current loop or is there an option for an externally powered loop? If you're always supplying the power to the loop, then you don't have to worry about isolating the system. (Unless it's a requirement.)

I don't see why a resistor would close the loop.
The resistor is in the return line of an external loop.
Only in circuit IF you connect an external loop.

Better to use a 51ohm resistor and 1.1volt Aref.
More accuracy and less volt drop then 250ohm and default Aref.

Good Morning,

Leo, sometimes I could kick myself I had the resistor in the output side of the loop. Cheers
Kallek, thanks for your idea would you happen to Know a typical number I could look at as that sounds like an option.
Duane, thanks for your help the AD694 is a component that is common on our systems and has proved very reliable.

…would you happen to Know a typical number I could look at as that sounds like an option.

I have not used optocouplers for analog input but I guess many would suit as long as they have phototransistor output. For digital inputs I have used SFH620A - you can also use it for AC input and dont need to worry about reverse voltage protection. The schematics would be like here - http://electronics.stackexchange.com/questions/36517/opto-isolated-arduino-input only we do not need R1 as the current is already limited. You have to find suitable “pull-up” resistor value based of the properties of the chosen part.
It should be possible to use a calibration table and let the micro to do interpolation between the points or devise a formula based on the measured values.
If you plan to experiment with the same part as mentioned above then there are three versions to choose from with different CTR. It is unlikely that the CTR is exactly the same for both diodes.
One disadvantage with LED’s is that the light could fade over time and you will loose calibration but if you plan to use it only for a temporary testing then it should be OK.