XTR300: Using XTR300/305 with AFE882H1 to produce bi-polar current output

Hello Sohail, 

Welcome and thanks posting on e2e :) 

The XTR300/305 are not recommended for HART modem applications. The XTR300/305 has an architecture (chopping current-mirror architecture to reduce device offset and the switching) that interferes with HART communication protocol - making it incompatible. 

The XTR108, XTR115, XTR116, XTR117 are 2-wire current transmitters that are compatible with HART. 
The XTR110 is a 3-wire current transmitter that is compatible with HART. 

Of the recommendations above, these do not support bi-directional current output. I will continue the conversation internally with the systems team.

All the best,
Carolina

Sohail,

I'd like to add a bit more detail to this post.

The AFE882H1EVM does have some connections available to connect the HART to the XTR300/305 (and XTR111). I have run some basic tests, and it does look like the communications will pass through. However, I think I'd agree with Carolina that at present, I wouldn't recommend using the XTR300/305 with HART.

I've discussed this with Luis Chioye, and he and Collin Wells both have reviewed this. There are some technical challenges with using the XTR devices and getting it to pass the HART registration tests. First the XTR devices have a noise spike just under 10kHz. This frequency is very close to the FSK frequencies of HART at 1.2kHz and 2.2kHz. Because the noise spike is so close to the FSK, we need some additional filtering in the output (either current or voltage output).

Because of the filtering, there are two tests in HART registration that become a problem in operation. First, the HART device must pass an output noise during silence test. This is basically the noise floor of the device when in operation. The maximum AC RMS voltage allowed is 20mVRMS while the current mirror is producing 20mA of current to the loop (with a 250Ω resistance in the loop). In my understanding, when they tested the XTR111 with no filtering, the device failed with 40mVRMS. The failure is from the XTR111 documented current sources at ~10kHz.

The second test is that the HART device must pass a transmit impedance test where the output impedance is high enough over frequency. At DC it must be greater than 100kΩ, and then it can fall off as the frequency. This frequency response should be equivalent to 5nF in parallel with the output, which sets the upper limit of the capacitance directly at the output. In their original tests, added filtering failed maintain an equivalent capacitance of less than 5nF.

So again, I think you can technically get HART communication through the XTR300/305. However at present, I don't think it would pass HART registration tests.

Joseph Wu