How to Read HART Variables With a Modem

Back to Blog

What You Need

You need a HART modem - a USB or Bluetooth device that converts between the FSK signal on the loop and your computer - and a host application or communicator software that can issue HART commands. You also need access to a point on the loop where you can connect across it, and the loop must have enough series resistance for HART communication to work, because the FSK signal needs a load to develop a readable voltage. A live 4-20 mA loop that already reads an analog value normally has this resistance in the form of the input's sense resistor.

Confirm the device is powered before you start. HART is a communication layer on a powered loop, so a two-wire transmitter has to be receiving its loop power for the modem to reach it at all. If the loop is dead, no amount of correct wiring or software will find the device, and you are troubleshooting power, not HART. The related page on the HART FSK physical layer explains why signal quality and loop resistance matter for the modem to hear the device.

Wire the Modem Across the Loop

Connect the modem across the loop so it can both hear the device's FSK signal and inject its own. On a standard point-to-point loop this means connecting the modem across a point that carries the loop current through the required resistance - commonly across the transmitter terminals or across a communication resistor - so the small AC tone has a load to work against. The modem is non-intrusive to the analog reading, so wiring it in does not disturb the 4-20 mA value a control system is reading, which is a property of how HART rides on the loop.

If the modem cannot establish communication, the usual culprits are too little loop resistance, a connection point that bypasses the resistance, or excessive noise, not the modem itself. A loop with no series resistance gives the FSK signal nothing to develop a voltage across, so the modem hears nothing even though the device is transmitting. Getting a clean physical connection across adequate resistance is the foundation the rest of the read depends on.

Discover the Device and Read Command 0

With the modem connected, have the software discover the device. On a normal single-device loop the device is at polling address 0, so a discovery or a poll at address 0 finds it; on a multidrop line you poll the range of addresses to find each device present. Discovery is covered in more depth in the page on finding a HART device at an unknown address, but for a standard point-to-point read, polling address 0 is where the device lives.

Once found, read command 0 first. Command 0 returns the device's identity - manufacturer, device type, and serial number - and its long-frame address, and reading it confirms you are talking to the device you expect and gives the host the unique address to use for the rest of the session. This is always the right first read: it anchors the session on a confirmed identity rather than an assumption, which the device ID page explains in detail.

Read Command 3 and Interpret the Values

Now issue command 3 to read the device's live data. Command 3 returns the loop current plus all four dynamic variables - the PV, SV, TV, and QV - each with its value, unit, and status, in a single exchange, which is why it is the efficient way to capture everything at once. Compare the loop-current field against what you expect the analog reading to be as a sanity check that the device and the loop agree, and note each dynamic variable's value and unit as the meaningful readings.

Finally, read the status. Every response carries the two status bytes, and the field device status tells you whether the device considers itself healthy. A value that comes back with a device-malfunction flag set is not a reading to trust, no matter how plausible the number looks, so checking status is what separates a reading you can act on from one that only appears valid. The device status byte page covers those flags; treating a flagged value as suspect is the difference between a careful read and a misleading one.

Verifying the Result

A correct read leaves you with three consistent things: a confirmed identity from command 0, live values from command 3 whose loop-current field matches the analog reading you can measure independently, and clean status bytes indicating the device considers itself healthy. If all three agree, you can trust the variables. If the loop current from command 3 disagrees with a meter on the loop, or the status flags a fault, you have found a real problem rather than confirmed a healthy device.

The most useful cross-check is the loop current. Because command 3 reports the current the device is driving and you can measure that same current with a meter in the loop, agreement between the two confirms both the device and your read are sound. A disagreement points at the device misreporting, a measurement error, or a loop problem, and it is worth resolving before you rely on any of the digital variables.

Common Mistakes

The most common mistake is connecting the modem where there is insufficient loop resistance, so the FSK signal has no load and the modem hears nothing - people then blame the modem or the device when the connection point is the problem. The second is skipping the status check and trusting a plausible-looking value from a device that is actually flagging a malfunction. The third is assuming the device is at a nonzero address on a normal loop; a standard point-to-point device is at polling address 0, and polling the wrong address finds nothing.

A subtler mistake is forgetting that the device might be reachable but at a polling address that has parked its current, so the analog reading is a flat parked value rather than a fault. If the analog current sits at a fixed low value but HART communication works fine, the device is likely in digital-only mode at a nonzero polling address, which the polling address page explains, and the flat current is expected rather than a failure.

Frequently Asked Questions