A pressure transmitter, a level sensor, a flow meter, most of the analog instrumentation on a plant floor that isn’t already routed through a PLC speaks the same simple language: a 4-20mA current loop. No protocol to configure, no register map, just a current that moves between two fixed points as the measured value moves between its own two fixed points. It’s the most common way to get a real physical measurement into a machine with no controller at all, and understanding how the loop works is most of what setting one up takes.
Why current, not voltage
A current loop’s whole design point is that the same current flows through every point in the loop, no matter how long the wire run is or how much resistance it picks up along the way, unlike a voltage signal, which droops over distance and picks up electrical noise a lot more easily. A sensor rated 4-20mA reports its lowest measured value as 4 milliamps and its highest as 20 milliamps, with everything in between scaled linearly across that 16 milliamp span. A pressure transmitter rated 0-100 PSI reports 4mA at 0 PSI, 20mA at 100 PSI, and 12mA sits exactly at the midpoint, 50 PSI. That’s the entire idea, and it’s why 4-20mA became the default for industrial instrumentation over a voltage-based signal that seemed simpler on paper.
Loop power: who’s powering the sensor
A 2-wire loop sensor is externally powered. The same two wires that carry the 4-20mA signal also carry the sensor’s own operating power. There’s no separate power pair. Check the sensor’s own datasheet for what it needs and where that power comes from in your wiring. Some are powered from the gateway’s own loop supply, others need an external supply wired into the loop. A 3-wire or 4-wire sensor separates power from signal entirely. Its own supply lines are distinct from the two wires carrying the 4-20mA current. Either way, wire the sensor’s signal output to one of the gateway’s ADC terminals and its return or common to GND, per its own datasheet, not from a generic wiring assumption. Loop power details are one of the places sensors differ.
Scaling: turning a raw count into a real number
The gateway’s onboard analog input doesn’t hand you milliamps or PSI directly, it hands back a raw count from its own analog-to-digital converter, a number with no engineering meaning on its own until you tell the tag what it represents. That translation happens through two numbers on the tag, scale and offset, applied as value = raw_count × scale + offset on every single read, automatically, once you’ve set them.
Work it from the two ends of the sensor’s own span. Say a 4-20mA loop sensor is wired through a shunt resistor, and its rated span is 0-100 PSI. The 4mA low end and the 20mA high end each correspond to a specific raw count on the gateway’s ADC. Read those two counts once (at zero flow, or from the sensor’s own known calibration points), and everything in between follows the same straight line: the low count maps to 0 PSI, the high count maps to 100 PSI, and a raw count of 600 partway between them maps to 50 PSI, no different math needed for any other reading on the same line. Enter the resulting scale and offset once on the tag, and Spall applies the conversion for you from that point on. Leave scale at 1.0 and offset at 0.0, and the tag reports the raw count, unconverted, useful for checking that a signal moves at all before you’ve worked out the real calibration.
What a clamp meter confirms, and what it doesn’t
A clamp meter is the fastest way to sanity-check a loop before or during setup. Clamp it around one leg of the loop and it reads the live current directly, no math needed for that one number. That confirms the sensor is powered, the loop is complete, and the current is somewhere in the expected 4-20mA band rather than pinned at zero (a broken loop) or something well outside spec (a wiring fault or a failed sensor).
What it doesn’t give you is the continuous record Spall builds from the same loop. A clamp meter shows one instant, whatever the current happens to be the moment you’re standing there with it, not a trend, not a history, not the value an hour ago when the pressure spiked. It also can’t tell you a PSI or a flow rate on its own, it reads milliamps, and turning that into an engineering unit still needs the sensor’s own span and the scale/offset math above, a clamp meter has no idea what 12mA is supposed to mean on your specific sensor. Use it to confirm the loop is alive during wiring and troubleshooting. Use Spall’s own tag reading, watched live while you move the actual process, to confirm the scaling is right.
Verifying before you trust it
The same discipline that applies to any sensed reading applies here: don’t trust a calibration you haven’t watched against a real change in the process. Move the thing the sensor measures, crack a valve, watch a tank drain, run the machine under load, and confirm the number in Spall tracks it in the direction and rough magnitude you expect. A scale and offset computed correctly on paper but never checked against a real swing is a plausible-looking number that hasn’t been proven, and that’s worse than an obviously broken one: nothing about it looks wrong until someone acts on it.
Three channels of onboard analog input is what a gateway carries on its own panel, and a plant with more loop sensors than that doesn’t need a second gateway just to fit them. A Modbus aggregator wired to the extra sensors, itself read as one Modbus TCP source, scales an install past the built-in three without adding another box’s worth of network configuration.
Quick recap
- 4-20mA current loops stay accurate over long wire runs because the same current flows at every point in the loop, unlike a voltage signal
- Check the sensor’s own datasheet for loop power, a 2-wire sensor draws its operating power from the same two wires carrying the signal
- Scale and offset turn a raw ADC count into a real engineering unit, computed once from the loop’s known low and high points, applied automatically after that
- Leave scale at 1.0 and offset at 0.0 to see the raw count first, useful for confirming a signal moves before working out the real calibration
- A clamp meter confirms the loop is alive and roughly in range. It doesn’t replace the continuous record or the scaling math
- Move the actual process and watch the number track it before trusting a calibration you’ve only checked on paper