Analog Scaling in SCADA: 4–20 mA to Engineering Units Without the Guesswork

If you have ever stared at a SCADA screen and wondered whether the number in front of you is real, you are not alone. A current transformer says one thing, a transducer says another, and the value on the screen sits somewhere in between. Most of the time, the gap is not a faulty sensor. It is analog scaling — the quiet piece of arithmetic that turns a 4–20 mA loop current into kilovolts, amps, megawatts, or whatever engineering unit your operators actually need.

Get that scaling wrong and everything downstream is wrong too: alarms trigger late, trends look plausible but are off by a fixed amount, and a perfectly healthy feeder looks overloaded. Get it right, and it is boring in the best possible way. So let us walk through it properly, once, with real numbers.

Why 4 mA is not zero

The 4–20 mA loop survives in substations for a good reason: it is simple, noise-resistant, and it can tell you when something has broken. That last point is exactly why the range starts at 4 mA instead of 0 mA.

A reading of 4 mA means the measured value is at the bottom of its range. A reading of 0 mA means something else entirely — a broken wire, a dead loop power supply, or a transmitter that has stopped talking. That difference, called a live zero, is the whole point. If your range started at zero, you could never tell a genuine zero reading from a failed loop.

So the first rule of analog scaling is: 4 mA is your low end, 20 mA is your high end, and 0 mA is a fault, not a measurement.

The formula

Linear scaling from a 4–20 mA signal is just a straight line between two points. If x is the loop current in mA, low is the engineering value at 4 mA, and span is the difference between the high and low engineering values, then:

y = ((x - 4) / 16) * span + low

The 16 is simply the width of the current range: 20 minus 4. Subtract 4 to find how far you are above the live zero, divide by 16 to turn that into a fraction of full scale, multiply by the span, and add the low value back in.

When the low value is zero, the formula shortens to the version you will see most often:

y = (x - 4) / 16 * span

That is the whole trick. There is no curve, no lookup table, and no guesswork — as long as the transmitter is linear, which a standard 4–20 mA device is.

A worked example: 0–150 kV

Say a voltage transducer is ranged 0–150 kV, so 4 mA means 0 kV and 20 mA means 150 kV. The span is 150 kV.

Now the SCADA screen shows a loop current of 12 mA. What is the voltage?

y = (12 - 4) / 16 * 150
  = 8 / 16 * 150
  = 0.5 * 150
  = 75 kV

Twelve milliamps is exactly halfway through the loop, so it is exactly halfway through the range: 75 kV. You can sanity-check any scaling this way — 4 mA should give you the low value, 20 mA the high value, and 12 mA the midpoint. If any of those three do not check out, the scaling in the RTU or SCADA database is wrong, not the sensor.

Try a less tidy number to make sure the habit sticks. At 16 mA:

y = (16 - 4) / 16 * 150 = 112.5 kV

That is three-quarters of the range, as it should be.

What about raw counts?

In practice you may never see milliamps at all. An RTU or PLC analog input card often hands you raw counts instead — say 0 to 27,648, or 0 to 32,767, depending on the hardware. The idea is identical; only the numbers change. Convert the counts back to a fraction of full scale first, or scale the counts directly between the same low and high engineering values. The trap is mixing up the count range: scaling 0–32,767 counts as if it were 0–27,648 will quietly shift every reading. Always check the card's range in its manual, and test the same three points — low, midpoint, high — before you trust the value on screen.

Mistakes I see again and again

• Scaling from 0 mA instead of 4 mA, which shifts every reading by a quarter of the span.
• Entering the span where the high value belongs, so a 0–150 kV range becomes 0–300 kV.
• Forgetting that a transducer and an RTU can each have their own scaling, and only checking one of them.
• Treating a 0 mA reading as a genuine low value instead of investigating it as a loop fault.

Each of these produces numbers that look reasonable. That is what makes them dangerous. A quick three-point check after commissioning catches all of them in minutes.

Check your numbers

If you want to check your own numbers, the Analog Scaling Calculator from GridCircuits shows the formula it uses, so you can see exactly how the result was calculated rather than trusting a black box. It runs in your browser and never uploads project data, which matters when the numbers belong to a real substation.

Analog scaling is not glamorous work. Nobody frames a printout of a linear equation. But it sits between every sensor in the yard and every decision made in the control room, and getting it right is one of the cheapest reliability improvements in a substation. Work through the three test points, write the range down next to the point in your database, and the next person to look at that screen will be able to trust what they see.