How Analog Transmitters Actually Worked - And Why It Still Matters

By Mike Glass, ISA CAP & CCST III | Orion Technical Solutions LLC

Why This Series Exists

A significant portion of the instrumentation and controls workforce is operating with an outdated mental model of how transmitter calibration works. Not the new hires — usually the veterans. The technicians who have been doing this for 20 or 30 years, who run the I&C shop, and who are actively mentoring the next generation.

The gap is specific and consistent: they understand analog transmitters well. They learned on them. They can calibrate them correctly in their sleep. The problem is that the equipment changed fundamentally about 30 years ago, and in many facilities the training, the calibration procedures, and the mental models never caught up.

I have had this conversation privately with hundreds of senior technicians across many industrial sites. The gap is real, it is widespread, and in SIS applications it has direct functional safety implications. This series exists to close it.

We will cover everything from the ground up — starting with analog transmitters, because you cannot understand why the smart transmitter is different until you understand what it replaced and why.

Part 1: How Analog Transmitters Actually Worked — And Why It Still Matters

To understand smart transmitter calibration, you first have to understand what came before it — what the zero and span concept actually was, why it existed, and why it was completely valid for its era. The vocabulary that still dominates I&C training and calibration procedures today was born here.

The Analog Era: One Circuit, Two Pots

Older electronic transmitters — and their pneumatic predecessors — were built around analog circuits. The sensor signal entered a circuit and was converted to a 4–20 mA output signal through a series of analog amplifier stages. There was one signal path, one transfer function, and two mechanical adjustments that controlled it.

The Span potentiometer changed the gain of the amplifier — the slope of the transfer curve. More span = steeper slope = larger output swing per unit of input change.

The Zero potentiometer shifted the entire output curve up or down — the offset of the transfer function. Adjusting zero moved the curve without changing its slope.

For those who prefer plain language: Zero moves the curve up and down. Span changes how steep it is.

The Transfer Function: y = mx + b

If you remember your algebra, the analog transmitter is a physical implementation of the linear equation y = mx + b, where:

• y = the 4–20 mA output
• m = the slope, controlled by the Span potentiometer
• x = the sensor input
• b = the offset, controlled by the Zero potentiometer

This is not a stored value in memory. It is a physical relationship built into the circuit by the position of two mechanical components. When you turn the zero pot, you are physically changing the DC bias at the summing amplifier. When you turn the span pot, you are physically changing the feedback resistance in the gain stage.

The Pivot Point Problem

Here is where analog calibration gets interesting — and where iterative passes come from.

Because zero and span interact through a shared circuit, adjusting span does not rotate the transfer curve around the zero point. It rotates it around a pivot point that is typically located at 20–30% of span on most analog transmitters. This means that every time you adjust span, the zero endpoint shifts. And every time you adjust zero, the span relationship moves slightly.

The location of the pivot point varies by manufacturer and model. Some transmitters have an adjustable pivot point to make calibration easier. Understanding where your specific transmitter pivots is part of knowing how to calibrate it efficiently.

Why Drift Rates Drove Annual Calibration Cycles

Analog transmitters drifted. The reasons are inherent to the technology:

• Component aging: resistors, capacitors, and potentiometers change value over time, particularly under thermal cycling
• Temperature coefficient: analog circuit performance shifts with ambient temperature
• Mechanical wear: potentiometer wipers are physical contacts that wear and oxidize
• Power supply variation: analog amplifiers are sensitive to supply voltage changes
• Vibration: physical components can shift position under vibration

The result was a typical drift rate of 0.25–1.0% of span per year for well-maintained analog instrumentation. [Source: ISA-RP105.00.01-2017 context] At those drift rates, annual calibration cycles were not arbitrary — they were statistically justified to catch meaningful errors before they accumulated to the point of affecting process control or measurement integrity.

What Made Analog Calibration Work

When the procedure and the equipment matched, analog calibration was straightforward:

• Apply a known traceable input at 0% of range
• Adjust the Zero pot until the output reads 4.000 mA
• Apply a known traceable input at 100% of range
• Adjust the Span pot until the output reads 20.000 mA
• Recheck zero (it will have moved slightly)
• Repeat until both ends are within tolerance simultaneously

The procedure worked because the transmitter was a single transfer function. Adjusting output to match input was the correct and complete action. There was nothing else to consider — no separate A/D section, no digital PV, no LRV/URV register, no distinct output stage. One circuit. Two pots. Done.

This is the mental model that is still embedded in most I&C training, most calibration procedures, and most technician muscle memory. It is not wrong — for analog transmitters. The problem comes when it is applied to equipment that does not have this architecture. That is what the rest of this series covers.

Try It Yourself

If you want to experience the pivot point interaction and iterative calibration process before moving on to Part 2, the simulator below lets you do exactly that. Adjust the Zero and Span sliders, inject errors, and work through correcting them. The pivot point slider shows you how span rotation affects the zero endpoint at different pivot locations.