Blog · 2026-08-23
Beyond the single loop: ratio, override and split-range control
Three control structures that go beyond one controller and one valve — ratio for blending, override for protection, split-range for two valves sharing one output.
One controller, one valve is the easy case
Most of the loops on a plant are simple in structure even when they are hard in practice: one measurement, one controller, one valve. The structures that go beyond that are less numerous and more consequential, because they decide how the loops interact before tuning is even on the table.
Three of them recur everywhere — ratio control, override control and split-range control — and each one is an answer to a different question the single loop cannot ask.
Ratio control: keeping two flows in step
Ratio control holds two flows in a fixed proportion. One flow is the wild stream and is only measured; the other stream's setpoint is that measurement times a ratio, so the ratioed loop follows the wild flow automatically as the load moves.
It is the natural structure wherever the relationship between two streams matters more than either absolute value: blending, air-to-fuel on a burner, reagent dosing against a feed. The ratio is the operator's handle — raise it and the proportion changes, whatever the total is doing.
The common extension is to hang a third loop on top, so a composition or temperature controller trims the ratio. The ratio then stops being fixed and becomes the mechanism the outer loop works through, which is where the structure meets cascade and feedforward thinking.
Override control: the selector that protects
Override control lets several controllers share one output through a selector. Each controller has its own measurement and setpoint, and the selector — low-select, high-select, or something stranger — decides which one drives the output at any moment.
The structure exists so a limit can protect the process. The temperature controller on a heater runs the fuel valve most of the time; a fuel-pressure controller takes over the moment its limit is threatened and hands the valve back when it is not. One is the regulating controller, the other is the override.
The failures live in the transfers. A controller that sits unselected for hours accumulates integral error the whole time, so anti-reset windup has to engage while it is idle; and the handoff has to be bumpless, or the process takes a step every time the override trips.
Split-range control: two valves on one output
Split-range control drives two valves from a single controller output, each covering part of the range. The classic arrangement is heating and cooling: the output drives the heat valve across the lower part of the range and the cooling valve across the upper part, so the loop never has to switch controllers.
The split point — where control passes from one valve to the other — is a configuration choice with consequences. An overlap leaves both valves open a little across the split, softening the handoff but wasting energy; a deadband has neither valve moving across the split, comfortable until the loop has to sit there.
Valve sizing is what breaks the scheme. Two valves sharing one range are two process gains meeting at the split, and if the installed characteristics do not match, the loop is stable on one valve and unstable on the other.
What PIDSnap does here
All three structures live in the control system's logic, and PIDSnap does not build them — it reads the loops inside them. A ratioed loop whose ratio is drifting, an override sitting saturated against its limit, a split-range loop stable on one valve and not the other: those are the sessions the pre-flight check and the per-loop guidance are built for.
The glossary entries and the valve characteristic selector cover the ground around the choices. None of it replaces the architecture; it makes the architecture you have behave the way it was meant to.
Questions that come up
What is the difference between ratio control and cascade?
Ratio control keeps two flows in a fixed proportion — the ratioed flow tracks the wild flow automatically. Cascade control is feedback on feedback: an outer loop sets the setpoint of an inner loop based on a process measurement. The two are often combined, with a composition loop trimming the ratio.
Why does my split-range loop stall in the middle?
Check the split point. A deadband across the split leaves neither valve moving while the output sits in that region, and a mismatch in installed valve gains makes the loop stable on one valve and not the other. Both are configuration problems, not tuning problems.