PID vs. Pseudo-Loop: Choosing How to Control the Process
A control loop should reflect the physics of the process, not just the name of the instruction. PID is an established choice for continuous regulation, but some applications benefit from an explicit step-and-wait strategy built around deadbands, timers, output increments, and exception rules. At Britton Electronics & Automation, we use the descriptive term pseudo-loop for that rule-driven approach.
What Is a Pseudo-Loop?
A pseudo-loop imitates continuous control through programmed decisions rather than proportional, integral, and derivative equations. It can use variable deadbands, variable timers, exception rules, fast overrides, preset starting points, incremental output adjustments, and conditional logic.
The deadband determines when an adjustment is needed, the increment determines how much the output changes, and the timer determines when another adjustment is permitted. Each can vary with error size, direction, equipment state, or information from another process.
Pseudo-loop is a descriptive term, not a standardized algorithm. Modern PID instructions, including enhanced controllers such as Rockwell Automation's PIDE, can also provide tracking, feedforward, limits, anti-windup, and mode management. The distinction is whether the primary regulation strategy is expressed through a tuned control equation or through explicit process rules.
Asymmetric Control: Bump Up and Bump Down
A pseudo-loop can use different adjustments in each direction. Consider a pump maintaining 60 PSI:
| Process condition | Output adjustment | Wait time |
|---|---|---|
| Below 59 PSI | Increase 2.5 percentage points | 10 seconds |
| 59–61 PSI | Hold output | No adjustment |
| Above 61 PSI | Decrease 3.0 percentage points | 10 seconds |
Illustrative settings, not recommended tuning values. Output changes are percentage points of the command range.
The larger downward step prioritizes reducing excessive pressure; the smaller upward step does not immediately undo the entire previous reduction. Combined with appropriate deadbands and waiting intervals, asymmetric adjustments can reduce hunting. They do not guarantee stability: the process response still determines appropriate increments and delays.
Near setpoint, a strategy might use smaller steps and longer waits. Farther from setpoint, it might use larger steps and shorter waits. The table shows only one band in each direction for simplicity.
Where Explicit Rules Can Help
Dead Time
In chemical dosing, a change at the injection pump may take time to reach a downstream analyzer. Repeated increases before the first change is measured can cause overdosing. A pseudo-loop can make an adjustment, wait for transport and settling time, then reevaluate. The wait can vary with measured flow. Conventional PID can also manage dead time with appropriate tuning, feedforward, or specialized compensation.
Equipment Staging
When a lag pump starts, the controller can suspend upward adjustments, select a new preset output, and change its waiting interval while header pressure stabilizes. Similar rules can respond to valve movement, backwash cycles, or demand changes.
Unreliable Inputs and Communications
A transmitter reading zero, full scale, or changing implausibly needs a defined response. Depending on the process, the controller might hold, use a validated redundant measurement, enter a restricted mode, or shut down. Loss of communications to a drive or remote I/O also needs a defined response. Holding the last output is not automatically safe.
Output Limits and Fast Overrides
If a pump reaches maximum speed while pressure remains low, further upward adjustments should not accumulate. The controller may alarm, request additional equipment, or change strategy. A high-pressure override may bypass the normal increment timer and command a larger immediate reduction, coordinated with any independent protective functions.
Transitions and Priorities
Suppose a 2.5-point upward bump is pending when a high-pressure override activates. The override takes priority; the pending bump must be canceled or explicitly suspended. When the override clears, the controller should track the actual output and deliberately resume, wait, or reinitialize. Setpoint changes, manual-to-auto transfers, and communications recovery also need defined transitions.
Windup and Output Tracking
A PID without effective anti-windup or tracking can accumulate integral error while its final output is constrained. When the constraint clears, that accumulated demand can cause an unexpected response. A pseudo-loop has no conventional integral term, but queued increments can create a similar problem. Both strategies need to cancel inappropriate adjustments, track the actual equipment command, and return to normal control under defined conditions.
Retuning and Portability
A well-structured pseudo-loop is easier to reason about. Deadbands, step sizes, and waiting intervals correspond directly to observable process behavior, so a technician can often adjust them without developing a mathematical process model. Those basic parameters can also be relatively portable between PLC platforms when scaling and execution timing are equivalent. PID parameters may differ between implementations because of equation form, time units, and scaling. Neither approach is plug-and-play: commissioning, documentation, and testing remain essential, particularly when changes affect exception priorities or equipment transitions.
Troubleshooting the Ghost Variable
A PID may calculate 65% pump speed while another routine limits the final command to 50%. Another routine may change the setpoint, and another may adjust the gains. All may be legitimate, but the technician examining only the PID cannot explain the actual pump command.
The same problem can occur in a pseudo-loop with scattered, undocumented exceptions. The important question is which routine currently owns the output, and why?
A technician should be able to see:
- Current mode and active control strategy.
- Process variable, setpoint, and error.
- Requested output and actual commanded output.
- Active deadbands, increments, and timers, when applicable.
- Output limits, overrides, and exception priorities.
- The condition responsible for the most recent output change.
Good diagnostics make the entire chain of authority visible, not just the instruction on the screen.
Choosing the Right Strategy
Conventional PID
Continuous regulation with understood dynamics, including many pressure, flow, temperature, and level applications.
Pseudo-loop
Deliberate step-and-wait behavior, independently configured upward and downward responses, significant or variable dead time, and process-specific exception handling.
Hybrid
PID performs normal regulation while supervisory logic manages staging, startup, shutdown, operating modes, and exceptions.
A hybrid design needs deliberate handoffs. When supervisory logic takes control, the PID should receive appropriate tracking or mode information so two routines do not compete for the same output.
Control the Process, Not Just the Error
PID, pseudo-loop, and hybrid control are not mutually exclusive. A facility can use each where it fits. The choice should account for dead time, interacting processes, equipment limits, abnormal conditions, and maintainability from the beginning.
The process doesn't care what the routine is named. It responds to equipment physics and the control decisions programmed around it.
How Can BEA Help?
Britton Electronics & Automation helps evaluate process behavior, define control authority, organize exception handling, improve operator visibility, and commission PID, rule-driven, or hybrid strategies. The goal is a control system that performs predictably and remains understandable to the people who support it.
Britton Electronics & Automation Inc. | Industrial Automation, System Integration & Troubleshooting
