Logix
®
Digital Positioners FCD LGENIM0110-0 05/16
flowserve.com
9
15.5 Outer Loop
The Logix
®
uses a two-stage, stem-positioning
algorithm. The two stages consist of an inner-loop (pilot
relay control) and an outer-loop (stem position control).
Referring to Figure 1, a stem position sensor provides a
measurement of the stem movement. The Final Command
is compared against the Stem Position. If any deviation
exists, the control algorithm sends a signal to the inner-loop
control to move the relay in a direction, depending upon the
deviation. The inner-loop then quickly adjusts the spool
position. The actuator pressures change and the stem
begins to move. The stem movement reduces the deviation
between Final Command and Stem Position. This process
continues until the deviation goes to zero.
15.6 Inner Loop
The inner-loop controls the position of the relay valve by
means of a driver module. The driver module consists of a
temperature-compensated hall-effect sensor and a Piezo
valve pressure modulator. The Piezo valve pressure
modulator controls the air pressure under a diaphragm by
means of a Piezo beam bender. The Piezo beam deflects
in response to an applied voltage from the inner-loop
electronics. As the voltage to the Piezo valve increases, the
Piezo beam bends, closing off against a nozzle causing the
pressure under the diaphragm to increase. As the pressure
under the diaphragm increases or decreases, the spool
valve moves up or down respectively. The Hall Effect
sensor transmits the position of the spool back to the inner-
loop electronics for control purposes.
15.7 Detailed Sequence of Positioner
Operations
A more detailed example explains the control function. See
Error! Reference source not found.
2. Assume the unit is
configured as follows:
Unit is in Analog command source.
Custom
characterization
is
disabled
(therefore
characterization is Linear).
No soft limits enabled. No MPC set.
Valve has zero deviation with a present input signal of
12 mA.
Loop calibration: 4 mA = 0% command, 20 mA = 100%
command.
Actuator is tubed and positioner is configured air-to-
open.
Given these conditions, 12 mA represents a Command
source of 50 percent. Custom characterization is disabled
so the command source is passed 1:1 to the Final
Command. Since zero deviation exists, the stem position is
also at 50 percent. With the stem at the desired position,
the spool valve will be at a middle position that balances
the pressures above and below the piston in the actuator.
This is commonly called the null or balanced spool position.
Assume the input signal changes from 12 mA to 16 mA.
The positioner sees this as a command source of 75
percent. With Linear characterization, the Final Command
becomes 75 percent. Deviation is the difference between
Final Command and Stem Position: Deviation = 75% - 50%
= +25%, where 50 percent is the present stem position.
With this positive deviation, the control algorithm sends a
signal to move the spool up from its present position. As
the spool moves, the supply air is applied to the bottom of
the actuator and air is exhausted from the top of the
actuator. This new pressure differential causes the stem to
start moving towards the desired position of 75 percent. As
the stem moves, the Deviation begins to decrease. The
control algorithm begins to reduce the spool opening. This
process continues until the Deviation goes to zero. At this
point, the spool will be back in its null or balanced position.
Stem movement will stop and the desired stem position is
now achieved.
15.8 Inner Loop Offset
The position of the spool at which the pressures are
balanced, holding the valve position in a steady state, is
called the Inner Loop Offset. The controlling algorithm
uses this value as a reference in determining the Piezo
voltage. This parameter is important for proper control and
is optimized and set automatically during stroke calibration.