Woodward's NetCon platform governs heavy turbine and industrial engines through a modular architecture where two module families do the visible work: actuator driver modules that own the output stages, and CPU modules that own the control logic. The 5501-432 (two-channel actuator driver) and the 5501-469 (CPU 040 with LL memory, the LM6000-class build) represent the two ends — and reading them as ends rather than competitors is the whole selection discipline.
The driver question: channels and torque
Actuator driver modules size by channel count and drive capacity: how many proportional actuator loops the machine runs, at what current class. A two-channel module serves a two-actuator scope; machines with more loops stack modules the way the I/O plan demands. The driver is where the control meets the metal — its diagnostics and its drive quality print directly on valve and actuator behavior.
The CPU question: logic, memory and the engine's future
CPU modules size by application: loop count, sequencing depth, and the memory class the application's history needs. The CPU 040 with LL memory exists for specific engine packages — the LM6000 family among them — where the application load and the installed base define the part. CPU selection is a growth decision too: the module that barely fits today's logic owns tomorrow's modification budget.
The decision rule
Specify drivers by actuator scope, CPUs by application class and memory history — then check the pairing rules in the platform documentation, because channel and CPU generations handshake. The platform lands in the plant's control architecture as the engine-tier specialist, and its spares philosophy inherits the same modularity.
Can one CPU serve several driver modules?
Yes — that is the platform's point: one logic module, scaled I/O. The constraint is the CPU's throughput and memory class versus the total application, which the sizing guides make explicit.
Why does memory class matter on a governor?
Application history — trends, event buffers, recipe storage — lives in memory, and heavy engine packages generate both. The LL-memory builds exist because their applications outgrew the standard class; size for the data your operations actually keep.


