Excellent PLC
By Richard Hammond – Independent Control Systems Consultant
The Honeywell 10012/1/2 CPU module was designed in an era when control systems were meant to be finished, not continuously edited.
That context matters more than most people realize.
What the Hardware Was Built For
When the 10012/1/2 was introduced, typical workflows assumed:
-
Rare application changes
-
Long validation cycles
-
Infrequent firmware updates
-
Static process logic
Flash memory was sized and rated accordingly.
It was never intended to behave like a modern development platform.
What Modern Operations Demand
Fast forward to today:
-
Online tuning during production
-
Frequent logic tweaks
-
Temporary diagnostic changes
-
Rapid rollback and redeploy cycles
Every one of these actions writes to flash.
Not once — but repeatedly.
Why This Creates a Hidden Failure Mode
Flash memory in the 10012/1/2:
-
Has finite write endurance
-
Lacks advanced wear-leveling
-
Does not actively monitor degradation
Each online change consumes lifespan quietly.
No alarms.
No counters.
No warnings.
The Illusion of “Safe Online Changes”
From the operator’s perspective:
-
The download succeeds
-
The CPU remains in RUN
-
No fault indicators appear
From the hardware’s perspective:
-
Another erase/write cycle completed
-
Another step closer to instability
The system rewards risky behavior — until it doesn’t.
What Failure Looks Like in the Real World
Eventually, sites report:
-
Configuration values reverting unexpectedly
-
CPU behavior changing after reboot
-
Intermittent startup failures
-
Logic that behaves differently day to day
These are not software bugs.
They are storage exhaustion symptoms.
Why Replacing Flash Is Not an Option
In the Honeywell 10012/1/2, flash is:
-
Soldered
-
Integrated
-
Not field-serviceable
Once degraded, mitigation options are limited.
The CPU must be replaced.
How Systems Should Be Managed Instead
Recommended practices:
-
Batch changes instead of incremental edits
-
Avoid frequent online downloads
-
Treat flash writes as lifecycle events
-
Plan CPU replacement proactively
What This Means for Long-Term Reliability
The 10012/1/2 is not unreliable.
It is simply being asked to do something it was never designed to do.
Modern flexibility applied to legacy hardware creates silent stress.
Final Perspective
The Honeywell 10012/1/2 CPU flash memory does not fail because it is weak.
It fails because expectations have changed — and the hardware hasn’t.
Respect the design era of your controller,
or the controller will eventually remind you.
— Richard Hammond