Regulus uREG, User Manual

Page: 100 / 107

uREG User Manual
100
17. TESTS AND EVALUATION OF PERFORMANCE
17.1. TEST CLASSIFICATION
uREG
contains circuits and software for evaluating its performance and maintaining its
readiness for operation. Test equipment and procedures are divided into five categories:

Initial performance test (when started):
POST procedures (power on self test - initial performance testing) inspect local and peripheral
hardware. If the test fails at this stage, the device will be immediately restarted and the test will
be repeated. If the defect is permanent, these operations will be repeated infinitely.
Performance confirmed during the test is required to continue the starting procedure.

Periodic testing and performance monitoring:
During normal operation, hierarchic procedures for periodic testing are activated. The scope of
the testing during operation is limited to actions which do not conflict with main protective
functions. Regular monitoring applies to: supply source, semantic correctness and information
interchange follow-up to confirm working order of computers, relay control circuits, bus
components, program memory and the panel (LEDs, display). For some main protective
operations, maximum times are set. If these times are exceeded, the device is recognized as
faulty.

Routine calibration and controlled circuit adjustment:
The measurement path is subject to regular testing during operation of the protective device.
Apart from the check, continuous calibration circuits are activated as well. The calibration
mechanism is based on statistical quality evaluation of the path and automatic application of a
corrective value. A persistently exceeded deviation (after two attempts) beyond the adjustment
range is the reason to recognize the path as faulty.

Failure alarms and failure recovery
Device activity (at the highest test hierarchy level) is checked by a special hardware-based
supervision system (the so-called watch-dog). The watch-dog requires that procedures of the
directly lower hierarchy should systematically confirm their ‘presence’ and ‘capability’. Failed
requests – caused by power failure or software self-lock – resulting in an immediate action to
restore normal operation of the system. If damage to an element is recognized by systems and
procedures in the mentioned categories, an error is alarmed and an attempt to resume a
normal state of the device is initiated.

On-demand performance tests of selected resources:
The fifth group refers to operational checks performed when requested by the operator, who is
responsible for evaluating the tested circuit. For example, this kind of check may concern LEDs.