Troubleshooting
Self-Test Overview
1-82
Self-Test Overview
Philosophy
The philosophy of self-test is, “It is better to pass a defective signal generator than to fail a good signal
generator. Self-tests are designed as a diagnostic tool used to check signal generator hardware. Self-tests do
not require the signal generator to be calibrated before self-tests will pass. This philosophy implies that some
failures will not be detected and reported as bad, but will pass. If the signal generator powers on and
self-tests can be run, troubleshooting should always start by running self-tests and resolving reported
problems first.
Overview
Embedded in the signal generator’s firmware is a group of tests (self-tests) that can be used to evaluate the
condition of the circuitry on many of the assemblies. Self-tests check hardware performance under controlled
conditions and compare test results to predetermine limits. If the results fall within the limits the signal
generator passes. If one or more test results fall outside the limits, the self-test routine evaluates the failures
and reports the most independent failure.
How self-test works:
1. When the signal generator is tested at the factory, self-test limits are loaded into the A18 CPU memory
from devices on each assembly or from fixed files. Most assemblies contain limit information while
microcircuit limits must be loaded.
2. When the program runs, latches and DACs are set to a known state.
3. An analog mux routes the detected dc signal from the selected rest node to the A18 CPU.
4. The A18 CPU compares the detected dc level to the test limit. If the detected value is within the limits, the
A18 CPU reports a pass.
5. If one or more detected values fall outside the limits, the most independent failure is determined and
reported.
Independent failures
Independent failures can best be understood by using an example. There are three assemblies in a series. A1
is the first assembly, A2 the second assembly, and A3 the third assembly. A1 output is the input for A2 and A2
output is the input for A3. If A1 output is bad, A2 and A3 pouts are also bad. If A2 output is bad, A3 output is
bad. A file in the signal generator contains this dependency knowledge for the signal generator to use to
identify the most independent failure. If A1, A2, and A3 self-tests all fail, the signal generator reports A1 as
the most independent failure. If A1 passes but A2 and A3 fail, self-test reports A2 as the most independent
failure. If A2 passes and A3 fails, self-test reports A3 as the most independent failure. If two independent
failures occur, self-test reports only one failure at a time and reports the second failure after the first reported
failure is repaired.
The troubleshooting procedure is based on the most independent failure and no lower level information is
needed. Although not needed for troubleshooting the signal generator, the self-test overhead provides
features to view lower level detailed information about the subroutines and select specific routines to run.
These features include:
• Select/Deselect Test
• Select/Deselect All
• View Details
• Run Highlighted Tests
To Next Page
Loading...
Need help?
General
