unmanned site. This can also be used for generating overnight test transmissions under
timer control.
The application of a voltage in the range +3 to +12V to pin 3 relative to pin 1 (see
appendix C.1) for at least 300ms initiates transmission of sequence 1. The sequence can
of course be programmed as required. If the voltage is still present when the sequence
has finished then it will run again. (Note that this does not interfere with serial
communication data which never lasts for more than 120ms even at 75 baud).
3.26 Breaking into a Repeating Sequence
When testing a line or link with the oscillator at an unmanned site, or even when testing a
tape machine, it is often convenient to leave the LA101 repeating a sequence which can
then be ‘broken into’ at any time. The LA102 will generate a
SEQUENCE ERROR 8
(missed segments), if it only receives the last part of a sequence, but this can be safely
ignored as the LA102 will stay in sequence mode until the whole sequence has been
received successfully when it will go into page results mode (this is true for single
channel and for stereo sequences).
To generate a repeating sequence the sequence definition must have ‘
<<’ added to the end
of it, and this is explained fully in section 4.17.
3.27 CCITT O.33 Compatibility EBU O.33 sequence CCITT O.33 sequence
Although work has been done using the EBU/CCITT O.33 specification for sequence
testing, this sequence has been removed with the development of our more versatile
system of segmented sequences. Our FSK format is based on the CCITT O.33
specification but the actual data carried in the FSK header and the test tones used are
different.
The O.33 sequence is primarily suited to instruments that measure simultaneously on both
channels. While this can potentially result in a shorter sequence time for given settling
times it carries a heavy penalty in complexity and cost and is only worthwhile if the
sequence has already been optimised for speed. Four factors determine the settling time
needed for each measurement on a line: group delay, echoes, autoranging, and switching
transients in the instruments. Group delay affects low frequencies the most, delaying
their arrival when frequency response is being measured, echoes are likely to effect low
level measurements such as distortion and crosstalk, which also require time to autorange
for greatest flexibility, and switching transients can be minimised by not changing
channels more often than necessary. For these reasons distortion and crosstalk deserve a
larger time allocation than simple level measurements, but the O.33 sequence allocates a
whole second to every measurement. It would be possible to implement a stereo O.33
sequence on the LA100 by reading both channels in turn during each tone, but it makes
more sense to minimise switching transients by changing channels only once and
63
3. Sequence Testing
To Next Page
Loading...