User's Manual

Table Of Contents
USER
MANUAL
MPD-1611291-A
EXACT-V2 High End Rack Modulator / Exciter
3/10/2017
MPD-1611291-A
Page 25/94
2.6.2 Manual Digital Pre-correction
In addition to digital adaptive precorrection, EXACT-V2 also performs manual linear
precorrection. Using the “MANUAL EDIT” mode, both amplitude and phase corrections
are performed over the full spectrum of the signal.
For this purpose, a table of 32 points can be loaded in the modulator with the Set Linear
Curves command. The table defines the amplitude and the phase precorrection that
will be applied on the spectrum. For each point, a correction of ±3 dB can be defined
for the amplitude and ±500 ns for the phase. The 32 points are equally spaced over
the useful spectrum. The module itself then computes the correction to each carrier
accordingly.
6 MHz
Useful bandwidth 5.71 MHz
Frequency spacing between
2 consecutive points
184 KHz
Table 2: Linear precorrection - frequency spacing
2.6.3 Adaptive Linear pre-correction for sharp channel filters (in ATSC 1.0)
With specific, sharp tuned filters, the EXACT-V2 Linear DAP algorithm could fail.
Generally, the complexity of the channel filter is due to the fact that this filter is so
tight. The compensation of such filtering requires a high level digital filter algorithm.
The channel filter is so tight that DAP synchronization mechanisms cannot stay locked
after a few iterations.
Hitachi-Comark has developed a dedicated method to improve the Linear precorrection
for sharp filters. An additional digital filter process located after the original Linear DAP,
known as the “Sharp Filter Profile” can be activated. This filter can be configured in the
Web GUI in the “Settings/Precorrections” screen:
Note: Using the Set Linear Curves command, the processing time required by the
module (for computing the corresponding coefficients) can be long. It is
therefore recommended to use the busy flag (available in Get Linear Status
command) before any further module request, or even better: the command
Set Linear Coefficients can be used instead in order
to directly load the
coefficients.