are expanded to 30 bits by adding six bits for error
detection and correction (EDAC). These six bits are
also called hamming bits. The value of these bits is
based on parity checks of specific combinations of the
24-bit data word.
During the receive cycle, the six EDAC, or
hamming bits, are examined for errors. There is
enough redundancy in the EDAC to allow for
correction of a single bit error. The operator can
control the selection of the error correction mode. If
the data word is not a control word, the word is
examined to determine if it is error-free, contains a
correctable error, or contains uncorrectable errors. If
the DTS is in the error detection and label mode, a
detected error is identified and labeled before the data
word is sent to the CDS computer. In the error
detection and correct mode, the DTS attempts to
correct an error before sending the data word to the
CDS computer. In both modes, the six EDAC bits are
deleted and replaced with two parity error status bits.
These status bits are defined in table 2-1.
Audio Tone Generation and Characteristics
The DTS converts the 24-bit data word, along
with the six EDAC bits, into a composite audio signal
consisting of 16 tones. This composite 16-tone signal
is the data frame. The tones range in frequency from
605 Hz to 2,915 Hz and are the odd harmonics of 55
Hz. The specific frequencies of the tones are shown
in table 2-2. The 605-Hz tone is used for Doppler
correction, and the 2,915-Hz tone is used for data and
synchronization. Each of the data subcarrier tones
(tones 2 through 16 in table 2-2) represents two binary
bits of differential quadrature phase-shift modulated
The Doppler tone (605 Hz) is not phase
modulated. It is used to correct for Doppler shifts in
the received tones caused by the relative motion
between the transmitter and the receiver. It is also
used to correct for the Doppler shift that may occur
because of differences between the transmitter and
receiver frequency standards.
The 2,915-Hz tone has two separate uses. During
the transmission of the preamble and Net Sync, the
2,915-Hz tone is used to identify frame timing. This
tone is phase shifted 180 degrees at the end of each
frame. When detected by the receiving DTS, the
phase shift indicates the start of a new frame. When
the DTS is in corrected timing, this information is
used to set the timing for the data frames that follow.
When stored timing is used, the frame timing that was
set during Net Sync is used.
The Doppler and sync tones vary from each other
and the other data-carrying tones in amplitude. The
Doppler tone is 6 dB greater than the other tones.
During the Net Sync and preamble frames, the
Doppler tone is transmitted at 12 dB and the sync tone
is transmitted at 6 dB.
The Doppler tone is
transmitted at 6 dB during the transmission of data
frames and the sync tone is used as a data tone. Data
tones are transmitted at 0 dB.
The audio tones are divided into data frames to
identify the separate parallel groupings of 30 bits. It
is the phase angle shift of each of the 15 data tones
that conveys the digital information contained in the
tone. During each frame, each data tone frequency
has a particular phase. At each frame boundary, the
phase of each data tone is shifted with respect to the
previous frame. The amount of this phase change, or
phase difference, determines the value of a two-bit
number. Two data bits yield the following four
possible combinations: 00, 01, 10, and 11. Each
combination is associated with a phase difference of
one of four values: 45 degrees, 135 degrees, 225
degrees, or 315 degrees from the previous position.
Each of these angles marks the center of a
quadrant, as shown in figure 2-14. Each 90-degree
quadrant is assigned a two-bit binary value. Any
falling within that quadrant
represents that binary value. This system of data
encoding can tolerate an error in the prescribed phase
shift of up to ±44 degrees before a single bit error will
occur. An error in phase shift that is greater than 45
degrees, but less than 135 degrees, will cause the
phase angle to fall into an adjacent quadrant. Notice
that the values are assigned to each quadrant in such
a way that if a phase shift error occurs, only one bit
error will be introduced as long as the quadrant into
which it falls is adjacent to the target quadrant.