This mode is included into the software MULTIPSK by Patrick F6CTE, there is the description of this mode as well as the specifications coming from his software :
Click on the copy of the screen in order to listen to the corresponding audio signal.
Creator : Patrick Lindecker (F6CTE) in 2006
Description :
Baud rate : 21
Speed : about 2 characters/sec so 20 wpm
Modulation : FSK 8 tones (3 bits arranged in a Gray form), with a shift between tones of 21 Hz (168 Hz bandwidth). A block is composed of 32 symbols of 3 bits (in other words it is a matrix of 32 columns (following the time) on 3 lines). Each of the 3 block lines of this block corresponds to a character which has been encoded, on a 32 bits vector, using a Walsh-Hadamard transform to provide a high level of redundancy.
Note: the implicite used matrix is 64x32 dimension, the 32 first lines are the 32 standard orthogonal combinations of the Hadamard matrix. The 32 next lines are their opposite. So any couple of lines are either orthogonal (scalar product=0) or biorthogonal (scalar product<0).
Reception mode : sensible to the side (USB or LSB), USB is recommended
Shape of pulse : specific shape (between a Hanning window and a rectangular window)
Character set : ASCII 6 bits (64 characters) but only 61 are usable.
Bandwidth : 168 Hz
Synchronization : automatic using the signal
Correction code : no
Convolution code : no
Interleaving : yes, spreading of 3 bits "vertically" in the block,
Scrambling : yes, with a 32 bits sequence applied on each line of the block but with a delay of 13 bits between two successive lines.
Drift tolerance : 10 Hz/mn (depending on signal the signal-to-noise ratio)
Pmean/Ppeak : 0.76
Lowest S/N : – 14.5 dB
Repetition of characters
This mode being done for blind or partially sighted Hams or SWL's, all received characters can be repeated. So a received character will be displayed on the screen and it could be, in the same time, repeated in an audible way.
To take profit of this function, the output plug of the sound card must be connected to an amplified speaker in parallel to the transceiver input.
Important: as the amplified speaker is going to be also active during transmission, not to collapse the output signal, the ideal would be to use the sound card "Speaker" output instead of the "Line out" output. If your sound card does not have this type of low impedance output, it must be checked that the sound card output can supports the amplificed speaker and the transceiver input, by looking, on a voltmeter, at the signal delivered by the sound card, which must not fluctuate. If it fluctuates, the only solution is to do an interface to "open" the circuit to the amplified speaker when the program is Txing (see the second example in the RS232_EN.DOC file, but replacing the relay on the diagram by an inverted relay, so as to open when Txing). Furthermore, this interface will avoid to hear the modulation when TXing.
Frequency tuning
Any AF frequency (200 to 4000 Hz) must be able to be used.
However, it is considered a standard central AF frequency of 1000 Hz for this mode, to simplify tuning. Furthermore, the "Freq. search" function permits to localize any Voice transmission in a 480 Hz band. So the ideal would be to set ON the XCVR 500 Hz CW filter and to tune the HF frequency on the XCVR instead of tuning the AF frequency on the waterfall.
Note: this mode is directly derived from the 8 tones, 62.5 bauds Olivia mode (Pawel Jalocha SP9VRC) but with a 32 bits vector instead of a 64 bits vector, a different set of characters and a different speed.
Note: this mode is directly derived from the 8 tones, 62.5 bauds Olivia mode (Pawel Jalocha SP9VRC) but with a 32 bits vector instead of a 64 bits vector, a different set of characters and a different speed.