[RTTY] FLDigi vs MMTTY?

[Joe Subich, W4TV]

 > If you don't want to build your own tone decoder, one of the
 > commercial products that implements it is the microHAM USB Interface
 > III.

microHAM will also implement tone decoder support for "p-FSK" in
a product to replace DigiKeyer that will be announced next weekend
in F'hafen.

 > A simple tone decoder can then turn the keyed audio tone into a
 > "hard" FSK keying line.  cocoaModem has the same function called the
 > on-off keying (OOK) option in the RTTY interface.

The same hardware can also convert "audio only" CW to "hard keyed"
output that will allow CW software that only produces sidetone
to be used with the CW mode of most transceivers.

 > Remember that you will still need to also supply the PTT keying
 > line.

If you're building your own tone decoder it is relatively easy to
add a second output (or a second detector) which includes a "delay
on release" (or "hang").  A delay of 1/4 to 1/2 second will hold PTT
closed during the spaces (FSK line open in most transceivers) in
FSK or during letter/word spaces in CW.

73,

    ... Joe Subich, W4TV
        microHAM America, LLC.
        http://www.microHAM-USA.com
        http://groups.yahoo.com/group/microHAM


On 6/18/2010 3:53 PM, Kok Chen wrote:
>
> On Jun 18, 2010, at 12:00 PM, Bill, W6WRT wrote:
>
>> I have been holding off from FLDigi due to lack of FSK keying
>> ability but now I see there is a pseudo-FSK option available. It
>> does require a small external circuit but looks easy enough to do.
>
> The pseudo-FSK keying is a single tone on-off keying (like mark-only
> receiving, but in the transmit direction :-) which in fldigi is sent
> out on the right stereo channel of the output sound card.
>
> A simple tone decoder can then turn the keyed audio tone into a
> "hard" FSK keying line.  cocoaModem has the same function called the
> on-off keying (OOK) option in the RTTY interface.
>
> On-off tone keying should be very easy to implement in any other
> software RTTY modem -- just suppress one of the two AFSK tones (and
> you need not worry about phase continuity of the tone since it is
> just being used by the tone detector; the on-off keyed tone does not
> modulate the transmitter directly).  Someone who has access to the
> MMTTY source code could look into this.
>
> If you don't want to build your own tone decoder, one of the
> commercial products that implements it is the microHAM USB Interface
> III.
>
> In case anyone is interested in why you would key a tone rather than
> "bit banging" a line such as the DTR or TxD in a serial port, it has
> to do the nature of modern operating systems.  Unless you write a
> "driver" inside the kernel of modern operating systems, you cannot
> control the timing of "bit banging" precisely (user processes get
> swapped out for short periods that you have no control over); the
> bit-banged keying ends up being a signal that has bit jitter and
> that, in turn introduces more errors at the receiving end when the
> SNR is poor.
>
> The USB audio device class is implemented as something called
> "isochronous" mode in the kernel of all the standard operating
> systems.  This ensures that any properly buffered sound data that is
> sent out to a sound card will have no timing jitters (or else music
> for the computer would sound funny).  Therefore, when you send AFSK
> or OOK (what fldigi call pseudo-FSK) tones to represent an RTTY
> signal, there will be no jitter.
>
> Our old time UARTS also don't have any jitter since the bit timing
> within a Baudot character was done in the UART hardware.
>
> Anyone familiar with listening to bit-banged CW will immediately
> notice that at higher Morse speeds, you can start hearing
> irregularities in the dit/dah spacings.  This is what led K1EL to
> develop the ubiquitous WinKey chip.
>
> In the good old days of MS-DOS, an interrupt-driven program can
> provide decently enough timing, but Windows (Mac OS too) required the
> use of an external chip to avoid the bit jitter problem, this is the
> "Win" part of the WinKey name.  The same jitter exists with
> bit-banged RTTY, except most people don't listen hard enough to know
> it is there -- but you can bet the demodulator does :-).
>
> Remember that you will still need to also supply the PTT keying
> line.
>
> 73 Chen, W7AY
>

--