ad: Alphaant-1

Does FCC permit General+ to do "non-MORSE" CW?

Discussion in 'Working Different Modes' started by N4VDI, May 11, 2021.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: abrind-2
ad: MessiPaoloni-1
ad: L-Geochron
ad: Left-3
ad: Left-2
ad: HRDLLC-2
  1. N4VDI

    N4VDI Ham Member QRZ Page

    Last night, I was looking through Amazon, and stumbled over all the $10-20 "Pixie" type CW transceiver kits... and I started wondering...

    If I cooked up a new digimode that combined on-off keying with precise timing and self-clocked Manchester coding (or something more sophisticated), would it be legal for a ham subject to US FCC jurisdiction who holds a General class or higher license to transmit, as long as the resulting output waveform would otherwise be acceptable to the FCC if it were a single-tone SSB soundcard digimode?

    On one hand, the FCC is pretty clear that "A1A" has a very specific definition that includes both "on-off keying" and "International Morse Code"

    On the other hand, it's also pretty obvious that the FCC regulations use the definition of "A1A" primarily as a regulatory proxy to limit what novice and technician class licensees are allowed to do on HF, and not out of any technical concerns.

    1. Most new radios don't really transmit "CW" by keying a carrier on and off... they "fake it" by transmitting a tone as SSB. And on lower-end radios, transmission begins on key-down, and continues (with the tone itself following the state of the key) until a few hundred milliseconds elapse with the key up. So, there's really nothing sacred or holy about toggling the transmitter itself anymore, and no waveform difference between a single tone modulated as SSB, and a carrier switched on and off.

    2. As a General-class licensee, I can transmit pretty much any novel digimode capable of being generated by a soundcard, as long as (generalizing a bit) it takes no more bandwidth than SSB phone (upper part of the band) or 1khz RTTY (lower portion), and I don't describe it as "spread spectrum". Presumably, that includes digimodes that would consist of a single tone.

    Given #1 and #2, it seems like any potential FCC objections to it are satisfied.

    Does this sound like a reasonable interpretation?

    Why: I'm toying with the idea of trying to cook up a new digimode that's compatible with CW-only radios, but has FT8-like robustness. If nothing else, it would give me geeky bragging points, and maybe something to list on my resume as a cool development project I worked on.

    Unfortunately, since point #2 doesn't (presently) apply to Novices & Technicians, a mode like this would probably only be allowed for General class & above. On the other hand, if the FCC finally gets around to giving novices & techs narrowband digimode privileges on all the novice HF bands, I can see modes like this becoming popular, since it would enable you to have a robust digimode that worked with even simple radio kits.
    KD2RDG, KO4ESA and KC3PBI like this.
  2. K7JEM

    K7JEM Ham Member QRZ Page

    From the FCC rules:

    (7) RTTY. Narrow-band direct-printing
    telegraphy emissions having designators
    with A, C, D, F, G, H, J or R
    as the first symbol; 1 as the second
    symbol; B as the third symbol; and
    emission J2B. Only a digital code of a
    type specifically authorized in this
    part may be transmitted.

    What you are talking about would be A1B, and is legal anywhere that RTTY is legal, assuming you have a documented code that is available. That is what PSK31 does, uses a special code that is well documented and available to anyone.
    KO4ESA, NQ1B, WD4IGX and 1 other person like this.
  3. AC0GT

    AC0GT Ham Member QRZ Page

    Depending on the specifics of the mode and the FCC rules I could see this legal for Novice and Tech as well.

    The FCC rules would appear to allow automated sending of CW/Morse by someone licensed as Technician. I didn't dig too far into the rules but it seems that "intended to be decoded by ear" is how A1A is defined and what is allowed on the "CW only" bands for Novice and Technician. Okay then, you create a new code that uses on/off keying and is, at least theoretically, decoded by ear so it's an A1A mode and allowed by Technician license holders. You only need to publish the code to avoid the FCC knocking on your door. What is sufficient documentation? I don't know but getting it published on your website and/or a magazine article in something like QST should be sufficient. I recall having to get a document in public records for something and it required taking it to some county office where it was photocopied, dated, signed by some county official, then kept in their records. I think I had a receipt for the document number given to me, or I made one myself. That or some similar government office for keeping records is an option to consider. If this is sufficiently novel then there's the patent and trademark office, but that sounds expensive as there's going to be lawyers involved.

    I find this a fascinating idea and I wish you well. This sounds more useful than some other modes I've seen being debated. Keep us posted on the progress. Consider the possibility that posting your work here as meeting the FCC requirement for being documented publicly.
    KO4ESA likes this.
  4. K6CLS

    K6CLS Ham Member QRZ Page

    This is a very interesting project. Of course whatever scheme you develop could be used on the air.

    As noted, it is crucial to publish your scheme, to make it easily readable by everyone.

    PSK31 and WSJT are good examples.

    PSK uses a Huffman encoded alphabet, called Varicose, which is like the way Morse organized the letters: most common get the shortest symbols. WSJT uses flat 6 bits per character.

    Both modes published the bit encoding mechanism, and provide open source software to encode and decode.

    PSK31 was quickly and widely adopted because of the available sources for software.

    Of course, it is your choice of open source license for your software and documentation. But something like MIT or Apache or GPLv2 would be considered most ham friendly. Share freely, and contribute back to help everyone.

    Notice ROS did none of these things, did not publish details or software, described it as spread spectrum... And FCC decided transmitting it isn't allowed for US licences. Fair enough.
    KO4ESA likes this.
  5. N4VDI

    N4VDI Ham Member QRZ Page

    I'm pretty sure that if not now, at least circa 1990, the FCC was absolutely clear that it regarded the "A1A" designation as the explicit combination of "CW" as the transmission method and "International Morse Code" as its encoding type. However, AFAIK, present-day language explicitly subjecting it to the same symbol-rate (300 baud) and bandwidth limits (1kHz) as other data modes wasn't added until around 2004.

    I know this, because back in 1989 when I was a novice, I seriously looked for loopholes in the regs that would have allowed a novice or tech to use cw transmission for faux-RTTY (using 'dot' for 0 & 'dash' for 1, or the presence/absence of a carrier to encode 0 & 1). I was disappointed to discover that the FCC had already thought of & closed that potential loophole.

    So... back then, I just had to be content with 100wpm computer-generated morse. From what I distantly recall, ~160wpm was the approx. limit of what the FT-107m's transmit relay could handle before "dropping dots", and 80-120wpm was the limit to what others could actually copy. I vaguely recall that a c64 with PK-Packratt(sp?) could go at least 20wpm higher than a PK-232 could, but I don't remember whether that was due to errors, or just the PK-232's firmware maxing out at 100wpm.

    I suppose there MIGHT be a bit of a gray middle ground for novices/techs if a Morse-based pseudo-mode sent FT8-like data as a straight-up morse callsign, followed by ~54 bits of data packed into strings of 22-24 letters (e/i/s/t/a/n + r/s/d/u) according to a published algorithm. It would stretch the definition of 'plain language' a bit, but you could argue that it's no different than aggressively using obscure Q signs & prosigns (just 16.8 million, instead of a few hundred).

    But... to really benefit from FT8-like SNR benefits, you'd really have to either scrap morse encoding, or rigidly impose EXTREMELY tight timing on both transmission windows & pip-timing.

    I'm not really sure yet just how tight the timing requirements could viably BE. Hell, even on FT8, there's ALWAYS someone who's 3-5 seconds out of sync, and most are only within 500-1000ms at best.
    KO4ESA and N5HXR like this.
  6. N4VDI

    N4VDI Ham Member QRZ Page

    @K6CLS, if I do it, I'll probably start with the source to wsjt or FLdigi & try grafting it on as a new mode in a fork of it. That way, once it worked, I could just submit it to them to merge into the mainline as another supported mode.

    The choice between them would mostly depend upon whether I pursued a strategy of "structured time-division on a single frequency around a watering hole" (to allow multiple people on one frequency to use it in a ft8-like manner), or if I settled for a more conventional one-to-one paradigm.

    The main appeal of a time-divided single-frequency watering hole is that it makes it friendlier to single-frequency qrp kits... it keeps any one person from hogging the whole frequency, while maximizing potential contacts ON that one frequency.

    Off the top of my head, I'm thinking of something like 64 slices per minute, structured in a way that each slice technically carries the full callsign+payload (so it wouldn't technically be TDMA... merely complete transmissions efficiently sent in random timeslices with automatic resending every 15 or 30s to minimize the impact of collisions.

    I contemplated a CS/MA scheme where someone would transmit a 4-8 bit value with exactly one high bit, followed by listeners remaining silent if they see a value with a single 1-bit, and shouting 'NAK!' in unison if they see 2+ '1' bits (because 2 people transmitted at once), but then I decided that it would impose so much delay & overhead, it's more efficient to just have people shout their ~900ms datagram at random, and re-send it if necessary.

    For a conversation-oriented variant, I'd use out-of-order windowing. IE, if you sent something that required 3 datagrams, it wouldn't wait for confirmation of #1 before sending 2 or 3... but after #4, would re-send 1-3 unless it heard otherwise first. I don't want to re-create AMTOR/ARQ's mistake of getting hung up on strict sequencing. If part of a sentence arrives out of order, so be it... draw a placeholder for the missing text, then fill it in after it finally arrives.
    KO4ESA likes this.
  7. K7JEM

    K7JEM Ham Member QRZ Page

    I just don't see how this would work to any advantage. There are problems with OOK as opposed to the more common modes of PSK or FSK or MFSK like most of the data/rtty modes use. These cheap transmitters will not have sufficient rise/fall times, and may introduce key clicks or other non desired anomalies. I think the speed will suffer, and the overall throughput is going to be much lower than what we are seeing with digital modes using other types of modulation.
    KO4ESA and N2EY like this.
  8. N2EY

    N2EY XML Subscriber QRZ Page

    Originally, RTTY was implemented with OOK, but FSK proved so much better that FSK became the standard way.
    KO4ESA and K7JEM like this.
  9. AG6QR

    AG6QR Premium Subscriber QRZ Page

    Most of us use CW to mean "Continuous Wave", or anything that turns a carrier on and off. But Part 97 doesn't use this everyday definition. 97.3(c)(1) defines "CW" as "International Morse Code telegraphy emissions having designators with A, C, H, J, or R as the first symbol; 1 as the second symbol; A or B as the third symbol, and emissions J2A and J2B."

    This "International Morse Code..." definition of CW is what novices and technicians are allowed to transmit.

    If you use continuous wave with an encoding scheme other than International Morse Code, it becomes "RTTY" or "Data" by the Part 97.3(c) definitions. Legal for General and Extra class licensees to use, but not for Technicians.
    KO4ESA, NQ1B, KA0HCP and 1 other person like this.
  10. AC0GT

    AC0GT Ham Member QRZ Page

    The advantage I see is people with older radios, kit radios, and such being able to get in on a digital mode with a simpler interface than other digital modes. Getting a radio to send some frequency or phase shift mode can involve a complex task of matching the computer audio output to the microphone input on a radio, matching the audio in the other direction, and operation of the PTT on a serial port or such. This can come with ground loop issues and such. With an OOK system the interface to the radio can be just a relay driven by a serial or parallel port for transmit, the receive audio by a microphone near the speaker.

    Having dealt with digital radio systems I can agree that in general PSK and FSK are easier and with fewer issues on the quality of the transmission. This assumes though that one has considerable freedom with how the computer to radio interface works. If one has limits in their hardware then OOK can simplify things.

    I do know that with some radios OOK is far from ideal. If OOK is more trouble than it is worth then this isn't for you. I don't know how popular this would be but I do see a use case. This use case is unique to Amateur radio, or at least I believe so. It can't be worse than computer assisted Morse/CW which is seemingly popular. If some new encoding method lowers the error rate on computer detected Morse code then I say it's worth experimenting with.
    KO4ESA likes this.

Share This Page