If you ever decide to apply for a Part 5 experimental license for your LF work, one thing you'll have to identify are the modes of operation. Of course with it being a government form, it can't have a simple answer like, "I'll be using QRSS-30."
Instead you must use a seven character emission designator based on a standard used by the ITU. Other documents explain what the seven characters mean (the first four are the bandwidth, the last three are the more traditional emissions designators we're use to seeing) so I won't go into details there.
You must also give the baud rate for the transmission mode as part of the Part 5 application. The thought of having someone want to know the baud rate of my shaky 7 WPM fist is kind of scary...
But hopefully this will save someone some time and effort at some point...
Morse code speed is typically given in words per minute (WPM). To convert WPM to baud, the standard word used is "PARIS" which has 50 symbols. 50 symbols in 60 seconds would mean that 1 WPM is equal to 50/60 or 0.83 baud. (Or the baud rate can be calculated by dividing the WPM by 1.2.)
Bandwidth depends on the keying constant K. A K of 3 is often used for "soft" keying while a K of 5 is used for harder keying. Bandwidth can be calculated using the formula BW = (WPM / 1.2) * K.
To do this calculation on QRSS keying, you can convert QRSS times to WPM by using the formula WPM =1.2 / (Dit Length in Seconds). So for QRSS-30 (30 second dit), you have 0.04 WPM speed.
Concerning QRSS, some people make the argument that because keying is so infrequent, that K effectively drops to a value of one. I'm not sure that there's an "official" answer to that, but in my calculations I've stayed conservative and used a K of 5. Bottom line is that with QRSS, the bandwidth is so small no matter what reasonable value of K you select that it probably doesn't really matter too much.
It is more appropriate to use the designation A1A instead of N1A for Morse code because CW code transmission is, in fact, an amplitude modulated signal. It's just that you only have two amplitudes of the carrier (all or nothing) in effect.
Also, A1A is for aural reception while A1B would be for "automatic" reception. So for modes like QRSS, the A1B designation is probably the correct choice.
FSK transmission speeds for modes such as RTTY are measured in baud where baud is simply the number of symbols sent per second. (And for simple FSK, each symbol only contains a single piece of information.) There are also some simple FSK modes used on LF where the dits and dahs are replaced by constant length transmissions using two different frequencies.
Bandwidth of FSK transmissions is calculated using the formula BW = (K * Shift in Hz) + B. K is a constant that depends on allowable signal distortion and B is the baud rate determined above.
Typical amateur FSK modes are designated as F1B if the carrier is actually shifted or a J2B where audio tones are feed to a SSB transmitter and the frequency "shift" occurs because of changes to those audio tones.
With LF work, you sometimes will see people using what is called either DF-CW or FSK-CW. In this mode the dots and dashes of Morse code are represented by two different tones but sent with the same length. Inter-character timing is the same length as a dot or dash and inter-word timing is usually just double that length.
So here's where I go out on a limb and do some calculating that I haven't found talked about anywhere. It seems like your baud rate could be calculated by the formula B = 60 / (Length of dot or dash in seconds).
I'm assuming the bandwidth equation still holds for this mode, so BW = (K * Shift in Hz) + B. I've seen references to K in the FSK bandwidth calculation of typically being around 1.2.
So if you think about what I've seen referred to as QRSS-6 FSK-CW, you'd have a baud rate of 10. If your shift is 6 Hz and if you assume a conservative K of 1.5, you're bandwidth would be (1.5 * 6) + 10 or 19 Hz.
If you consider using 30 second symbols, you'd have a baud rate of two. You'd be running your decoding program (like Argo) at a lower bandwidth, so you'd probably want to consider a smaller shift, maybe something like 0.5 Hz. In this case you'd get a bandwidth of 2.75 Hz.
And if you think about using a 60 second symbol with a 0.2 Hz shift, you'd get a baud rate of 1 and a bandwidth of 1.3 Hz.
Again, those are just some assumptions I've made and I haven't seen a detailed discussion of FSK-CW's baud rate and bandwidth anywhere.
WOLF (Weak-signal Operation on Low Frequency) uses BPSK. From information on the web, it seems to send (best case) a 15 character (compressed) message in 24 seconds. Information available does indicate that it has about a 40 Hz bandwidth (at -20 dB). It seems to use two carriers--a pilot tone and the data stream. There is also mention that you can control a transmitter to send WOLF using a 300 bps serial portmentioned.
So making some informed guesses with this, it would seem the bandwidth would be 40H0. The Part 2 rules say that when modulation designation F (frequency modulation) is used, phase modulation (typically designated as G) can also be present. Since there are two cariers (one pilot carrier that just contains timing info today and the other carrier with the actual data stream) it would be type 7 for the next character. And since it's for computer reception the final character would be B.
It is also mentioned that a 960 bit packet takes 96 seconds to send. So the baud rate should be 10. There is also mention of control using a 300 bps output of a serial port, but I'm assuming they divide this down to the 10 baud rate.
So I believe the correct emission designator for WOLF would be 40H0F7B at 10 baud.
PSK31 uses phase shift keying and has the emissions designator of J2B. By definition, the protocol uses a rate of 31.25 bps. The bandwidth can be calculated as BW = 31.25 x 1.2 = 37.5 Hz.
There are slower flavors of this approach that have the "look and feel" of PSK31 such as PSK08, PSK02, and PSK01. There are also alternate character sets to provide better throughput at slower speeds such as PSKAM10 supported by MultiPSK.
The specification for the WSPR protocol indicates a bandwidth of about 6Hz and a baud rate of 12000/8192 or 1.46 baud.
From what I can tell, this uses a 4 tone FSK with the tones separated by 1.46 Hz. It is typically sound card generated, so it is sent by SSB (suppressed carrier) and is intended for reception by another computer. So it would have an emissions designator of J7B.
This is a work in progress. Other modes I need to look into include CASTLE and JASON. Also need to look at the "image" transmission modes like MT-Hell and ASK HELL.
Using this information, here are some common designators that I've derived. I would strongly encourage you to verify these before using them for anything "official" like a Part 5 license application.
It's also not clear to me how detailed you really have to get. If I receive a license that allows me to operate at 15 wpm, it seems reasonable that I would be "authorized" to adjust my speed based on condition and slow down to 10 wpm or 5 wpm based on conditions.
But can you use that same argument for QRSS? Or what about going from PSK31 to PSKAM10?
Note: For all "normal" Morse code baud rate calculations, a K of 5 was assumed.
Sources of information used for this included the ARRL Extra Class License Manual, FCC Part 2 for detail on emissions designations, W1TAG's excellent comparison of various LF modes, and information from pages for packages that support various modes.
Any feedback or corrections on the things in this page would be welcome.
My Home Page / E-Mail Me
Copyright © 2009 by Jim McClanahan All rights reserved.
All trademarks and brands are the property of their respective owners.