The History of Amateur Radio
The 1990s : Digital modes and Internet (XV)
For over 80 years, RTTY (ITA2) was the king of the ASCII transmission and was used all through the world to transmit at 50 bauds scientific and commercial data, and even much more (remember those ASCII artworks).
All began in 1910 when the first "Morkrum systems" (from Morkrum company), the ancestor of RTTY terminals, were set up on the postal lines between New York and Boston. RTTY was used until the '80s as it provided a better comfort and security than CW but the signals occupied a wide spectrum over 120 kHz not really compatible with ham activities.
At the time of computers, modems and digital transceivers, amateurs looked for a conversational communications means like RTTY, but somewhat upgraded and taking advantage of the latest digital technologies like DSP and FFT. The 1960 teleprinter for example could be replaced by the screen of a computer and the teletype interface with a modem or similar interface, and an HF transceiver.
The objective was to keep the emission class of the RTTY, the "keyboard-to-keyboard" communication type, to keep the very appreciated life contact but without introducing waiting delay or very short, to work as fast as the user can type but not faster, so 31 bauds or 50 WPM can be kept, and at last, as the transceiver were now more stable in frequency, the bandwitdh could be much narrower.
Operating side, PSK31 kept some concepts of its ancestor, like the shifted keying of RTTY, now phase shifted of course instead of frequency shifted, the low speed, a very short transmission delay (400-800 ms), the lack of error-checking, and the mandatory keyboard interface. Its bandwidth dropped from approximatively 120 kHz for RTTY to as narrow as 31.25 kHz in PSK31.
In the beginning of '90s, the Polish amateur Pawel Jalocha, SP9VRC, had already developed various DSP systems (noise suppressor, Q15X25, FSK interface for Fax, etc) and Peter Martinez, G3PLX who developed AMTOR, used these concepts to created PSK31, standing for "Phase Shift Keying, 31 bauds", a new digital mode more suited to the "computer assisted amateur".
Two versions were developed, PSK31 BPSK that does not correct errors, support only 128 ASCII characters (the standard US-English and Internet ANSI format) but which transmission delay-time is fast 64 ms, and PSK31 QPSK correcting errors. This last mode displays a longer delay to 800 ms and a round-trip delay to a two-way contact of 1.6 sec due to the special algorithms included to correct errors and extend the support for all ASCII codes, or 255 characters, even if a hundred of them are practically never used.
included DSP filters (software) too. Associated to the special shape of
the audio signal, they ensured that the signal didn't extend over 40 Hz at
-3 dB below P.E.P. and 80 Hz at -40 dB as we can see on the oscillogram
displayd above, the stronger signals (double peak) displaying a bandwitdh
of 31.25 kHz. The worst distortion products are at ±
At first, Peter named his product "varicode", because it used variable length encodings to represent all 255 ASCII characters, similar to Morse code. It requested a computer because the audio signal containing the encodings must be generated and decoded. Then the encoded audio signal modulated the RF carrier of the transmitter. Peter provided initially converter kits but soon most amateurs took advantage of the DSP capabilities of the sound card that equipped their PC or the one provided by VOX and other packet or SSTV interfaces.
As poweful as CW, it can also be used for QRP intercontinental transmissions (5-10 W) and was able to make its hole in a crowded band. As it was much more comfortable than CW, included DSP features, PSK31 received immediately a warm reception. You will find on PSK31 website all technical details including a PDF file wrote by Peter Martinez dealing with this new mode.
In respect with the IARU bandplan, the recommended frequencies (for Region 1 mainly) are next : 1838.150, 3580.150, 7035.15 for Region 1 and Region 3, and 7080.15 for Region 2, 10142.150, 14070.150, 18100.150, 21080.150 (and often 10 kHz down), 24920.150 and 28120.150 kHz.
Almost all electronic loggers include today a PSK31 feature, and support many other digital modes too, to name DX4WIN. Among the dedicated software name WinPSK.
A plethora of digital modes
RTTY, AMTOR, PSK31 and Packet represent some of the many digital modes accessible to radio amateurs. Several others are born is the '90s. Let's review each of them in a few words.
PACTOR means "mediator" in latin, TOR being also the acronym for "Teleprinting Over Radio". PACTOR was developed in Germany in 1991 by Ulrich Strate, DF4KV, and Hans-Peter Helfert, DL6MAA, to improve AMTOR/SITOR and Packet radio in weak conditions of work. It combines the efficiency of packet, the error-correction (CRC) and automatic request (ARQ) of AMTOR, with a higher data transfer rate. From PACTOR I to PACTOR III the speed increased from 200 bps, to 1200 bps and now 5200 bps with data compression ! Today PACTOR-III uses a maximum bandwidth of 2.4 kHz at -40 dB. PACTOR-III is today considered at the standard digital mode on HF.
GTOR or Golay-TOR was developed by M.Golay from Kantronics, a manufacturer specialized in modems. This is a proprietary FSK mode similar to Clover or PACTOR, but cheaper.
GTOR works at low speed, only 300 bauds, but drops quite easily at some 100 bauds under bad conditions. It never had many fans.
Clover was released in 1993 by HAL Communications as a means to gain new fans on HF bands in creating a product able to break through the worst band conditions, as well as to detect weak signals. Clover is a PSK mode, full duplex, coding the signal in 4 or 8 tones (Clover-2000 or XClover). It uses the DSP technology in its last version Clover-II. Unlike Pactor, it includes a very narrow bandwidth of 500 Hz at -50 dB below the peak amplitude, where PACTOR needs from 1.5 to 2.4 kHz for each station, HF packet signals 2 kHz, and AMTOR 1 kHz.
Clover offers a great efficiency and error-correction. Unlike the other mode, thanks to DSP Clover is able to adapt to conditions on bands by constantly monitoring the received signal. Expensive in its first version, Clover-II is today much more accessible although it required always a dedicated HAL processor.
Hellschreiber, Hell for short, means "Brightly writer". It is not a new mode. It was pioneered and patented in 1929 by Rudolf Hell but this is only nowadays that amateurs discovered all its advantages. Unlike all other digital modes discussed so far, Hell displays the signal on your screen like does a fax or your television using facsimile technology. There are several versions of Hellschreiber, two of them being mostly used : the Feld-Hell, single-tone, mainly used on HF. Transmitting in CW, dark areas (text) are transmitted keyed on, while all white areas (space) are keyed off. A Fel-Hell transmission works as fast as 122.5 dots/sec or about 35 WPM using a narrow bandwidth of about 75 Hz. The second version is called FM Hell. It provides a better quality print but requires a greater duty cycle. At a third version use different tones (frequencies) to code black and white areas, it is called Multi-Tone Hell or MT-Hell. Hell is interesting because of its narrow bandwidth and compatibility with any sound card. However like many precursors it does not include any error correction system.
THROB was developed by L.G. Sear, G3PPT as an improvement of MFSK to take advantage of DSP to establish communications under difficult propagation conditions at a reasonable speed. THROB use Fast Fourier Transform technology to display waterfalls and code/decode a 5-tone signal. Today at version 2.6, Sear upgrades regularly his application. Its use stays however limited.
MT63 looks really strange at audio and it is no much used. Like PSK31 this is a keyboard-to-keyboard digital mode taking advantage of the sound card and DSP to send text in conditions of strong fading and over QRM from other stations. Unlike Hell the signal is embedded in a complex matrix of 64 tones spread over time and frequency. Due to its special modulation stucture, MT63 provides a high level of error correction, and is capable to ensure a transmission at a speed as fast as 100 WPM ! The default bandwidth is wide, 1 kHz, making this mode not really appreciated on crowded bands like they could be on weekends or during contests, where PSK31 or MFSK16 is warmly recommended.
MFSK16 is the newest "super-RTTY" mode to date (the sound that you heard in loading this page, press "Esc" to stop it). This is an advanced THROB mode encoded on 16 tones. There is a weak-signal variant, MFSK8, using 32 tones. Based on MFSK developed by 1935 by L. Devaux and F. Smets it is similar in many aspects to Hellschreiber "fuzzy" image mode, as it used the human pattern recognition ability to read the text, and the signals are not encoded. Today MFSK16 shows outstanding performances when you apply it to decode weak-signals. Some users consider it as exceeding the performance of PSK31.
Taking advantage of the sound card for DSP, unlike PSK31, MFSK16 is capable to suppress all trouble caused by multi-path reception, Doppler effect, static noise and RFI (common on low bands). In fact this is the first software providing a true digital mode. Like PSK31 it uses a constant phase frequency shift keying (CPFS) to send the coded signal. To ensure a continuous forward error correction (FEC) all data are sent twice with an interleaving technique to reduce errors from impulse noise and static crashes, making this mode almost immunized from multi-path phase shifts.
MFSK16 uses a relatively wide bandwidth of 316 Hz in which are sent the 16 tones - one at a time - at 15.625 bauds and spaced 15.625-Hz apart, allowing to reach a speed equivalent to 42 WPM. Each tone represents four binary data bits. In fact it works exactly as RTTY, in respect with ITU-D specification 316HJ2B but with 16 closely spaced tones instead of two wider-spaced tones. This reliable and performing mode is available in several popular programs, the most used of them being "Stream" developed by Nino Porcino, IZ8BLY.
This close our review of main digital modes born in the '90s. After have listed so many modes - they are others - some of them offering outstanding performances, I think that if you are interested in HF digital, you have exciting times to spend in front of your computer ahead !
Here we enter in another type of digital communications, packet that we introduced in the '80s. You probably know OH2AQ and his team who provide on their website, among others, what we call "spots information" displaying in real-time a summary of QSO established on various bands and modes. In the '80s such tools were not available yet.
A cluster or node is a computer system connected to the Internet on one side and to shortwaves the other side using VHF (2m) or UHF (70 cm) packet connections. In the field, this or these computers are connected to one or more dedicated transceivers, themselves connected to a TNC (e.g. Pakratt or Rigblaster), a sort of multimode modem able to process hig tones if its bandwidth is large enough. The information is transmitted in small packets, hence its name (see '80s).
These computers manage information about ham traffic (QSO) as well as mails sent between hams, and it provide many other interesting services that we are going to review hereunder.
Licensed hams (only) who heard or worked a station and want to publish the information to the attention of the ham community have first to connect to a near cluster using the TNC linked to their 2 m or 70 cm transceiver or using their Internet connexion to reach a dedicated website like OH2AQ.
In 1996, packed radio and specially cluster-based TNC systems (hubmasters) can provide high speed packet radio, up to 256 kbps on the 33-cm (904-916 MHz) and 23-cm bands (1.3 GHz) with a maximum data rate of 100 kbps to 1 Mbps !
Although some contests request amateurs to not use "assistance" of cluster to check activities on bands, most organizations support today the principle as in the past they had to support the widespread of SSB instead of CW. This is an evolution, that one appreciates or not. But as it is there, at range of your DSL or TNC connection, best to use it. You will find more information about clusters on this dedicated page.
Internet, the global village
Go back to 1969. This year, after more than ten years of discussions and developments, Dr. J.C.R. Licklider now at the U.S. Department of Defense, invented Arpanet, an information network built at intelligence and information purposes that was rapidely more used by the end-users that by its developers so much it was convenient to exchange data between user groups.
This "galactic network" has it was nicknamed when it was still a concept, became rapidly the most used information network through the world.
This true computing revolution allowed the development of Bulletin Board Systems (BBS), beginning in the U.S.A. and Japan, and ten years later European entered in the course for good. At the end of '70s, some companies were dedicated to the exchange of messages and ensured a technical support like CompuServe (bought by AOL in 1998), the pioneer of this concept. In 1980, CompuServe offered the first real-time chat with its "CB Simulator". CompuServe expanded to Europe in 1989.
Thanks to this provider, many amateurs linked by modem to this company began to share their own information with friends and developed true international networks. The most involved users uploaded their database on Compuserve where others can download what they needed for free or thanks to the paiment of a license.
With time BBS evolved and became true servers accessible 24 hours a day from all over the world at the sole condition to be registered on the server. Soon, private companies and governmental institutions signed agreements with this provider and others to share or sell services or information to the users.
The web however as we know it today, was invented and developed entirely by Tim Berners-Lee and a small team at CERN between 1989-1994. It was online at CERN for the first time in 1991. The story of the Internet can be read in "How the Web was born".
Unfortuantely, although CompuServe is always present on the web, today the glorious time of this pioneering company is over due to its many challengers. Indeed most of its large accounts up to the simple end-users have developed their own online service, at home, in their HQ or to a hosting company.
Internet takes its name from the concept of creating multiple independent networks of rather arbitrary design, and linked together. This "global network", because distributed all over the world, had to embody a key underlying technical idea, the open architecture networking, hence the lack of general method for federating networks as there was before; each subnet could be out of any constraint, excepted that all them had to use the same procol, TCP/IP to provide peer-to-peer services. In fact Internet included Arpanet or, say differently, Arpanet has grow inside the frame of Internet. I suggest you to check the previous links if you desire to know the little story hidden behind these inventions.
Recorded at BBC on August 10, 2005
Internet is first of all a network, hardware, constituting of adapters and cabling systems connecting computers on a peer-to-peer basis. Since its introduction, everywere, individuals, professors, and IT administrators have began to set up their computers, modems, memories, hard disks, and peripherals to get the most of this new... medium ? No. Internet is not a medium, but simply a network tool, hardware, that requests some habits to not be agressed by all its power.
We will see later that Internet is of course also used to establish (virtual) communications. Today, for 15% of the world population, so for almost one billion people, to own a computer and have an Internet connection is as common as having a clock, a transistor or a television at home. Due to its importance, we will come back on Internet on a dedicated page dealing with the future of communications.
To read : World Internet Usage
Code-less US licensees access VHF bands
During these '90s, the US hams were alway in pray with harassments because FCC refused always to remove the Morse code to access bands over 30 MHz. A small fraction of the US ham community wanted a still more restricted regulation but, like in many other countries, most disapproved this option for a simple reason : if FCC erects too much barriers, this activity will no more attract new amateurs.
Until then the code speed for the US Novice and Technician licensees was still set at 5 WPM and continued to discourage many beginners, as it did also decades earlier to european or asian amateurs, but now freed from this problem.
In fact, some authorities continue to believe that the Morse code played a major role in military communications and it could in any case be removed. An amateur had always have the capability to answer to a message coming from another service sent in code... However, many militaries could have tell him that for years, military used more efficient systems, digital and crypted, like this Stanag-4285 message send by a NATO member.
After debates with ARRL and the ham community, in 1991 FCC granted and proposed a code-free license for Technicians, allowing them to access all bands over 30 MHz. As expected, this allowed to thousands amateurs ready for decades in their starting block to enter the long lasting large family of amateurs radio. By the end of the decade, FCC had released over 700,000 licenses ! (compared to the 3 million in Japan, and the 6,000 released in Belgium). The amateur community was revitalized !
At last, remember that in 1994 was created the U.S. International Broadcasting Bureau (IBB) that provides the administrative and engineering support for U.S. government-funded non-military international broadcast services, like "Voice of America" to name one of them.