Broadcast Engineer at BellMedia, Computer history buff, compulsive deprecated, disparate hardware hoarder, R/C, robots, arduino, RF, and everything in between.
3697 stories
·
3 followers

Shut Up and Say Something: Amateur Radio Digital Modes

1 Share

In a recent article, I lamented my distaste for carrying on the classic amateur radio conversation — calling CQ, having someone from far away or around the block call back, exchange call signs and signal reports and perhaps a few pleasantries. I think the idle chit-chat is a big turn-off to a lot of folks who would otherwise be interested in the World’s Greatest Hobby™, but thankfully there are plenty of ways for the mic-shy to get on the air. So as a public service I’d like to go over some of the many digital modes amateur radio offers as a way to avoid talking while still communicating.

Of Modes and Modulations

Hams speak in terms of modes and modulations when describing their radio transmissions. The difference between the two terms is mostly not important to our discussion, though, and in practice a lot of hams use the terms interchangeably. But for completeness, modulation is a way of impressing information on a radio wave, and a mode is a way of using a modulation to communicate. Modulation schemes include amplitude modulation (AM), frequency modulation (FM), and single sideband modulation (SSB). Modes include continuous wave (CW), analog voice, digital voice, images, and data.

The digital modes I want to discuss are the ones where you can easily sit down at a keyboard and have your message appear magically on another ham’s terminal across the world. I’m not going to cover CW as a data mode here, even though it clearly was the first and is arguably the most successful digital data mode ever. International Morse Code has been going strong for 140 years, and with the many advantages of CW modulation it’s likely to remain a powerful tool for as long as people care to learn their dits and dahs.  Yes, there are applications that will translate keystrokes to Morse and back, but that just feels like cheating.

Equipment

Those of us with long enough memories will recall the early days of the interwebz, when dial-up connections were the only way to get online. The sound of a modem dialing Compuserve or AOL and negotiating a connection was the soundtrack of the pre-Internet days. The modem was modulating the data signals from your computer into audio tones that would fit down the analog phone line, and demodulating the returning audio signals. All ham radio data modes basically boil down to this same process — with the addition of a little outboard equipment, data from your computer is turned into audio tones that are fed to your transmitter, and audio from your receiver is decoded back into data.

In a lot of cases, the extra equipment required to tap into most data modes these days is minimal. There was a time when special converters were needed, but with a powerful DSP built into every computer sound card, pretty much any PC will do. Many ham transceivers now have sound cards built in, too, so sometimes all you need is a USB cable and the right software. FL-Digi is a popular package that supports most of the popular digital modes, provides a waterfall display that lets you easily visualize a huge swath of bands, and even controls your rig, tuning it to the selected frequency and keying the transmitter when needed.

Data Modes

There is a bewildering number of data modes out there, and cruising through the HF bands at night can sound a little spooky. The warbling tones that seem to drift across the bands as the ionosphere does its nightly dance are a little eerie. The audio signature of each data mode is pretty distinctive, and experienced practitioners can pick out the mode just by the sound, or maybe with a little help from its appearance on the waterfall display. Noobs can get help with identifying the modes from any number of websites, or can rely on their software package to autodetect the mode.

Which mode to choose is largely a function of what is going to work best under the given conditions. Unlike modems connected by a telephone line, the physical medium in ham data modes is subject to a lot more potential for interference, both natural and man-made. Signals can be interrupted by crashes of static from electrical storms, two signals can arrive by different paths and suffer phasing problems, or the signal strength can be so low as to be barely above the noise floor. Any useful data mode has to take these vagaries into account, and some do a better job of dealing with one set of conditions than another.

Here’s a run-down of the major data modes you’ll run across and the relative benefits of each:

RTTY

Radioteletype, or “ritty” as hams call it, is the original digital data mode. It dates back almost as far as commercial radio does, with the first RTTY service established between San Francisco and Honolulu in 1932. Then as now, RTTY uses the 5-bit Baudot code to encode each character. The simplest modulation scheme for RTTY is audio frequency-shift keying (ASFK) with a 170Hz difference between the mark and the space bits. This results in a whopping 45-baud connection (you’ll notice that most ham digital modes tend to be on the low side with regard to throughput thanks to the limited bandwidths available at the relatively low frequencies needed to take advantage of the ionospheric skip needed for long-distance contacts.) As slow as it sounds, that’s still about 60 words per minute, which is plenty fast enough to keep up with most typists.

RTTY has been joined by a raft of other data modes, but there are still RTTY aficionados out there plying the airwaves. The lower end of the 20-meter band is a good place to find RTTY operators.

PSK31

One of RTTY’s advantages is that it’s technically easy to implement. But it doesn’t perform particularly well at very weak signal levels. To fix that, [Peter Martinez (G3PLX)] decided to come up with a better RTTY. In 1998, PSK31 was introduced, and it has become quite popular since then.

[Martinez] took a two-pronged approach: first, he developed a new encoding method for alphanumeric characters, called Varicode. Instead of a fixed word length like the Baudot used in RTTY, Varicode’s word-length was more Morse-like, with frequently used letters represented by shorter codes than rarer letters. Then, to modulate the code, [Martinez] leveraged the DSP in a computer’s sound card to shift the phase of an audio signal by 180° to represent a zero in the Varicode, while unshifted audio represented a logical one.

This phase-shift keying (PSK) results in a bit rate of 31 – slower than RTTY, but designed to keep up with the average typist. PSK31 is more efficient than RTTY in terms of bandwidth — only 31Hz wide — and coupled with the fact that receiver and transmitter have to be synchronized and the DSP algorithm lends itself to predicting when to expect the phase transitions that signal data being transmitted, PSK31 excels at pulling data from weak signals.

Packet Modes

The user experience for RTTY and PSK31 is pretty simple — the sending party types a terse, abbreviation-rich message on a keyboard, and the receiving party reads the message on some sort of alphanumeric display. But lest you think that Amateur data modes are just for sending straight text messages like those that were sent by Model 33 teletype terminals back in the day, there are plenty of packet modes for sending more complex messages, including email.

PACTOR is a set of modes that are based on frequency-shift keying (FSK) with a 200Hz shift. Unlike RTTY and PSK31, PACTOR encodes data as 96- or 192-bit packets, which allows the use of the Automatic Repeat Request (ARQ) error control protocol to request packets that fail a CRC to be resent. PACTOR clocks in at around 200 baud.

Unfortunately, PACTOR requires an expensive piece of equipment called a terminal node controller (TNC) between the radio and the computer. To remedy this, the WINMOR protocol was developed. Similar to PACTOR in that it’s a packet mode with error correction, WINMOR does away with the TNC by using an inexpensive USB audio link, or by leveraging the sound card built into many modern transceivers.

Both WINMOR and PACTOR are gateway protocols to the Winlink 2000 network that provides email service via HF radio. Winlink is an extremely diverse hybrid network of HF and VHF radio links into internet-coupled message servers. Emails can be composed with the full-featured RMS Express client that looks and feels pretty much like any other email client. Emails can include attachments and can be sent peer-to-peer or through the network to any other Winlink user.

As useful as the Winlink network is — it has been a huge boon to emergency communications in natural and man-made disasters where local internet service is disrupted — using it is about as exciting as sending an email, because that’s exactly what you’re doing. For my money, digging a one-to-one contact out of the noise with a couple of watts on PSK31 sounds like a lot more fun. I’m glad the Winlink network is there, and it pays to practice with it from time to time, but there are a lot more challenging data modes to explore, at least in my opinion.

I’ve only scratched the surface of the digital modes available to the mic-shy ham. Here’s hoping this gets a few more new people into the hobby, or maybe even gets those licensed but largely inactive hams on the air. After all, we all need more people to not talk to.


Filed under: Engineering, Featured, radio hacks



Read the whole story
tekvax
1 day ago
reply
Burlington, Ontario
Share this story
Delete

Retrotechtacular: Tinkertoy and Cordwood in the Pre-IC Era

1 Share

It is widely accepted that Gutenberg’s printing press revolutionized thought in Europe and transformed the Western world. Prior to the printing press, books were rare and expensive and not generally accessible. Printing made all types of written material inexpensive and plentiful. You may not think about it, but printing–or, at least, printing-like processes–revolutionized electronics just as much.

In particular, the way electronics are built and the components we use have changed a lot since the early 1900s when the vacuum tube made amplification possible. Of course, the components themselves are different. Outside of some specialty and enthusiast items, we don’t use many tubes anymore. But even more dramatic has been how we build and package devices. Just like books, the key to lowering cost and raising availability is mass production. But mass producing electronic devices wasn’t always as easy as it is today.

Pre PCB

At one time, electronics were assembled by hand with point-to-point wiring and a variety of terminal strips and connections (for example, see the 1948 Motorola TV set to the right). Wires formed connections to the terminal strip (the component with five lugs near the top left), sockets, lugs on controls, and components.

This is in stark contrast to today where all the components would mount on a printed circuit board (PCB). Actually, PCBs in some form have been around since the early part of the 20th century. Even [Thomas Edison] tried to plate conductors on paper.

Despite some limited use, PCBs didn’t really take off until World War II. German mines and U. S. proximity fuses used them. Still, it would be well into the 1950s, though, before consumer electronics started to really use PCBs.

There are some technical advantages (and disadvantages) to using PCBs. But the most obvious advantage is just in labor savings. Assemblers used to use photographs and checklists to be sure they made every required connection. Not only was this labor-intensive, but it was also prone to error. The photo below shows an RCA radio factory in 1937.

Despite the name, printed circuit boards are not always printed (in fact, today, they are rarely printed). But the idea that you can make one template and automatically make tens, hundreds, or thousands of identical copies is the same idea of the printing press.

Pre IC

The other printing-like process that changed electronics forever was the integrated circuit. While the PCB allowed wires to be reproduced flawlessly, the IC lets you create entire circuits with many components and then reproduce them relatively easily. There are other advantages, too (miniaturization, close matching of active devices, etc.). But the ability to produce a CPU, for example, with all its components and wiring repeatedly using a reasonably simple process has driven price and innovation in the electronics business since it became available.

If you think about it, the IC makes a lot of things practical. Let’s say you are going to create a fish finder that uses a sonic pulse to find your dinner (or the bottom of the lake). You probably need an instrumentation amplifier. How much could you spend to develop it? You probably can’t sell millions of fish finders, so you won’t have a lot of time to refine your design. It probably can’t have too many components in it either, or the price will go up and you’ll have even fewer sales.

You can buy an instrumentation amplifier as an IC very inexpensively. The company that makes it probably spent years getting it to the current state that it is in, going through multiple product iterations. Although more components do drive up the cost of an IC (due to driving up the die size), it doesn’t raise it very much, especially at smaller die sizes where manufacturing processes have very high yields. So your choice is to design your own inferior amp using a few devices at great cost or spend the buck or less to get a well-tested design with dozens of devices and great specifications. Easy choice. Of course, if you can find a highly-integrated fish finder IC (don’t laugh, the LM1812 was a thing; see page 81 of this old Popular Electronics) then you can use that and be done with your whole design in an afternoon.

Better Living Through Military

The military is often the first to find ways to pay for new technology. They had been searching for a reasonable way to get more reproducible electronic assembly for some time. The U.S. Navy, for example, had project Tinkertoy. The idea was to make little modules out of ceramic with silver patterns painted on them. Components like resistors and capacitors could also be placed on the boards using automated processes. At the end, modules stacked together and little tabs around the edges served as a guide for interconnection wires as well as keys for orienting the boards.

You can see a very detailed–and a little stiff–video from 1953, below, explaining how the system worked. NIST (the technical muscle behind Tinkertoy) also has a photo gallery of both the devices and the pilot plant.

Cordwood

This was an improvement, of sorts, over another technique often used in military electronics back then known as the cordwood module. Cordwood modules were the ultimate in packaging density and shock resistance when using normal components and found use in military, space, and high-speed computer systems (since the density allowed wiring to be shorter).

The idea was to use two insulating cards. Components like resistors would bridge between the two boards and nickel ribbon would form wiring on the outer surfaces of the insulating cards. Thinner cards were used where ribbons intersected. Later, single sided PCBs would sometimes act as the cards and copper traces replaced the nickel ribbons. The module in the picture below uses this method.

Of course, if a component in the middle of the module went bad, you had a lot of work to get to it.

Don’t Forget the Army

The Navy wasn’t the only one thinking about this. The Army had their MM (Micro-Module) concept that looked a lot like Tinkertoy. The document covered late 1962 which was the 19th quarter of the program, so it was a little bit behind Tinkertoy. The MicroPac computer used as a test case looked pretty interesting and surprisingly compact for its day.

There were probably other module standards floating around (The NSA’s “flyball modules” come to mind), but really what everyone wanted was the integrated circuit. Then again, the first IC was in 1960, so the Micro-Module and its siblings were doomed almost from the start.

Modern Times

Once in a while, you’ll still see something done in the cordwood style (like the blinking light board in the video below). The board in the video is actually sold as a kit with no instructions as a puzzle challenge. Your job is to build it without looking at instructions (like we would do that, anyway).

Of course, you still see standardized modules around. It is just usually, they take the form factor of an IC. The Basic Stamp comes to mind. You can get whole ARM systems on a little DIP board. There are still a few places where hybrid integrated circuits are used instead of pure integrated circuits.

The Time Has Come

It is interesting that looking back you can see the pattern. The industry clearly wanted cheap standardized modules that didn’t require a lot of hand assembly. Just as the printing press allowed mass production of books and other reading material, PCBs allowed mass production of wiring and ICs the mass production of entire circuit “boards” with components.

There are more modern examples, too. Everyone was moving towards small boards that could run Linux when things like the Gumstix and the BeagleBone appeared. But the Raspberry Pi made it cheap and that’s what pushed high adoption rates. We are already seeing those get pushed into the chip level, too.

What you have to wonder is what trends are we in the middle of today? Will someone come out with a cheap multimaterial 3D printer (that includes metal or circuit components)? We are pushing to where the whole planet will be bathed in wireless networking, but at great cost. Will a Hackday author in the year 2110 write an article about how we used to build cell towers and satellites everywhere to get network devices connected? Spotting those trends can be lucrative or–if you fail to act on them–frustrating.

Cordwood module photo by [ArnoldReinhold] CC BY 2.5


Filed under: Hackaday Columns, History, Retrotechtacular



Read the whole story
tekvax
1 day ago
reply
Burlington, Ontario
Share this story
Delete

CP/M 8266

1 Comment and 2 Shares

Hands up if you’ve ever used a machine running CP/M. That’s likely these days to only produce an answer from owners of retrocomputers. What was once one of the premier microcomputer operating systems is now an esoteric OS, a piece of abandonware released as open source by the successor company of its developer.

In the 1970s you’d have seen CP/M on a high-end office wordprocessor, and in the 1980s some of the better-specified home computers could run it. And now? Aside from those retrocomputers, how about running CP/M on an ESP8266? From multi-thousand-dollar business system to two-dollar module in four decades, that’s technological progress.

[Matseng] has CP/M 2.2 running in a Z80 emulator on an ESP8266. It gives CP/M 64K of RAM, a generous collection of fifteen 250K floppy drives, and a serial port for communication. Unfortunately it doesn’t have space for the ESP’s party piece: wireless networking, but he’s working on that one too. If you don’t mind only 36K of RAM and one less floppy, that is. All the code can be found on a GitHub repository, so if you fancy a 1970s business desktop computer the size of a postage stamp, you can have a go too.

There’s something gloriously barmy about running a 1970s OS on a two-dollar microcontroller, but if you have to ask why then maybe you just don’t understand. You don’t have to have an ESP8266 though, if you want you can run a bare-metal CP/M on a Raspberry Pi.


Filed under: classic hacks, Microcontrollers



Read the whole story
tekvax
1 day ago
reply
loved CP/M!
Burlington, Ontario
Share this story
Delete

The Best Pi Emulation Console You Can Build

2 Shares

By far the most popular use for a Raspberry Pi is an emulation console. For an educational device, that’s fine – someone needs to teach kids how to plug a USB cable into a device and follow RetroPi tutorials on the Internet. These emulation consoles usually have one significant drawback: they’re ugly, with wires spilling everywhere. Instead of downloading a 3D printed Pi enclosure shaped like a Super Nintendo, [depthperfection] designed his own. It looks great, and doesn’t have a donglepocalypse hanging out the back.

The biggest factor in building an enclosure for a Pi Zero is how to add a few USB ports. There’s only one USB port on the Pi Zero, although if you’re exceptionally skilled, you can solder a hub onto the test points on the bottom of the board. This stackable USB hub solves the problem with the help of pogo pins for the power and USB pair. It’s only $17 USD, too.

With the USB and power sorted, [depthperfection] set out to design an enclosure. This was modeled in Fusion360, with proper vent holes, screw bosses, and cutouts for all the ports. It’s designed to be 3D printable, and with a little ABS smoothing, this enclosure looks great.

For software, [depthperfection] turned to Recallbox, a retrogaming platform that also doubles as a media player. It’s simpler than a RetroPi installation, but for playing Super Mario 3, you don’t really need many configuration options. This is a great project that just works and looks good doing it. The world — and the Raspberry Pi community — needs more projects like this, and we’re glad [depthperfection] sent this one in.


Filed under: Raspberry Pi



Read the whole story
tekvax
7 days ago
reply
Burlington, Ontario
Share this story
Delete

The BeagleBone Blue – Perfect For Robots

1 Share

There’s a new BeagleBone on the block, and it’s Blue. The BeagleBone Blue is built for robots, and it’s available right now.

If a cerulean BeagleBone sounds familiar, you’re not wrong. About a year ago, the BeagleBone Blue was introduced in partnership with UCSD. This board was meant for robotics, and had the peripherals to match. Support for battery charging was included, as well as motor drivers, sensor inputs, and wireless. If you want to put Linux on a moving thingy, there are worse choices.

The newly introduced BeagleBone Blue is more or less the same. A 9-axis IMU, barometer, motor driver, quad encoder sensor, servo driver, and a balancing LiPo charger are all included. The difference in this revision is the processor. That big square of epoxy in the middle of the board is the Octavo Systems OSD3358, better known as a BeagleBone on a chip. This is the first actual product we’ve seen using this neat chip, but assuredly not the last – a few people are working on stuffing this chip onto a board that fits in mini Altoids tins.


Filed under: hardware



Read the whole story
tekvax
8 days ago
reply
Burlington, Ontario
Share this story
Delete

NASA’s 2017-2018 Software Catalog is Out

1 Share

Need some help sizing your beyond-low-Earth-orbit vehicle? Request NASA’s BLAST software. Need to forecast the weather on Venus? That would be Venus-GRAM (global reference atmospheric model). Or maybe you just want to play around with the NASA Tensegrity Robotics Toolkit. (We do!) Then it’s a good thing that part of NASA’s public mandate is making their software available. And the 2017-2018 Software Catalog (PDF) has just been released.

Unfortunately, not everything that NASA does is open source, and a substantial fraction of the software suites are only available for code “to be used on behalf of the U.S. Government”. But still, it’s very cool that NASA is opening up as much of their libraries as they are. Where else are you going to get access to orbital debris engineering models or cutting-edge fluid dynamics modelers and solvers, for free?

We already mentioned this in the Links column, but we think it’s worth repeating because we could use your help. The catalog is 154 pages long, and we haven’t quite finished leaf through every page. If you see anything awesome inside, let us know in the comments. Do any of you already use NASA’s open-source software?


Filed under: news



Read the whole story
tekvax
8 days ago
reply
Burlington, Ontario
Share this story
Delete
Next Page of Stories