Skip navigation
1 2 3 Previous Next

The Engineering Mind

125 posts

Hello once again!  I hope that you have been enjoying my recent series on single event effects with high speed ADCs.  So far we have looked at TID (Total Ionizing Dose), SEL (Single Event Latchup), and SET (Single Event Transient).  This month the focus is on SEU (Single Event Upset).  This is quite closely related to the SETs we looked at over the last two months.  It is possible that a SEU could result in an SET depending on the ion strike and the resultant behavior of the circuit.  For the purposes here though we will look at them separately to help provide a bit of clarity into the underlying cause.  For a high speed ADC we will consider an SEU as a configuration register upset.  The AD9246S is the device under test.  As it turned out there were no configuration register upsets observed for the AD9246S but in my Planet Analog blog I discuss what an SEU might look like.  A brief idea is that an ion strike results in a configuration bit(s) upsetting.  It is expected that the bit(s) would not permanently flip but would revert back to the expected value.

SEU - Single Bit

Configuration Register SEU - Single Bit

SEU - Multiple Bits

Configuration Register SEU - Multiple Bit


Recall that it is important to perform this testing so that the device behavior in space can be predicted.  The Weibull fit curve is generated from the data and then input into a model factoring in the expected orbit to generate a probability of device upset.

Example Weibull Fit Curve

Example Weibull Fit Curve - AD9246S


It is imperative to perform the testing to generate this data and predict the device performance in space since, as we all are well aware, it is not very easy, dare I say impossible in many cases, to repair/replace devices once they are put into an application in space.  Obviously there are things like the ISS (International Space Station) that are routinely visited and repaired, but something like Juno which is orbiting Jupiter that would not be able to be repaired nor worth the cost required to attempt to do so. If you'd like to learn more I encourage you to check out my blog over on Planet Analog.  I hope you have learned a lot so far and found SEEs interesting.

Have you ever listened to AM radio at night and wondered how you could hear stations from great distances away, but not hear them during the day? So have scientists and researchers, who have spent decades trying to understand what's going on above us to cause radio signals to "skip" like that at night. That led to more questions, such as why a signal sent from the same location at the same time of the day - but at a different frequency - skips farther or shorter distances? Or why signals from the same location at the same time of day and at the same frequency - but sent during a different season of the year - can behave differently? Or how in some years, even with all those conditions (location, time of day, frequency, and season) the same the signal travels great distances, but a few years later will barely make it to the border? With all of these variables at play, are we able to predict the behavior of the ionosphere and its effect on communication?


Sure we can.


But no, we really can't.


Okay, so first a quick bit of history to put these seemingly conflicting answers in context. Humans have long observed - and early on exploited for navigation - the fact that a magnetized object which is allowed move freely will point north. In 1839 a scientist named Frederick Gauss postulated that there was region of space above the earth that was electrically charged. For the next several decades a lot more postulating, theorizing and experimentation went on that lead to the development of an early form of wireless communication called spark gap. The name is very descriptive. You know how sometimes you will touch someone and you both receive a small electric shock? That's basically spark gap. When early experimenters (such as Guglielmo Marconi) figured out how to detect the spark from a distance, radio communication (using a variation of Samuel Morse's code) was born.


Among its many disadvantages was that spark gap required massive amounts of power to work over any appreciable distance, and was limited to a small range of frequencies, near today's AM band. What happened next reminds me of lining up dominoes and watching them fall against each other, because the timing could not have been better for the advancement of radio and our understanding of the ionosphere. The first domino is the passing of the Radio Act of 1912. Since wireless "state of the art" in 1912 was spark gap, which produced signals below 1200 KHz (basically the middle of today's AM radio band) the Act graciously handed experimenters and ham radio operators - what Congress blithely considered to be a useless part of the radio spectrum, above 1500 KHz.


Spark Gap Transmitter at Awanui, New Zealand

Spark Gap transmitter at the Awanui, New Zealand station.  Note the blowers in the bottom right of this photo.  These high-power stations got hot.  Very hot.


Hear that sound?  That's another domino falling.


Just before the First World War the pioneering work with vacuum tubes by Lee DeForest, Walter Schottky, and others enabled the design of transmitters that produced a signal that was more than just a hash of static, it was a tone called Continuous Wave, or CW. Amateur experimenters who could not afford the extremely powerful transmitters and massive antennas of commercial and military spark gap stations found they could make CW contacts with much less power and smaller antennas. These experimenters also began to notice patterns in their ability to reach certain distances based on some of the variables mentioned earlier such as time of day and frequency. Leveraging that experience, hams were soon making record-setting connections between the Eastern seaboard of the United States and Europe, and the West Coast and Hawaii by radio. Now get ready for another domino to fall. In the beginning most hams were crowding close to that 1500 KHz Congressionally-mandated wall. It got very crowded down there very quickly so, leveraging the ever-improving frequency range of the new tubes, higher frequency transmitters and receivers were built.


Now a seriously important domino was about to fall because in the early 1920s, as experimenters moved up in frequency they found longer distance communication was easier at certain times of the day. As more and more data was collected (hams are notoriously good record-keepers, making this like a global lab experiment) the physicists put the pieces together and, in 1923, concluded that the ionosphere - postulated 87 years earlier by Gauss - was responsible.  Now a whole line of dominoes began to fall as several things happened. First of all radio, considered to be a toy in its early days, was quickly becoming commercially viable. Hundreds of local stations were being built around the country, their owners happy to sell airtime to local hardware and feed stores, evangelists, and hawkers of patent-medicines, all of whom were happy to pay to have the ears of thousands of potential converts and customers. Listener-ship skyrocketed with many manufacturers getting into the radio receiver market, including the purveyors of cheap crystal sets, which allowed folks whose homes did not have electricity to hear those broadcasts.


Crystal Set

No electricity?  No problem with a crystal set!


The government quickly realized that this new industry required oversight by an organization dedicated to taming this new "Wild West," so in 1926 it formed the Federal Radio Commission. A year later the Radio Act of 1927 was passed which, among other actions, rescinded the Radio Act of 1912. Now fall more dominoes, as those frequencies above 1500 KHz that the government had tossed away as "useless" back in 1912 were now being allocated to commercial, amateur, military and other government services. This put thousands more technicians and experimenters up and down the frequency spectrum, working with more efficient transmitters and better performing tubes, collecting tons of data on radio reception. In 1926, building on the work of theorists and physicists such as Oliver Heaviside, Arthur Kennelly, and others physicist Robert Watson-Watt first used the term "ionosphere" to describe that place where radio signals propagate or "skip" (as it was more popularly known by excited radio listeners) in the atmosphere.


What they found was that a radio station operating at the right frequency and at the right time of day could be heard thousands of miles away - even across oceans. This was not lost on some governments in Europe, where tensions had been rising since the end of the First World War. Germany and England, two nations at the forefront of wireless research, built stations that could be heard here in America, allowing them to present their views unfiltered to the world - not unlike the way Twitter and other social media are used today.  Being market-driven, it wasn't long before manufacturers of consumer radios included one or more shortwave bands in their radios. This meant that many Americans who were listening to Franklin Roosevelt's "Fireside Chats" broadcast here in the States were also able to hear Hitler, Churchill, and other foreign leaders via shortwave, thanks to the skipping of signals in the ionosphere.


 Zenith Model 6-S-222 "Cube" Radio from 1936

 Zenith Model 6-S-222 "Cube" Radio from 1936, with three radio bands


Ionospheric skip took on even greater geopolitical importance after the Second World War, as both sides in the Cold War built mammoth shortwave stations, pumping out up to a million watts of power to extol the virtues of their political system and the disadvantages of the other. Yet our understanding of the ionosphere, especially our ability to predict the behavior of signals up there, was not advancing fast enough for those tasked with the development of rockets, whether they were being designed to deliver nuclear payloads to the enemy or humans into orbit. But it wasn't just the ionosphere that was of concern. Based on observations made as far back as the early 20th century, scientists were speculating that there was a region of charged energized particles beyond the ionosphere. In January 1958 experiments aboard Explorer 1 (the first United States satellite to reach orbit) proved the existence of what became known as the Van Allen radiation belt.


Very early observations of the Van Allen belt confirmed what had been speculated by scientists, that from 93 million miles away the sun was producing a "solar wind" of electrically charged particles that pushed against the belt at various times and at varying strengths. While there was some predictability to it (such as an 11 year cycle of sunspots) the sudden appearance of solar flares were very unpredictable. So, for those of you keeping score, between the discovery of the radiation belt and the sun's impact there were dozens more variables added to an already long list (frequency, time of day, time of year, etc...) that need to plugged into any equation for ionosphere skip prediction. And what an equation it is! (Don't worry, I won't give you any more math than I can handle... which is why I am giving you none.)


At the very beginning of this piece I asked if we could predict the effect of the ionosphere on communication. My answer back then was: "Sure we can." And then I said "But no, we really can't." I hope you can see that I was not being deliberately evasive. The fact is that even today, after over 180 years of theorizing and testing and collecting data (from both terrestrial and space-based observations) physicists and mathematicians kind of throw up their hands when asked for a definitive predictive model for what's going on up there by giving us this word:


stochastic: /stəˈkastik/ adjective

Derived from the Greek, the word means "randomly determined; having a random probability distribution or pattern that may be analyzed statistically but may not be predicted precisely." What that means for us is that instead of using lined up and orderly dominoes for my ionospheric analogy I should have employed the game of Craps, because Einstein was wrong when he said God does not play dice with the universe. Not only does He play dice, but He's constantly shooting for a hard eight.


I know this from personal experience as a ham radio operator, and now that you have all the background, I want to tell you about the time I cashed in on the long odds. To start, hams are always looking to improve their chances of making contacts with other hams in far-off places (in ham parlance, we call that DX) and there are a number of online tools which provide predictive reports on ionosphere skip. Here's an example of one such report, posted on the American Radio Relay League's website, which shows what happens when all those variables we've been talking about get plugged into a tool that charts predicted activity between any two parts of the world, in this example the East Coast of the United States and Eastern Europe: 


East Coast to Europe Propagation Forecast

Source: American Radio Relay League


The vertical axis is frequency in MHz (for reference, an AM station at 1030 would be down at 1.3) and the horizontal axis is any 24-hour period at UTC (what used to be called Greenwich Mean Time) during the month for this chart. Now, here is why we have to swap our dominoes for dice, because the three different colored lines represent only the probability that at a given time and frequency two hams could make a connection between those two places on the planet. Red represents a 10% chance, green is a 50% chance and blue is your safe bet (kind of like using house money) with a better than 50% probability that those two hams could hear each other at that time and frequency. 


The chart says that If I tuned my radio to 14 MHz at 0800, I'd have a 50/50 chance to connect with a ham in Prague. If, at the same time, I transmit at 10 MHz I'd have a better than even chance. That drops to only a one in ten chance were I to be on 18 MHz. So 18 MHz at 0800 is not a good place to try to make that contact. But we also see (by following the green line) that the chances for a contact at 18 MHz improves as it gets later in the day. Come back in six months, when the northern hemisphere is tilted differently towards the sun and the chart will be different. Come back in six and half years (during the other half of the sunspot cycle) and the chart will look very different, as well.


So thank you Gauss, Heaviside, Van Allen and others for all your pioneering work. Thank you to the folks who collect data from satellites, radio beacons, solar observatories and other sources to create this and other reports. But you know what?  While they are great guides for those of us who surf the ionosphere, I have personal experience with the predictably unpredictable nature of skip and how, sometimes, even when all the charts and tables and reports say otherwise, you gotta go for the hard eight.


In 2010 the sun was at a low point in the sunspot cycle, meaning that the ionosphere was not conducive to skipping signals, especially up on the 28-MHz band. The data said that at 0200 UTC (9:00 pm EDT here in Massachusetts) there was a very low probability for making a contact with anyone more than a few miles away. After being on my radio I was ready to concede that as fact but, before I shut the radio down, decided to tune across the band just in case. Suddenly, I caught the sound of a voice, weak at first, but quickly getting stronger. As weak as his signal was, I could tell he was calling CQ - that's ham shorthand for "can anyone hear me?"  Within 20 seconds it was as if he were in the room next to me.


"This is E51JD in the North Cook Islands calling CQ and listening for a call."


Wait. What? No, that's impossible. The North Cook Islands are (...type a quick lookup on Google Maps...) over 7,000 miles away.  And the charts and tables say that... okay, just in case... "E51JD this is WB2HTO in Massachusetts, do you copy?"


North Cook Islands


"I do and good afternoon, WB2HTO this is E51JD on North Cook Island in the South Pacific. Name here is Jim..."


So I began a conversation with a location that the charts and tables said I should not be able to have. It was a quick exchange of basic information (signal reports, names, and a few other details) because just as his signal had been growing in strength it was now starting to fade.  So we quickly signed off while we could still hear each other. As soon as we were done I could hear other stations trying to reach Jim, but whatever was happening up there in the ionosphere was changing so rapidly that I'm not sure if anyone else made a contact. I had been listening at just the right time and on just the right frequency to take advantage of what the physicists call a sporadic E opening, the "E" referring to a layer of the ionosphere that sometimes acts... stochastic.


Since, like Forrest Gump's box of chocolates, you never know what you're going to get (or how long it will last) when the E layer starts acting up like this, I moved up in frequency (to give Jim a clear space to make other contacts) flipped on the microphone and called out a "CQ" looking to see if any stations would reply. No sooner did I release the mic button that I heard "WB2HTO this is VK2HOT in Port Macquarie Australia, how copy?"


This... is... impossible. Port Macquarie Australia (another quick Google search) is 2,000 miles farther away than the North Cook Islands.  And I'm talking to someone there. At night. On a band that is supposed to be dead.


Port MacQuarrie

The distance from Port Macquarie, Australia to Boston is over 9,000 miles


A quick exchange of basic information followed and the exchange ended while we could still hear each other.  Meanwhile the band continued to shift. I only had two data points - the North Cook Islands and Eastern Australia - but it looked like the skip from Boston was moving westward. You can bet I tried calling for another contact, but after several tries it was clear that whatever magic had happened up there in the ionosphere was over. Or perhaps there just wasn't a ham in the right place at the right moment to hear me.


Several weeks later I got something in the mail from my new friend. It was a QSL, or confirmation, card. They are mostly sent electronically today, but I'm glad to have this hanging in my radio shack to remind me that the ionosphere is predictably unpredictable. Or, perhaps, unpredictably predictable.



Confirmation, or QSL card, from the contact that the numbers said should not have occurred

If you've been reading my blog for the last several months you know I have been discussing a lot of radiation effects and have given an example with these effects on a high speed ADC.  Last month I started talking about single event transients (SETs).  I felt like the subject did not quite get fully covered so this month read along as I spend a bit more time on the topic before we move on to single event upsets (SETs).  I realize this topic is most important to those working with the space industry and is not often thought about with commercial products.  However, this could be shifting as process technologies shrink and devices become more susceptible to low level radiation that makes it through the Earth's atmosphere.  In addition to that there are companies like SpaceX who are working to put more COTS products into you commercial guys may not be off the hook! 


I encourage you to take a look at my latest blog continuing the discussion on SETs over at Planet Analog here: Planet Analog - Jonathan Harris - Single Event Effects (SEEs) with High Speed ADCs: Single Event Transient (SET), Part 2. Here is an example of what an SET for a high speed ADC might look like:


AD9246 SET Run 64


If you have not read any of this series of blogs yet hop on over to Planet Analog and check them out.  You can find all my blogs here: Planet Analog - Jonathan Harris - Latest Content. The radiation effects series started in my December 2017 blog. There are also lots of other topics that I've discussed on my blog at Planet Analog.  Feel free to add comments are questions there or here on EngineerZone.

Working as I do here for the global leader of high-performance signal processing solutions, I often speak or write about Analog Devices and how our semiconductor products represent the "state of the art."  The challenge for ADI (in fact, for any company seeking a technology advantage) is to advance that state of the art and make it commercially viable.  As I've learned in my hobby of antique radio restoration, the same was true in the early days of mass-market consumer radio.  The technology might be 80 to 90 years old but, as I'll show you now, we can see great examples of how circumstance and ingenuity led some companies to find - or miss - that sweet spot.


From its founding in 1892 as the Helios Electric Company (manufacturing carbon-arc lamps) then, in 1906, as the Philadelphia Storage Battery Company (making batteries for electric vehicles) and finally, in 1919, adopting the name by which we know it today, the Philco Company (when it turned its attention to providing consumers with storage batteries for the burgeoning radio industry), the company was always looking for technology that could meet the needs of the market.  For example, because those early batteries were expensive and messy (they required the monitoring and refilling of liquids) in 1925 Philco would introduce the first "Battery Eliminators."  Today we call them power supplies.  Back then it was a boon for those home owners who had electricity because now they could run their radios directly from a wall socket.  (Just a quick side note: many in rural areas did not have power in their homes and still required batteries to run their radios.  Philco continued to market their batteries to serve that market and maintain their brand so that when New Deal projects such as the TVA brought power to farms across the South, Philco was the preferred brand when home owners upgraded.)


In 1926 Philco decided to get on the broadcasting bandwagon, but it took almost three years before they released their first radio.  That's because they saw how radio manufacturers, such as Atwater-Kent, were individually (and expensively) assembling their products and, as a result, having to charge high prices, limiting the market and, by extension, the radio audience.  The planners at Philco saw how Henry Ford was dramatically reducing the cost of manufacturing with assembly lines, which they co-opted for the mass production of their radios.  According to Wikipedia, only a couple of years after introducing their first mass-produced radio in 1928, Philco was already the leading maker in the country, grossing $34 million with the sale of over 600,000 radios.  For the mass market these included the well-made, beautiful, yet inexpensive "Cathedral" model, and for customers who liked cutting edge tech they developed and sold an innovative, one-tube wireless remote, which they marketed as the "Mystery Controller."


Philco Mystery Front

The Philco Mystery Controller was the first mass-market remote


Then, came the war...


Following U.S. entry into World War Two in late 1941 the economy, still moribund due to the Great Depression, was jolted into a frenzy of war-related production.  Philco was among the top 100 providers to the government for the next four years, as they pushed aside the design and manufacture of just about any product that was not war-related.  In 1945 the war ended, and as restraints on commercial production were lifted manufacturers, Philco among them, looked for products they could quickly produce and sell to a country anxious to enjoy the benefits of peacetime.  Both of the following radios are Philco model 46-250, meaning that they were produced that first full year after the end of hostilities.  They look the same, don't they?


Philco 46-250 Bakelite

From 1946, two Philco model 46-250 radios (author's collection)


Both radios are housed in Bakelite cases with only two knobs; one was a combined On/Off Switch & volume control, the other for tuning up and down the AM radio band here in the U.S.  Inside, both have an almost identical "All American Five" design found in most tube radios built from the 1930s onward.  AA5 radios eliminated the big, bulky power transformers prevalent in the large consoles that were popular before the war, which greatly reduced the cost of manufacturing and owning a radio, something especially important during the Depression.  The "trick" to an AA5 was, first, to reduce the number of tubes needed to the bare minimum.  I'll cover this in a future blog, but for now it's important to know that in the 1920s and 1930s our improved knowledge of tubes (and what caused them to distort a signal) allowed designers to build tubes that combined two or more stages of the receiver into one tube, which helped to bring the tube count down to five.  That typically included an RF converter, IF amplifier, audio detector/first amplifier, audio output, and rectifier which enable the most cost-efficient, best sounding radio.* 


Was there a drawback to the AA5?  Well, there was that nagging problem of electric shock.  That's because one side of the power line was connected to the metal chassis, so you didn't want to touch it when the radio was plugged in and turned on.  And who wouldn't want an appliance like that for the wife and kids?  (Full disclosure: I received a couple of lessons on AA5 power supplies the hard way.  Hurt like hell, too.  You'd think I'd have learned after the first time.  Or the second. Or the... never mind.  That was one of the selling points of Bakelite, by the way - it's a great insulator.)


Okay, so now let's talk about these specific radios and how they play into the story of changing the "state of the art."  As stated before both are the same model number (46-250) but, as we see from the stickers affixed to the base of each radio, there is a difference in the code numbers.  These numbers are also called chassis numbers, which delineate versions of the basic circuit employed in each radio:


Philco Inside side-by-side

Product stickers for the two radios


The primary difference between the two chassis is that the 122 on the left uses a mix of "Octal" and "Loctal" tubes, while the 125 on the right uses only Loctals.  Octals, as you can see above, have a black base that was made with Bakelite.  They had thick pins that fit snugly into metallic sleeves arranged circularly in a Bakelite socket that was mounted on the chassis.  Octals were the workhorse of radios for decades.  Loctals were a relatively new type of tube developed by Sylvania in the very late 1930s for use in automobiles.  They had an aluminum alloy base that, as the name implies, locked into place in the socket.  As detailed in a Wikipedia page on tube sockets, Loctals had the advantage of being "pin-for-pin" compatible with the older Octal tubes.  Interesting to note that pin-for-pin compatibility is a selling point still used today in the field of semiconductor manufacturing.


Loctals are in both versions of the 46-250 we're talking about today, taking the roles of RF converter, IF amplifier, and audio detector/first amplifier stage.  But chassis 122 on our left still has two Octal tubes: a 35Z5 for the rectifier and a 50L6 for the audio amp.  This tells us that the 122 is an older model, as it is well-known that manufacturers such as Philco did not want to throw away their stock of older components.  But, as the three loctal tubes indicate, Philco's stock of old Bakelite tubes for the first three stages must have reached a point where it was financially viable to use the newer loctal technology.


As it turned out, the move to Loctal tubes is a great example of incorporating "state of the art" that sometimes fails to live up to the promise.  The tubes, according to Wikipedia, were "...prone to intermittent connections caused by the build-up of electrolytic corrosion."  And if you tried to take advantage of the pin-for-pin compatibility, you found the smaller pins of the Loctal tubes in the bigger Octal sockets had a tendency to "wobble."  So much for an upgrade.  Those problems would not surface for a few years, and Philco would use loctals in many of the home tabletop radios they built after the war.


So loctals have kind of a sketchy history, and demonstrate one challenge of developing new technology that may, at first, be considered "state of the art" but later turns out to have many flaws that diminish its marketability.  However, we will now see how the story of the audio output tube will, no pun intended, light the way forward not just for radios but all electronics.  We turn again to Wikipedia, which explains that "in 1938 a technique was developed to use an all-glass construction with the pins fused in the glass base of the envelope. This was used in the design of a much smaller tube outline, known as the miniature tube..."  The advantages of the mini tube are echoed today in the semiconductor industry where companies, such as Analog Devices, build and market products that are smaller, use less power, and dissipate less heat (primarily because the tube's filaments are smaller).  Yet, despite these reductions, they outperform their predecessors.  That may explain why, for the next version of model 46-250, Philco designers tweaked the audio output section of the radio to accommodate a next-generation mini tube, the 50B5.  Clearly there was a cost benefit to Philco to use the smaller tube, because they also had to install a mini tube socket in the place where previously an Octal 50L6 sat.  We can see that in the close-up of that portion of the 125 chassis:


Mini Socket adapter

Mini Socket adapter the Audio Output tube


Lower power dissipation and smaller tube size meant that Philco, along every other radio manufacturer who wanted to stay competitive, would abandon Bakelite cases and start housing their radios in cheaper, lighter, and less expensive plastics developed during the war.  They were not as resistant to heat, but didn't have to be.  State of the art and commerce often walk hand-in-hand, which they did in the case of Bakelite vs. Plastic and octals vs. loctals vs mini tubes.


Me?  I have a few colored plastic pieces in my collection, but I like the old school Bakelites with the bigger, hotter octal tubes.  As one of my ADI colleagues is fond of saying, "real radios glow."  You can hear the 46-250 125 chassis in action in this YouTube video.  





* The AA5 design worked (shock potential aside) because the filaments of the radio's five tubes were connected in series and the voltage drops across the tubes came close to line voltage (a number that changed over the years - there's another blog for another day, for now let's just settle on 120V as the voltage drop goal).  Since the leading number in a tube designation is the voltage drop across that tube, we can see the Philco 46-250 adds up to 106V.  Philco added a 2 watt 80 ohm resistor to the series of tube filaments to drop the additional 9V.  And yes, that meant that inside the cabinet not only did we have heat dissipating from the tubes, but also off that resistor.  That's why radio backs had large openings, to provide air for cooling.  


Philco PS with 80 ohm resistor

Close-up of the Philco 46-250 power supply showing the 80 ohm resistor

We had our most successful year as a group of ADI teams at the World Championship events in Houston and Detroit. It was record-setting, both for individual teams and for the program! We had a total of eight teams attend Championship and ADI hosted a booth again in the Robot Service Center this year at both events, where we got to talk to teams from all over the globe about what we do and what we have in store for teams next year.


ADI had four teams in Houston this year spread across three subdivisions. We were lucky to have the Robot Service Center right behind the Turing Division field, where Team 2655 Flying Platypi and Team 2471 Mean Machine were competing. One field down from us was 254 in the Hopper Division, and another two down from them Team 1577 Steampunk was competing in Galileo. I have to admit I was particularly interested in watching the Turing division teams, not just because my team was there but because... yeah, no, it was because my team was there. But what made Platypi's situation so unique was that we had zero control over our own destiny because of the robot's design. We couldn't climb, and we couldn't touch the scale. Winning or losing a match was up to our alliance partners. The only ranking point we could control was the Auto Quest. But we knew our robot could fill a niche that very few others could, and we were praying for other teams to see that value we could bring to an alliance.


After finishing 55th in the Turing Division, 2655 was selected to be part of the 8th seed alliance by our friends 1533 Triple Strange, another Greensboro team whom we've worked with nearly enough times to earn a spot in the "most successful team-ups" list on The Blue Alliance. It was going to be a long road clawing through this division as the number 8 seed, particularly for the first set of matches. To get out of the quarterfinals, we had to win against the top-seeded alliance, who was strongly favored to win. And for me, this was so difficult because this meant playing against 2471 who is also supported by ADI, and I've come to know this team as great friends in this community.


To be honest, I actually completely missed the first match. I came running up the stairs to the stands as the crowd erupted at the match score. And to my shock, the 8th seed had upset the 1st seed. As I emerged at the top of the stairs and looked to where 2655 was sitting with 1533, they were going crazy. And I looked closer to see that members from every North Carolina team were sitting with them, all cheering. I ran over to sit with the team while we waited for the next match. One field over, I could see 254 competing on the Hopper field, partnered with 148. This was an all-star alliance, the favored champions. I watched as 148 distracted teams so that 254 could load the scale with no interruptions. With the reality of 2655 advancing to Einstein looking more and more likely, the prospect of playing against this alliance became more and more real. On the one hand, I knew that realistically we probably weren't going to make it to the Einstein finals. But what if we did!?!


So match 2 between the 8th seed and the 1st seed comes up, and it was SO painful to watch. Alliance partner 1296 died in the middle of the match, 2655 threw a chain and couldn't drive. The 1st seed won the game with ease. I stood up and ran down the stairs to the pits with one of our other head coaches to see what was wrong. One of our drive team members ran up to the rail and explained that one of the drive chains had busted and fallen completely off. After a nail-biting and excruciating seven minutes, the robot was fixed just in time to head onto the field for the rematch.


And would you believe it, the 8th seed won it. I watched from the stands as the 8th seed alliance went on this Cinderella-esque journey to the Turing Division finals. I hardly remember each individual match and who did what, I just remember shouting, cheering, losing my voice, and then entirely losing all resolve when at the end of the trial we made it. We made it to Einstein. I shouted with what little voice I still had, I cried right along with the students (so, so much ugly crying, it was kind of pathetic). Hugs were going around everywhere. It was such a beautiful moment to see all of the North Carolina teams celebrate the success of these two teams as one, to watch these students I've worked with for so many long months achieve something they never thought they could do. At that point, I didn't care about any matches past that. We made it to Einstein and won our first Championship award for the team's fantastic business plan. I couldn't have asked for more.


At this point, I had no voice left. Zip, zero, zilch. I watched as our battered robots struggled on the Einstein stage, conversing via text with the drive and pit crew down on the field. When it became clear that we were locked out of Einstein finals, I looked up at the screen to look at who our final match was against. And surprise, it was against the "Black and Blue" alliance from Hopper. I only had one thing to say to the drive team before their match:


"Give them the best defense they've ever seen, and go have fun. No, actually, just go have fun out there. I don't even care about the score. It's your last regular season match, just have fun with it."


The match score was actually pretty sad. FMS sent our robot the wrong information for autonomous, so we scored on the wrong side of the switch and never fixed the switch position. The match score was left at 15 for the majority of the match for our alliance. At first, I was frustrated. Then the most beautiful thing happened, and I just about fell out of my seat.




It wasn't even planned. They just did it. Teams 148 and 1296 just decided to spin out of control because they had literally nothing else to do for the rest of the match. It was by far the funniest thing I've ever seen on an FRC field. I don't think I'll ever be that excited and tickled to lose a match ever again. The only way the season could have ended better was for us to have operational robots and defeat that alliance. But I almost prefer the way it ended because we didn't play to win the match, we played to have fun. In my opinion, that's more important than winning.


The Black and Blue alliance went on to the Einstein finals in Minute Maid Park, and they swept the finals to win the event. The Cheesy Poofs became the first team in FRC history to go an entire regular season undefeated.


Tune in next week for a wrap-up on the Detroit Championship and a look inside the Robot Service Center Booth!



This blog is part of a series covering the 2018 season of the FIRST Robotics Competition, FIRST POWER UP. Stay tuned for more updates, including coverage of the Championship Events in Houston and Detroit at the end of April! Get to know the ADI teams, learn more about our donation boards, and meet the employee mentors that make it all happen!

I’m back again to talk about my hobby of ham radio. Last time I wrote about the experimental antenna I recently built on my property for operating on the low (1.8 – 2.0 MHz) Amateur Radio bands. Today I’m going to go back to my early days in the hobby when I first got to know OSCAR. No, not the Grouch or Felix’s roommate, this was the acronym for the Orbiting Satellite Carrying Amateur Radio. 


Hams have always been at the forefront of communication technology, starting with the first wireless stations which sent and received Morse code. Hams would later experiment with some of the first AM, FM, and even television stations, so it’s no surprise that they were early in communication utilizing space. In fact, the Space Race was only four years old when OSCAR 1 was launched on December 12, 1961. OSCAR 1 was a very humble beginning to ham radio in space, as it was literally used as ballast to balance the payload of the rocket, although it holds the distinction of being the first satellite to be ejected as a secondary payload and enter a separate orbit. According to the OSCAR Wikipedia page, “…the satellite carried no on-board propulsion and the orbit decayed quickly. Despite being in orbit for only 22 days, OSCAR 1 was an immediate success with over 570 amateur radio operators in 28 countries forwarding observations to Project OSCAR.”


In 1974 after a series of successful launches of increasingly sophisticated satellites, Oscar 7 was launched, and that was followed fours later by the launch of Oscar 8. Both satellites were in low earth orbit (about 115 – 120 miles up) which meant they would only be in range at certain times and for a short period of time, but could both send and receive Morse code signals, permitting actual conversations. Both satellites acted as “repeaters,” receiving signals (the uplink) on the 2-meter band (144 – 148 MHz) and transmitting on the ten-meter band (28.000 – 29.000.)  As a youngster who grew up thrilled by the exploits of Shepard, Grissom, and Glenn it was exciting to think that in some small way I could experience the Space Program first-hand. Using a polar azimuth projection of the earth and a plastic overlay of Oscar 7 orbits (which had been included in an issue of QST, a ham radio magazine) it was easy to determine when the satellite would pass within range.


Azimuthal projection

From an edition of QST magazine.  Plastic overlays (with either Oscar 7 or 8 orbital paths) sat over this map


I set my transceiver (the rig was a Tempo One) to the down-link frequency on 10 meters and, when the timing was right, I could actually hear a number of CQs and QSOs coming from space. That was pretty cool. But I knew that the real fun would begin when I transmitted through the satellite, and so in the spring of 1977 I bought a used Hallicrafters VHF1 “Seneca” 2 meter transmitter, then constructed a 2-meter ground plane out of coat hangers, which I hung outside the shack window. Nothing fancy, as I had recently purchased a used TH3 beam for mounting on my parent's roof (these were GREAT parents) and knew eventually I would stack some sort of antenna for 144 - 148 MHz up there. Once I had the 2-meter transmitter working, I checked my QST Magazine for the satellite's latest schedule and, with my high-tech plastic overlay, saw that Oscar 7 would pass right over New York on the afternoon of June 6th, 1977.



A rendering of OSCAR 7 


I started hearing CW signals from other stations a few minutes after 1:00 pm (EDT) on 10 meters and began sending CQ (the internationally recognized abbreviation for “calling any station”) on 2 meters (remember that OSCAR was a 2 meter receive/10 meter transmit repeater). That’s when I heard… nothing. Huh. Well, since I didn't have a directional antenna I kind of figured the satellite, which was orbiting about 115 miles above, would have to be a lot closer to pick me up. So I tried again and - BINGO! - there it was, coming back to me on 10 meters, my own signal! Now here's the really cool part: as I'm transmitting and the satellite is racing towards me at 17,000 miles an hour, I could hear the tone of the signal getting higher in pitch - the Doppler shift of my signal coming back from space. It got stronger too, as Oscar 7 passed almost directly overhead, then almost immediately the tones began to shift lower as the bird raced away and the opposite Doppler shift took effect. Well, I just thought that was the coolest thing to be able to actually hear that happening. As of this blog, the experience is as vivid as any ham radio memory I have. 


Repeated attempts at connecting with another station were unsuccessful, not surprising given the transmit antenna was non-directional and, geez, made of coat hangers. Two weeks later I installed the TH3 beam (for 10, 15, and 20 meters) and a Cushcraft circularly polarized 2-meter antenna, stacked on top of the HF beam. Here’s how it looked shortly after installation. (My mom says she still gets heart palpitations thinking about me climbing on the roof.)


Kruh residence in New York, topped with two beamsThe roof of my parent's home in Merrick.

That's me, my TH3 HF beam and Cushcraft 2 meter antenna.


With both transmit and receive directional antennas in place, I was ready to go for a contact during OSCAR’s next pass.  On July 24, 1977 (Oscar 7's 12,294th orbit) I made my first space-based QSO, with a ham in Florida (WA4JID). I subsequently would connect with dozens of hams up and down the Eastern Seaboard and western Europe through Oscar 7 and then, in the summer of 1978, dozens more through Oscar 8. Yet I don't recall any of those QSOs as vividly as I do when I first heard the dits and dahs of my Doppler-shifted CW signal coming back at me from space.

In my last few blog post we've been looking at radiation effects on high speed ADCs.  We started the journey discussing TID effects and moved over to the current topic of SEEs (single event effects).  This month the topic in particular is a class of SEEs known as SETs (single even transients).  These are events that are transient in nature as its title suggests.  These types of events occur for short periods of time and do not require a device reset to return to normal operation.  With these events some sort of transient device upset is observed and after a short period the device returns to normal.  In the case of a high speed ADC this can be illustrated by a short duration where the output code is beyond a specified threshold and returns back to normal levels without requiring a reset of the ADC.


I hope you are enjoying the journey so far looking at all these radiation effects.  You can find more details on my blog at Planet Analog: Planet Analog - Jonathan Harris - Single Event Effects (SEEs) with High Speed ADCs: Single Event Transient (SET) .  I trust you will stay tuned as we continue looking at SEEs in my next blog where we will take a closer look at SEUs (single event upsets). 

It surprises some people to learn that, though cell phones offer inexpensive, instantaneous communication worldwide, there are still millions of active ham radio operators around the world. We like to say that there is something for everyone in our hobby. Some of my fellow hams are actively engaged in supporting disaster relief efforts (they were an especially critical part of the still-ongoing recovery in Puerto Rico). We have astronaut hams orbiting in the International Space Station who have set up a station to talk to us earth-bound hobbyists. Other hams design and build high-gain antennas for bouncing signals off the moon. We do these things-and more-while using methods that range from state-of-the-art digital modes (that run on our computers) to good old-fashioned Morse code.


The facet of the hobby I enjoy is operating HF, on eight different sets of frequencies between 1.8 MHz to 28 MHz, called bands, that have been allocated to ham radio. Lucky placement of trees on opposite ends of the long side of my property provided me with a place to hang a 100’ center fed dipole. Dipoles are great. They are inexpensive - they’re just wires, really, with a feed line and some rope to hold up both ends in available trees - plus they are easy to design, install, trim, and fix - and for a bonus they even have a little gain (although it’s directionally fixed because they don’t rotate.)


Here’s an aerial of my property, showing the location of the dipole and, in the inset, a picture of what is called “twin lead” (the feed line from the transmitter to the antenna) and the antenna tuner, intended to “fool” the transmitter into seeing a 50Ω load (even though the actual antenna is not resonant on the frequency.)


Aerial of my property

Using this dipole on HF frequencies I made contacts with all 50 states and over 200 countries as far away as southwest Australia – practically on the other side of the planet. However, when I operated on the lowest ham band, 160 meters (1.8 to 2.0 MHz,) performance was sub-par, since the dipole was only 25 feet above the ground, and that’s just a fraction of a wavelength on 160 meters.  That meant most of my signal was going up, not out. My longest-distance contact on the band was only about 1100 miles away.


During a lunch-time conversation with Analog Fellow Woody Beckford (WW1WW) I explained the problem. He suggested that I convert my 100-foot dipole into a vertical by twisting the leads together and feeding them into one side of a balun. A vertical would provide a lower “take-off” angle for my signal, increasing the distance that my signal would be heard. This is dramatically shown by EZNEC, an antenna analyzer program popular with hams: on the left is the model for the pattern of a dipole at 25’ and on the right side was the model for a 25’ vertical at the same frequency.  Note the lobes in the right-side model showing the lower take-off angle of RF:


Antenna model for dipole and vertical

 Because the dipole – and therefore the vertical - would be limited to 25’ height (a small fraction of a 160 meter wavelength,) we knew getting it to tune up on frequencies as low as 2.0 MHz was problematic. Using EZNEC, Woody calculated parameters for a base-loaded coil that would not only electrically extend the antenna but also eliminate the need for a balun. Here’s the coil I designed from Woody’s specs, with 38 turns of #10 wire around a 2" piece of PVC.  The wire out of the top connects to the twisted pair leads of the twin-lead, an SO-239 provides easy connection to coax from the shack, and the lugs allow for easy addition and removal of the radials:


160 meter Coil


Ah, yes, radials. You see, at this point, the antenna was quite literally only half complete. That’s because verticals need something that dipoles don’t. Vertical antennas can be said to be only "half there," the other half being a "reflection" in the ground, which means they rely on the return currents through the ground, via wires radiating from the bottom of the vertical (hence the word “radials"). So basically that means if the ground system sucks the antenna will, too. It may load up nicely but as Doug Grant - another former ADI employee and ham operator (K1DG) reminded me - so does a dummy load.


To quantify the effectiveness of the ground radial system we can use radiation resistance (R) which, in one term, expresses loss from the entire system: the antenna, the feed lines, and the radial system. We need R – and a special meter (I used an MFJ-259 antenna analyzer) to help answer a number of questions: How many radials do I need to install? How long should they be? Should they be elevated or buried? What effect will my house, my neighbors’ houses, soil conductivity (which can change from season to season) and buried gas, water, sewer, telephone, and cable lines have on R or the vertical’s broadcasting pattern? There were simply too many variables to calculate. Only testing (by laying out radials) would reveal the answers. And this is where the fun began.


With a 1000’ spool of insulated wire in hand, I began a few months of laying out the radials and measuring SWR, R, and X (which I haven’t mentioned yet, is reactance, which is the opposition to alternating current due to the combination of capacitance and inductance inherent in any antenna system), all of which we also want as low as possible. For the purpose of this blog we will focus on R, since X tends to follow R up or down and although a naturally low SWR would be preferable, I could always use the tuner to present a 50Ω load to the transmitter. The first test was with four radials at ground level, laid out at the edge of my property and next to the house, as shown in this annotated aerial:


Annotated Aerial of property with radials


The results were encouraging although at its lowest R, was above my soft target of 50Ω and gradually increased with frequency until, at the top of the band, it was over 60Ω. More radials were clearly going to be needed, but where could I put them? The answer lay in a series of articles authored by Rudy, N6LF, titled QEX articles on verticals and radials in which he details how elevated radials can be as effective - and sometimes more effective - than ground-based radials.  After reading this I ran a pair of radials approximately 3 1/2" off the ground, along the upper support of a wooden fence that ran around my backyard. The results were brilliant, with a drop of at least 10Ω across the 160 meter band.


With nothing but time and plenty of wire left from the 1000’ spool, I experimented with the addition of several radials in various layouts, including one test in which I figured “hey, if one set of elevated radials reduces R by 10Ω, why not add a second set along the middle of the fence?” Imagine my surprise when I measured R actually going up across the entire band! (Who wants to tackle the math to explain that?) Ultimately, after trying a few more layouts and lengths, I found that the addition of two more 70’ radials laid out on opposite sides of the vertical performed best, reducing R across the entire 160 meter band by about another 5. Here’s a plot of performance of the original four radials (in red), with elevated radials on top of the fence (in blue), and with the final two, 70’ ground-level radials in place (in green).


showing measured results of 4, 6, and 8 radials
After sharing these results with Woody and Doug (and asking "what should I try next?") I was given my favorite piece of advice ever: Stop, already, and just get on the air! By then it was October and, with winter coming (a time when 160 meters has optimum operating conditions) I was happy to stop digging up my backyard. Over that winter - and since - the antenna has performed way above my humble expectations, with contacts made in all the lower 48 states and over 35 countries, some with hams over 4,000 miles away in Eastern Europe, Russia, and South America - a HUGE improvement over the dipole. Well worth the effort and strange looks from the neighbors as I was digging and burying wires in my yard.


For more information, data, and pictures of this vertical project, please visit my ham station website.

As the early competition season wraps up, the season for many teams is now over. But for lots of our ADI teams, the excitement is just beginning. New England Championships is this coming weekend, and other teams across the globe are now learning if they have qualified for the World Championship in Houston and Detroit.


North Carolina

Team 2655 has had one of the wackiest and most successful seasons in the team's 10 year history. Platypi went to their first competition with a tall robot with a custom elevator system one of the students developed himself, and it was a struggle. Every match we were tweaking and re-tensioning chain, with a handful of matches ending with the robot flopped over. After the first competition, the team convened and collectively decided to re-work the entire robot from the drive base up. The elevator was stripped off, and a pneumatic arm put in its place. Sure, the robot could no longer reach the scale, but we were now blazingly fast. Platypi held their own at their second event and went on to win their event in Forsyth County near home, earning enough points to squeak into the state championship. Team 900 Zebracorns also came home with a hard fought win for the Chairman's Award at the same event, earning them admission to State as well.


Platypi rose through the ranks at the state championship, earning enough district points to advance to the World Championship in Houston. This year has been a record-breaking year for the team. This is the team's first year without full access to a well-stocked machine shop and tools, yet the Platypi brought home their first ever Robot award in the team's history for the complete redesign and rebuild. The team also took home the Entrepreneurship award twice including at the state championship.


2655 is the only ADI team of three from North Carolina advancing to the World Championship in Houston.


Pacific Northwest

Many of you will remember my interview with Quality Engineer in Camas, WA Bruce Whitefield. His team, 2471 Mean Machine, has been making quite the ruckus in the FRC community, and for good reason! This team had one of the most talked-about robot reveal videos, featuring some impressive autonomous work and a solid "buddy climb" support system which has allowed them to climb to the top ranked team in all of the Pacific Northwest District. With two event wins and two well-deserved Robot Design awards under their belts, Team Mean Machine earned their rightful place at their district championship. Team 2471 had a strong showing and went on to win the event, earning them a spot in the World Championship in Houston. Personally, I'm expecting to see this robot on the Einstein fields! Check out what they were able to do with our ADIS16448 IMU board!




Team 1577 Steampunk has long been an Israel powerhouse, and this team even made the FIRST Updates Now Network FRC Top 25 for Week 5, which takes opinions from the greater FRC community to decide who the best robots are each week. This year they finished all three events, including their district championship, ranked 1 or 2 at the event. Steampunk earned their place as the top ranked team in Israel this year and a rightful place at the World Championship in Houston.


Come Find ADI!

These teams aren't the only ones going to Houston - ADI will once again be present in the FRC Robot Service Center with gyro/IMU support and a show off some of the ways that ADI sensors you see on your robots help revolutionize the way we interact with the world. Come stop by our booth at both Houston and Detroit!


New England


Curious about our New England teams? Their district championship is this weekend! Here are all the teams you should watch for at the event! And don't forget to check back on the FIRST District Rankings website to see who qualifies for the World Championship in Detroit!


Going to the New England Championship Event...

  • 1153 RoboRebels
  • 5422 Storm Gears
  • 4909 Bionics
  • 5962 perSEVERE
  • 5735 Control Freaks
  • 4905 Andromeda One
  • 1058 PVC Pirates
  • 2342 Team Phoenix




This blog is part of a series covering the 2018 season of the FIRST Robotics Competition, FIRST POWER UP. Stay tuned for more updates, including coverage of the Championship Events in Houston and Detroit at the end of April! Get to know the ADI teams, learn more about our donation boards, and meet the employee mentors that make it all happen!


I may work at the world's leading designer and manufacturer of analog and mixed signal semiconductor solutions, but my first love was an old tube radio in my parent's basement. Growing up during the 1960s I would listen to Top-40 AM radio stations from all over the country on this old Philco. (I later had a very brief career as an AM radio disc-jockey, but the story of my misspent youth is for another blog.) Though I later got my Masters in Engineering, my colleagues will attest to my continued love for old wooden and Bakelite radios, as I have more than a few in my cubicle here on the Wilmington campus.


Like many of my fellow antique radio collectors, I always seem to have a few “someday I’ll get to it” radios; the ones we collectors buy, put on a shelf, only to gather dust as other radios get priority. So it was for me with this GE220, a post-war Superhet Bakelite tabletop that I bought for $20 at a flea market many years ago. The fact that the radio was missing its back didn’t trouble me - I liked that it had a shortwave band and was a lot heavier than later, lighter, All American Fives that would soon flood the market. It felt sturdy, almost like manufacturers wanted to assure the American public that our wartime and Depression years of sacrifice were over.



This past summer, after a nice layer of dust had settled on the rig - and with no other radios to work on I took it off the shelf and began with the basics: cleaning the piece (inside and out), a recap and new power cord. Once I had the confidence that the radio powered up safely, I then tackled the challenge of an antenna.

Now the following comment comes from a long-time marketer: it might have been a trend of the times, but GE really went overboard with the brand names. The radio itself was called a “Musaphonic,” (probably an early positioning against upstart FM which would threaten - and eventually conquer - AM radio’s dominance for broadcasting music.) But the marketers at GE went a step further, even giving the loop antenna its own name: “Beam-O-Scope.” Other models, such as the floor-model combination tuner and phonograph H-77, H-78, and H-79, were equipped with the equally impressively named ‘Super Beam-O-Scope.’” (The italics are theirs, from the Rider Manual, Volume 11.) Other models, such as the pre-war tabletop GE L-740, included another grandly named antenna called the De Luxe Beam-O-Scope.


So what is the “Beam-O-Scope?” The GE220 schematic didn’t elaborate, but in the documentation for GE’s H-7x series it explains that: “The ‘Super Beam-O-Scope’ is essentially a tuned coil antenna wound on a frame and shielded by a Faraday screen against electrostatic disturbances” (Again, looks like the battle against no-static FM has already been joined!) But, marketing aside, this left me without an antenna - one that, as seen in the schematic below, was more than just a few loops of wire; it also included a “pick-up” loop for an external antenna that required a 470 Ω resistor and .002 µF capacitor in series. Furthermore L1, the built-in broadcast antenna (the heart of the Beam-O-Scope) also had a 1.5 – 15pF variable cap in parallel.


GE220 Schematic

The loop antenna is critical, since it is part of the first tuned circuit in the radio (feeding the grid of the 12SK7 RF Amp.) If I were to create a replacement I would have to get within the right range of inductance required for the tuned circuit. Further complicating the task was that there were only four wires coming from the radio, three from under the chassis and one off a variable cap that was tied to the ganged capacitor of the RF detector stage.  While researching for other GE220 owners I found this old thread on the forum, one that had been started by another collector with a GE220 and the same issue – no Beam-O-Scope antenna and four wires coming from the radio. The thread included photos of working rigs and one with a list, wire-by-wire, of the five connections from the radio required to the antenna to complete the RF detector circuit.


Unable to find someone with an existing Beam-O-Scope to sell, I resigned myself to having to build a replica from scratch. But then I remembered that a few months earlier, at the New England Antique Radio Club spring flea market, I had spent $5 on this Philco E-808:


Philco E808

Philco E808-5


“Worth five bucks,” I told myself. “It’s not that nice-looking but I can use it for parts.”  When I went to the shelf I was happy to see that the Philco's loop antenna was there.  A quick check with an ohmmeter showed the loop was unbroken (that would have been a bummer) so I began the process of converting it into a Beam-O-Scope.

From a picture posted on the thread I counted 25 loops of wire on the Beam-O-Scope. Now, we all know there’s a whole lot of math that goes into the design of a loop so it will collect RF within the broadcast band and at just the right level to feed the grid of RF Amp (in this case a 12SK7.) That math includes many variables, including the number of loops, the width and permeability of the wire and the size of the space in the middle, just to name a few. But, with all due respect to the designer of the “Beam-O-Scope” and associated circuitry- this isn’t rocket science, and I banked that the tolerances were pretty wide and that the 12SK7 would accept signal in a range that the Philco loop, although smaller than the GE’s Beam-O-Scope, would provide.  The picture below shows the Philco antenna soon after the conversion was started, showing the .005uF fixed and 5-15pf variable caps and 470 ohm resistor and connecting lugs that I added.  I then laid a single loop of wire around the outside of the main antenna for the pickup loop.


Philco E-808 loop antenna at the start of conversion



That left just one more connection to be made: that missing fifth wire from the chassis, which the schematic showed going to the side of the main loop with the junction of the 5 – 15pf variable and C1-A (one of the three sets of ganged capacitors in the tuning section.)


I found it interesting that the person who wrote the original post on that forum had a radio that was also missing the fifth connecting wire. A design or manufacturing flaw, perhaps, that induced the wire to come loose?  Whatever the reason, it was a simple matter to trace the AVC (Automatic Volume Control) line in the radio, finding what looked like the connecting point, and securing a wire there and to the main loop of the hybrid Beam-O-Scope I had created.


GE220 schematic with AVC highlighted


What a treat to have it work the first time, as you can hear and see in this YouTube clip:  The re-assembled radio now sits in a more public place, on a shelf upstairs in the house. I cannot walk past it without feeling a bit of pride, having channeled Dr. Frankenstein to produce a working radio from the parts of two. It’s ALIVE!

I've been covering radiation effects over my last few posts and have transitioned into a discussion on these effects with high speed ADCs.  It is a logical progression for me considering my time here at ADI with the high speed applications team for high speed ADCs.  As I looked at these parts and thought about my current role as an applications engineer for the aerospace team I thought it would be great to talk about radiation effects with these products.  In my last installment we looked at TID (total ionizing dose effects). 


Recall that the main idea there was to do a pre- and post-irradiation test on the ATE.  This gives us information on any long term effects experienced by the ADC when subjected to radiation for an extended period of time.  The irradiation is basically like an accelerated life test that exposes the part to a certain amount of radiation to gauge the part's expected performance over its life in space when exposed to radiation repeatedly. 


In this month's installment on Planet Analog I discuss the testing for single event latch-up (SEL) on a high speed ADC.  You can find my blog here: Planet Analog - Jonathan Harris - Single Event Effects (SEEs) with High Speed ADCs: Single Event Latch-up (SEL).This concept is much like you'd expect in typical testing for device latch-up however, in this case the latch-up would be induced by the irradiation.  This is a test that is typically done before other single event effects testing because it can be destructive.  There are ways to mitigate the risk and use additional circuitry to detect and prevent a latch-up condition but in many applications a product with latch-up at lower LET values is undesirable. The additional circuitry takes up space and uses power as well as increases cost so it is not preferred but can be implemented in some cases. 


As an example in my blog this month I look at the AD9246S test report that shows the device performance for SEL.  To look for SEL the device power supply currents are monitored to look for a sudden increase in current during the irradiation of the device.  The AD9246S performs really well and exhibits virtually no change in supply current for the test.  I've included an example plot below showing the AVDD current for the AD9246S during the test where the device is exposed to an LET of 80 MeV-cm2/mg out to a total fluence of 1.0E+07 ions/cm2. Take a look at my blog on Planet Analog to find out more details.  Stay tuned also next month as I continue to look at single event effects for high speed ADCs.


I have been covering the basics of radiation effects over my last few blogs on Planet Analog and this time I discussed a bit about TID effects.  These are a little simpler to evaluate as the test procedure is relatively simple.  Of course, I say it is simple because the product test engineer has done a lot of the work for us already.  When checking a high speed ADC for TID effects the product is tested pre- and post-radiation exposure on the ATE.  A control unit is also tested for a comparison baseline.  In this case a control unit plus 4 units exposed to radiation were tested. The idea in this type of radiation analysis is to check for performance shifts in the ADC after it has been exposed to a specified amount of radiation.  In the case of the AD9246S in my blog the product was tested out to 100kRads.  As you can see below the performance has very little shift after radiation exposure which is exactly what you'd want for a space application.  Ideally there would be no shift, but in the real world things are rarely ever ideal.  However, in this case it is about as close as you can get and the AD9246S shows it holds up well to radiation for the case of TID.  If you'd like to read more please hop on over to Planet Analog and check out my blog: Planet Analog - Jonathan Harris - Total Ionizing Dose (TID) Effects with High Speed ADCs.


AD9246S HDR Report AC Parameters Plot

Every year I always look forward to bag day. It's usually filled with excitement, scramble, and way too much coffee way too late in the evening. But any way you cut it, Bag Day is the culmination of a rough 6 weeks of nonstop work.


On Team 2655, it also means some combinations. After being dared by a student last year on Valentines day to try dipping my donut in a bowl of queso, and subsequently discovering it actually tasted mildly like a cream cheese Danish, I accidentally started the most polarizing debate on the team to date. Several students requested I bring the odd pairing back again for bag day and, well, the rest is history. I'm kind of hoping it becomes this joke of a tradition. Another student tried dunking apples in honey mustard this year. Some people will try anything. Hey, an engineer has to keep an open mind, right?



Weird foods and sleepless nights aside, it's fun to see lots of cool robot reveal videos start to flood YouTube after being up past 1am last night. We've already spotted a few instances of our donation boards making appearances on robots. We're also seeing reveals from our other ADI teams across the globe. It's such a fitting time for bag day this year, since it happens to line up with National Engineer's Week. Seeing the work we put in culminate in a running robot makes all the blood, sweat, and tears more worth it than you can imagine.


Here's a team with a great cameo from our ADIS16448 IMU board about 30 seconds in.



Here's a great robot from ADI team 2471 Team Mean Machine from Camas, WA. Bruce told me about a cool tip-over avoidance/correction application they're trying to implement using the ADIS16448 IMU to help their driver out with controlling this beast of a robot when it's fully extended. Don't tip the robot guys! That thing is scary!



As more videos trickle out, I'm getting more and more excited for the competition season to get rolling. A couple of our ADI teams competed last weekend at the official Week 0 event, including 1153 Walpole Robo-Rebels and 4905 Andromeda One. Both even advanced to the playoffs. Looks like we're in for an exciting season across all of our ADI teams this year.




This blog is part of a series covering the 2018 season of the FIRST Robotics Competition, FIRST POWER UP. Stay tuned for more updates, including coverage of the Championship Events in Houston and Detroit at the end of April! Get to know the ADI teams, learn more about our donation boards, and meet the employee mentors that make it all happen!


2018: A Space Odyssey?

Posted by J.Harris Employee Feb 14, 2018

Now that I am working in the space products group at Analog Devices I pay much more close attention to various events occurring that are related to space and its exploration.  We are now just a few short weeks into 2018 and we have already had some very cool events some of which are pretty historic.


To start off 2018 on January 1st we had a full super moon.  If you do not know a super moon is termed as such because it is closer to Earth and appears brighter than usual in the night sky.  Since this particular one was the first full moon of the year it is often termed the Full Wolf Moon which is a term that goes back to Native Americans since they would often hear howling wolves outside their villages in the month of January.  This was not necessarily a historic event but turned out to be a precursor for some cool things to come. Only one day into 2018 and we had the first full moon and it was a super moon...pretty cool!


At the end of January we had the Super Blue Blood Moon with a lunar that is a mouthful!  Talk about the stars (ahem moon) lining up!  This was quite an historic event to have a super moon that was also a blue moon (second full moon in a month) and on top of that a lunar eclipse!  What a nice cherry on top! Take a look at a screenshot I grabbed from NASA's live stream of the event.  What a beautiful site to behold!


2018 Super Blue Blood Moon Lunar Eclipse


If you were up early enough on January 31st I hope you were able to catch this event on the live feed.  It wasn't terribly early on the east coast (start time was about 6:50AM) but was quite early for the west coast folks (that's 3:50AM).  The best view was from the west coast and thankfully NASA had the live stream for folks like myself who did not have a great view locally.If you missed the whole event, don't worry, you can watch a nifty time lapse from Griffith Observatory here: Super Blue Blood Moon Lunar Eclipse Time Lapse.


There will also be another lunar eclipse in late July of 2018.  It won't be lined up with a super moon or a blue moon however, but is still very neat to behold.


Next up we had an historic rocket launch when SpaceX launched their Falcon Heavy on February 6th.  The rocket is the most powerful operation rocket by a factor of two.  That is pretty amazing in itself.  The plan was for the Falcon Heavy to launch, send its two secondary boosters back to Earth landing them two platforms, subsequently send its main booster back to Earth landing it on a mobile platform in the ocean, and ultimately send "Starman" out in an orbit that would take it around Mars.  Starman is a SpaceX space suit riding in a Tesla, how is that for cool???  You can see more about the plan via the animation at


SpaceX Falcon Heavy Launch Animation


You can also see the video of Starman riding in his Tesla through space at the same link on the SpaceX website.


SpaceX Starman

Finally, if you want to see the real thing you can also see that.  The video is quite long so plan ahead, but it is really neat to watch.  I won't spoil the outcome for you so you'll have to go check out the video to see if SpaceX was completely successful with the launch.  I will say that you won't be disappointed!


SpaceX Replay of Falcon Heavy Launch


There are many other neat space events to see coming up in 2018.  For a pretty good list of events you can go to the calendar provided by over at Space Calendar 2018 - Rocket Launches & Night Sky Events


Another neat place to find a list of must-see events in 2018 is here: Top 8 Must-See Sky Events for 2018


I'd encourage you to take a space odyssey in 2018 and see all the cool events occurring in the sky overhead. Take a break from looking down at your computer all the time and look up and be mesmerized by the incredible sights to behold in the skies above us!

One of the blogs I follow regularly is that of John V-Neun from Team 148 the Robowranglers, a team that many in FRC are very familiar with. People like to think that all of the powerhouse teams just magically come up with a robot that works, or, heaven forbid, have a robot that is designed/built by the mentors and not the students themselves. No struggles, no difficulties, it just works. But if you read JVN's blog, you find that this just isn't the case. Higher caliber teams face many of the same struggles that even the most basic team has, and these struggles often parallel those we face in the engineering world.


JVN's most recent blog last week talks about the constant "two steps forward, one step back" pattern that is inescapable when working through the development process, and how disheartening and frustrating it can be. You show up to work or to the meeting with exactly one goal to accomplish for the day. And by some stroke of horrible luck, that one singular task does not get done, or worse, fails spectacularly. For JVN and his team, this was exactly what happened on Day 31 of the build season (for those not counting, that was February 5th). And believe me, both 5679 and 2655 have had their fair share of Day 31's this build season. I've definitely had my fair share of Day 31's in the 3 years I've been with ADI.


Think of all of the great accomplishments that we have seen in our lifetime... We have seen the revolution of IoT. We've witnessed the dawn of a new age of space exploration. We put the Juno probe in orbit around Jupiter! We saw SpaceX successfully land two booster rockets simultaneously, side by side. We sent a TESLA ROADSTER...TO MARS...JUST because we COULD! Think what would have happened if anyone in any of those accomplishments would have just given up after having a Day 31. None of these things would have happened. (JVN points to the countless failed rocket return landings from SpaceX before they had any successes, just as an example.)


One of the most important lessons any engineer can learn is to stare at failure in the face, laugh, and carry on looking for another solution with your head held high. If FRC doesn't teach that lesson to students I don't think there's a program around that will. It's the proverbial trial-by-fire of Day 31's. But darn it if all of those struggles don't make the successes that much sweeter when they do come.




This blog is part of a series covering the 2018 season of the FIRST Robotics Competition, FIRST POWER UP. Stay tuned for more updates, including coverage of the Championship Events in Houston and Detroit at the end of April! Get to know the ADI teams, learn more about our donation boards, and meet the employee mentors that make it all happen!