Skip navigation
1 2 3 Previous Next

The Engineering Mind

121 posts

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.

 

#VictorySpins

 

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.

 

OSCAR 7

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!

 

 

Israel

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.

 

GE220


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 Antiqueradios.com 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 antiqueradios.com 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: https://www.youtube.com/watch?v=kooZwRG5C_0.  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 rather...interesting...food 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!

J.Harris

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 eclipse...now 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 www.spacex.com/webcast.

 

SpaceX Falcon Heavy Launch Animation

www.spacex.com/webcast

 

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

 

SpaceX Starman

www.spacex.com/webcast

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

www.spacex.com/webcast

 

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 www.space.com 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!

I started talking about radiation effects last month in my blog on Planet Analog.  In that blog post I looked at the cumulative effects that are observed.  In my latest installment this month on Planet Analog I discuss the different single event effects (SEEs) that are recorded when a device is exposed to radiation. This is a very important aspect of working with space products.  It is imperative to know what the performance of a particular device will be when exposed to radiation.  As you can imagine there is no shortage of radiation when outside the protective atmosphere of Earth.  There is a good bit even here in terrestrial applications but not on the order of what a device can be exposed to in space.  Think about how much radiation that Juno is exposed to while orbiting Jupiter...I'll give you a hint: its a LOT!  This is the reason we do such extensive radiation testing in cyclotron facilities. 

 

As I mentioned last month I had previously worked with ADCs so it was a natural fit to provide insight to the radiation experts in our group when looking at testing ADCs.  It is quite fun to be able to learn new things (radiation effects) and apply previous experience (from ADCs).  I hope that you enjoy learning some things about SEEs in my blog on Planet Analog, A Quick Overview of Radiation Effects: Single Event Effects. Stay tuned as I continue the series and explore how these different effects apply to high speed ADCs.

Before I begin this week's blog, I wanted to let you all know that jchong made some major updates to the IMU library code on GitHub. There is now complete support for all three official languages (LabVIEW, C++, and Java) and handy guides specific to each language. Check out the most recent drivers and user guide here!

 

If you have been following this blog series for a while, you likely know all of the teams that have enjoyed support from ADI employee mentors and sponsorships in the past. But with the merger with LTC last year, we were able to bring a few new teams to the list. This week, I talked some more with Bruce Whitefield, a Quality Manager at the Camas facility in Washington.

 

I had the pleasure of meeting Bruce in person at the Houston World Championship last year and snagged a cool team t-shirt (in exchange for a Platypi t-shirt of course...t-shirt trading is apparently pretty big in the FRC community, particularly at World's). Is this not the most clever use of a team name ever? And I thought 2468 Team Appreciate was a cool team name/number combo.

 

24 hours, 7 days a week, 1 build season24 hours a day, 7 days a week, 1 build season

I may or may not have been wearing mine a lot this build season...

 

Bruce is one of the core mentors for 2471, Team Mean Machine. He started off helping his son, who was one of the founding student members of the team. His son has since graduated high school and college as a Mechanical Engineer. Now, he designs 3D printers. Bruce jokingly said "I'm still playing around with robots!"

 

As mentors, we typically get to see a lot of development during the four years a student is on the team. They rarely leave the team the same person they were when they came in, and it's one of the things I love most about mentoring. Their successes are our successes, and their struggles are our own. I asked Bruce what the most amazing thing he's seen one of his students accomplish and he told me about one of his students that went on to become a Dean's List winner at the 2014 World Championship. To be selected as part of that small group of students is an amazing feat. It typically brings extra scholarship opportunities and makes a nice addition to college resumes, sure. But the students that get selected usually went to great lengths to help out in their communities, and do robotics, and keep their grades high. It's a special caliber of student, and they are usually the most amazing to watch do their work on a team.

 

Like any job, being an FRC mentor has its rough patches. I've had my fair share in just the 3 years I've been doing this with 2655. (Bruce has done 11 years so far!) However though it gets, though, there's something that keeps us coming back for more even when we get to Week 3 of Build Season and wonder "Why did I decide to give up life outside of work for this again???" For Bruce, it's watching what students are able to accomplish because of the impact that being on the team had on their lives. "The growth in learning and capabilities and confidence is amazing," Bruce said, "They launch out of high school on a much higher trajectory because of their FRC experience." One of their lead mentors bad to pull away in the middle of build season last year due to a death in the family. "The next week we had four team alumni show up at the shop saying 'We're here to help,'" Bruce told me. If that's not an indication that their lives had been so touched by this program, I'm not sure what is.

 

In the spirit of last week's article, I asked Bruce if there was anything he had learned by being a mentor that he might not have learned otherwise. "Quite a bit about CAD software, machining and CNC tools that I never would have had the opportunity or motivation to explore on my own," he said. "You don't learn until you have to teach it."

 

His favorite story though?

 

"I often remember the freshman who was watching from the periphery of the action on a preseason project. I handed him a drill and pointed to where the hole needed to be. He looked at the drill and then at me and asked incredulously, 'You mean you guys will let me drill a hole?" Turns out he had never held a drill before. I said, 'You can drill them all!' After we got the drill pointed the right direction, he proceeded to do so wih the biggest smile I've ever seen. Simply being trusted to touch the equipment started changing his life."

 

 

-----------------------------------------------------------------

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!

 

TheFeminineEngineer

When Mentors Learn

Posted by TheFeminineEngineer Employee Jan 15, 2018

Most of the time, FIRST is a fun way for me to volunteer my time to help create the future generation of engineers and to better the lives of young high school students in my local community.

 

But then there are those special days when I actually learn more than the students did. And on Saturday, it was one of those days.

 

When mentoring, I tend to gravitate towards the students that are struggling the most. Maybe I see my own timid self in them, maybe I just find joy in helping them achieve something they thought was impossible to surmount. Whatever the case may be, this year that resulted in taking charge of the CAD design team and guiding the climber subsystem team through the prototyping and designing process. Our climber team is in sort of a bind at the moment. They know exactly what we're doing for our climber and the concept is one we've stolen from last year's climber, so not much prototyping is involved. And since our mechanism relies on the cube lifter team's mechanism to be complete before we can even begin to assemble it, they're feeling pretty unproductive at this point.

 

This is where I would normally be able to say I gave them some side project and hooray everyone felt better and everything was rainbows and unicorns. But that would be lying. I'm not even 3 years into my career and about as many years into mentoring a robotics team, so I have exactly zero experience with managing a team. I tried bringing the team to the local tool shop to get some inspiration on how exactly to execute the part of the climber that actually grabs the rung, looking at different hooks, carabineers, and latches. I tried getting them all to Google some ideas. We watched videos of past years with similar climbing field elements. But nothing I did would get them thinking. Needless to say, by the end of Saturday our entire subsystem team left feeling uninvolved, unenlightened, and perhaps even dejected.

 

I was beginning to feel pretty terrible. Another day gone in the build season and we had accomplished just about nothing in our subsystem team other than "Look! I found last year's climber over here!" and "Let's use Velcro to attach it to the shaft again." Even Juan couldn't get them motivated to up and do something. The only part of the day in which they looked engaged was when we went to the machine shop to actually help the intake team with prototyping because they needed more than four hands. Why was I having so much trouble getting them excited about their system? And how much longer could they flounder before they lose interest in the team completely?

 

The nice part about mentoring an FRC team like ours is that there is usually a wealth of knowledge to be gained from other mentors, be it others on your team or others across the globe. Luckily, I didn't have to look far. I decided to turn to Kevin, an ADI engineer with plenty of experience leading teams between managing a team of engineers here at ADI and his numerous years as a Boy Scout leader.

 

He explained to me that when you are working with a team, regardless of what skill the task at hand requires, they will typically fall into one of four quadrants, and each quadrant requires a different coaching style for them to be as successful as possible.

 

Confident Competence - When you can give someone a task and they can go run with it without you worrying about them making too many mistakes along the way, that person is probably in this quadrant. These are the people that dive in head first and can be autonomous enough (and knowledgeable enough) to return to you once they've solved the problem at hand. Through most of my time mentoring the team, I've dealt with team members like this. These are the kids that are typically still trying to solve robots problems at home when the meeting is over, or that one programmer that decides to write a program to test the program he wrote because he didn't feel like typing in a bunch of random possible inputs. Very few people actually start here right out of school or when they first join the team.

 

Confident Incompetence - They'll dive in alright! But they'll ask you for a map and a GPS. Maybe they need a textbook or some other resource to learn more before they can start, assuming they know where their knowledge is lacking. If they are blissfully unaware of their knowledge deficiencies, they might charge forward so fast that they just run straight off the cliff because they thought they didn't need a map. Simply put, these people will have the confidence to jump in but they won't know where to start or might start in the wrong place. They need a little coaching to bring their knowledge up to speed, but once they're there you can let them go.

 

Competent, Lack of Confidence - To be honest, this is where I started in my career. Sure there were things I didn't know, but for the most part I knew what I was doing. But at a certain point, I realized that my input to a discussion was just as valid as the engineer with 20 years of experience sitting next to me and that it was worth speaking up about. People in the Competent Lack of Confidence quadrant know exactly what needs to be done. But they are too afraid to jump in and do it. This is where I feel like most new engineers tend to start. And to get them going you'll need to help them recognize their strengths before they truly blossom.

 

Lack of Competence, Lack of Confidence - This is where most rookies fall, and in many areas new engineers as well. They don't know where or how to start and usually either don't know enough to know where to ask for help or are too afraid to ask for help. Much like the Competent Lack of Confidence quadrant, they're quiet, and rarely attempt to add any input in a group conversation even if the room is dead silent. They require the most guidance to get them to the point where they can be autonomous. These are not the people whom you can just give a task and they will go run with it. This is where half of the Climber subsystem team falls, as half of them are brand new to the team.

 

That was my issue. I was trying to get these students going like they were in a Confident Competence quadrant when really the majority of them lacked both confidence and knowledge to get the job done on their own. Of course anything Juan or I tried wasn't going to work! With that in mind going into tonight's meeting, I'll have to come up with something different to do with them to keep them engaged.

 

It's situations like this that are the reason why I encourage everyone I meet to try mentoring a team, whether they have technical skills or not. Technical skills can be learned outside of robotics. But these soft leadership skills can only truly be learned with experience. It's much safer to try and fail in leadership on a robotics team than on the job. I really can't say it enough times, sometimes I think I learn more than the kids do on the team.

 

 

-----------------------------------------------------------------

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!

It's that time of year once again as thousands of high school students across the globe begin the scramble to design, build, and test a robot in 6 weeks. This year's challenge has just been released this morning and I'm just as excited for FIRST Power Up as a dog is for tennis balls.

 

You are doing doggo a heckin excite there fren!

Throw the ball already!

 

You can watch the game reveal video on the FIRST Robotics Competition YouTube page!

 

Rather than talk about the teams today, I wanted to take some time to introduce everyone to what ADI has to offer to give your team an advantage, and why you want to use them!

 

ADXRS450 Gyro Board - Available on FIRST Choice and for Purchase on AndyMark

This is the same gyro board we included in the Kit of Parts last year. Code to use these boards is already integrated into the WPI library, so you can plug it into your SPI connector on the RoboRIO and get started right away. We talked to several teams at Championship and saw that many had great results during autonomous using them last year. Using a gyro is an easy way to determine which direction your robot is pointing, making it great for performing turns during autonomous or keeping your mecanum drive robot facing in the correct direction while strafing. The ADXRS450 gyro board is a great place to start for teams that are brand new to the world of gyro heading on their robot.

 

ADIS16448 10-Degree-of-Freedom IMU Board - Available on FIRST Choice

Back again on FIRST Choice this year is the ADIS16448 industrial grade inertial measurement unit (IMU). If you really want to get precise motion control, the IMU is the way to go. Each unit is calibrated individually at the factory, meaning more precision for your robot. High vibration environments (like being mounted on a competition robot with motors and pneumatics and crashing into walls) can also lead to significant errors on less expensive gyros, however the ADIS16448 happens to excel at rejecting this influence - it was designed for industrial grade UAVs after all! All of that adds up to superior performance when it comes to helping your robot know where it's pointing. Code is available on github here courtesy of jchong and there is even a handy installer available for teams that are using LabVIEW.  In addition, we know teams have had issues with this board in the past, and we are pleased to announce that with the newest RoboRIO image, teams can expect improved performance from the IMU compared to previous years. This applies to all of our IMU boards regardless of when they were obtained.

 

The metric most teams like to use when selecting a gyro for their robot is drift. Let's explore what that means for your robot.

 

Most gyros, including the ADXRS450 and the ones inside of the ADIS16448 IMU, produce an output that has a unit of degrees-per-second. It's a measure of the rate of rotation in a particular unit of time t. The higher the number, the faster your robot is rotating. A gyro does not simply output the degrees you've rotated your robot. A gyro will take hundreds of readings over the course of a given time period, indicated by the purple, red, and green lines in the figure below. If you were to add each of those readings together, you end up with the actual change in angle that occurred during that set time period (one second in the example below). The calculus term for this is integration. For those who have taken calculus (or those of you "older folk" that remember your calculus classes "back in the day"), this is the same as calculating the area under the curve. This math is done for you in the WPI library and in the code linked above for the IMU.

 

When you first turn on the robot, the orientation you place the robot in will be set as 0 degrees. Think of this as your robot's idea of "due-north" or straight forward (for example, towards the opposing team's driver station).

 

Top is the front of the robot in this case

 

However, since no gyro is perfect, that direction will drift over time. Remember how we have to obtain the actual direction in degrees by adding every single sample together? Over time, every little error in each individual reading begins to add up. These errors could be from any number of different internal and external sources. (If you're interested in learning more about all of these error sources, you can check out this article.) The biggest contributing factor to this drift is the gyro's Bias Stability spec, which is a measure of how stable the gyro measurement is over a long period of time. If you were to leave your robot powered on for a full hour without moving it, the gyro value would read something very different from zero at the end of that hour.

 

Again, the top is the front of the robot

 

This is what teams typically refer to as gyro drift. Once you take into account the different error sources mentioned in the article linked above, teams can expect the IMU to drift somewhere between 20 and 30 degrees per hour. At first glance, that might seem like a lot of drift! But if you look at what this means over the course of a match, we see this drift is actually not so bad. If we assume your robot stayed powered on for a full 7 minutes before the match even started, your robot will think it's heading has only changed by about 3 degrees. By the end of the match, that number is closer to 4 or 5 degrees, which is much smaller and more manageable.

 

Now that all of you are experts in gyro drift (right?), you are armed with the knowledge to make the most epic autonomous routine ever!

 

This year we want to give a huge shout out to Samtec for providing the connectors on both of these boards as a donation to FIRST Robotics! Samtec offers a wide range of connector solutions, ranging from simple digital/analog I/O to RF and even optics. They have been a long-time partner for the iSensor product family, and we're excited to have them on board with this year's donations! You can check out their website here and their FIRST Kickoff blog post here.

 

Are you ready to POWER UP? Share your thoughts about the game reveal below or how you would solve this year's challenge. Be sure to follow The Engineering Mind for updates throughout the robotics season!

 

 

-----------------------------------------------------------------

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!