 Okay, so this next talk by B. Dale is about open avionics for high-powered model rockets If you want to ask any questions, please wait until you get a microphone. So peace love and rockets Thank you very much and thanks to all of you for hanging around to hear this sort of Last real talk session on the last day always is a challenge to get the energy level up for This really is just sort of a something I do for fun that other people seem to find interesting to kind of talk Don't look for any hugely deep social significance or Debian project, you know release schedule insights or anything like that I've been asked about the title. This is actually the the Model of our local rocketry club in Colorado Springs and in fact, that's not something I put together in the GIMP That's a photo taken of one of the club t-shirts used as a background here I've been accused over the years I think pretty accurately as having as my principal hobby turning all of my other hobbies into open source projects and This is sort of another example of that out of curiosity how many of you have seen me talk about Rockets or satellites or something like that before I Went back and looked in my notes and the first deb-conf I ever went to was deb-conf 2 in Toronto Just after I was elected Debian project leader and at that time I gave a talk that was explaining how it was that I had come to be involved in Debian and One of the things I pointed to was the fact that it was an amateur radio satellite project that had caused me to first discover Debian then I think it was in Brazil which would have been deb-conf for That I gave a talk about the satellite project that I had by that time Gotten essentially to completion And then several times since then I've talked about Satellite or rocketry projects is is sort of fun things on the side for the last couple of years My son and I have been very involved in Playing with model rockets. My son is nine years old and you'll get to see a photo or two of him Later in the stack, but what I really want to do this afternoon is it's give you sort of a gentle introduction to this Model rocketry hobby at least as we practice it in the United States Talk a little bit about the role of avionics, which are electronics designed for use inside these sorts of hobby rockets And talk about some of the products that already exist out there that I've had Some personal experience using and then I want to introduce this project that I started called altos metron When my daughter is in the audience as a latin student she gives me a hard time about Mixing genders and tenses and things, but I like it this way so tough I'll talk a little bit about some of the tools I used because I'm one of those Folks that in order to really have fun at my hobby. I find myself often Doing things the hard way in order to make a point and this is another one of those times where everything I used For tools were open source as well. I'm gonna talk a little bit about where we are in some future plans So what's this whole basic idea about well? We like to build launch and successfully recover Hobby-sized models of rockets There are a bunch of elements of this hobby that I think make it really interesting first of all It's another place where there's sort of an ongoing learning process every time you build another rocket and try some new stunt You learn something in the process and as my son says it's loud sign He has a t-shirt in fact that says, you know Be quiet. We're practicing loud science or something like that I really like things where there's sort of a progressive level of technical achievement By which I mean the more time you spend and the more things you do the more you learn the more you can share with other people and The more sort of sophisticated things can get there is a huge element of craftsmanship involved in designing and building the actual rockets you'll see a Couple of photos of things being built in a few minutes and then I really enjoy the camaraderie that comes from collaborating with other Enthusiasts in local clubs and at regional launches The spirit among people in this hobby is actually very similar to the spirit that we experience here It's something like debcon a lot of people all interested in the same thing working together to learn more and have a good time and Every time people in the rocketry world start to do more sophisticated things They end up incorporating electronics and software Now there are several different categories of participation in hobby rocketry And even though there are lots of different things you can play with like Rocket launched gliders and radio-controlled planes and stuff like this the way that people Usually delineate the different classes of participation is almost entirely on the basis of how big the motors are I Guess that's not surprising. You know big motors big rush Big money all those things kind of go together We talk about model rockets as being the little rockets I don't know how many of you had the opportunity But some of us as kids played with Estes or quest model rocket kits put them together and flew them with little Pre-manufactured motors out of curiosity how many of you have had a chance to do that at some point in your life It's pretty cool. It's actually I mean I don't hopefully you would all agree. It's a fun thing to do We generally talk about this category of model rockets as being everything up through about D motors the letter prefix in the motor classification is a Rough indication of how powerful the motor is a motors in the sort of Estes nomenclature up to two and a half Newton seconds of total impulse and Each letter prefix is a doubling so by the time you get to D You're talking, you know enough power to carry a couple hundred gram rocket up a few hundred feet Those are typically rockets. They're sort of in the 10 to 24 millimeter motor diameter The rockets themselves are usually a little bigger than that There's then a category called mid power rockets, which is where my son I think currently has the most fun. These are rockets that can easily be you know two three feet long a Couple of inches in diameter, you know, however many centimeters that works out to and or in the sort of EF and G motor range It's possible with a G motor and a 29 millimeter motor casing to carry a rocket that weighs I don't know 500 grams or so to carry that up eight or nine hundred meters above ground level so That's that's getting to be fairly serious toys It's not the kind of thing you do in your own backyard very often These are the kinds of things you tend to go to some sort of an organized club launch for But then the things that I've been spending the most time with are high powered rockets And these have motors whose diameters are anywhere from about 29 to 150 millimeters And they start in 160 Newton seconds of total power and go up from there They're different levels of certification and in order to buy and fly this kind of commercially prepared motor You have to be certified and for some of the larger ones if you're going to store them overnight in the u.s At least you have to get a license from the a From the Federal Bureau of Alcohol Tobacco and Firearms So I am in fact a licensed user of low explosives with the Bureau of Alcohol Tobacco and Firearms I have an explosives magazine on my property in Colorado and the ATF gets to come and inspect it every once in a while In fact, I'm told with the upcoming Democratic National Convention being in Denver that I should expect an extra inspection There's then a whole another category which are research or experimental rockets sometimes called EX rockets And what differentiates these is that they're using homemade motors and not commercially produced motor Reload kits. There is a significant difference in the risk That's involved when you start building motors yourself and part of the reason those whole hobby rocketry thing has been so wildly successful in recent years is that the Consistency and predictability of behavior of the commercially produced motors has gotten to be really quite good And therefore it's reasonably safe to to build and fly these things even when their kids involved So, you know, it's it's on the other hand my daughter is Now interested in working with me to make some motors. I don't know. We'll probably play with that too over time This is a photo from a while back of Some of what I guess at the time it was sort of the family portrait of the various rockets that my son And I had in our possession at that time the biggest one on there Which is the one that's yellow black and red is a 98 millimeter Airframe with a 38 millimeter motor mount a little red one to its right is my son's favorite rocket of that era which is It's a 2.6 inch diameter. Whatever that works out to 50 60 some millimeters with a 29 millimeter motor mount And as you can see we had a bunch of little sd's ones There's an interesting one over in the right somewhere. It's hard for me to pick out from here That's interesting because it was actually made out of rolled paper And I mentioned the craftsmanship these are some photos of my son about a year or so ago Working on one of the rockets sanding fins marking a body tube As you see this particular rock it's sort of interesting down a lower right. It's got five fins There's something Optically strange about trying to get the fins aligned right when there's five of them Somehow our brains are used to dealing with things divided in three and in four And when you try to figure out if they actually look straight or not when there's five at least for me It was a real challenge, but that turned out to be a rocket that he really likes It's called a sky torpedo with a 24 millimeter motor mount It can fly on either the largest sd's black powder motors or some of the smaller composite propellant motors Another thing that I really like when I'm doing hobby things is to combine hobbies and One of the other things I like to play around with is computer-controlled Milling machines and related Metalworking tools. This is a small tabletop milling machine that I'm using to cut centering rings For centering the motor mount tube in the body tube for a custom rocket. These are being cut out of Seven or eight millimeter thick plywood using a I guess it's about a two and a half or three millimeter cutting bit And there are a couple of interesting things that you can do when you cut centering rings yourself For example, you notice there's some slots cut in the ring This is so that the fins which are being mounted through the wall of the outer tube can be locked into the fins And I get perfect alignment of the fins without having to stare at them very hard and figure it out And there are also a couple holes that have been drilled in there for mounting hardware since this is the aft ring that goes at the Back of the rocket We put some mounting screws in there so that we can screw down the motor assemblies and hold those in place when we're flying and As a consequence of this I've actually ended up writing a Python library that emits gcode, which is sort of the assembly language of machine control and so when I go to Create a new rocket. I sit down in a few minutes I can whip out a Python script that will emit gcode to cut the rings with however many fins and the different Diameters and so forth. There's nothing terribly complicated about this But I have to tell you my friends in the rocketry hobby just drool when they see me walk in with you know It's no it's not packaged because unfortunately this is one of those classic examples of something that's very specifically tied to the particular Little milling machine which in since I took this picture. In fact, I've reworked this and to have a much stronger motor It can cut rings much faster than it used to All of it is stuff. I would happily share and in fact I'm thinking about writing an article for one of the Model rocketry magazines about how I do this And then I realized I could actually also submit it to one of the There's a magazine now called digital machinist, which is all about people playing with these tools at home and I Don't know. I love it when I can create a talk or write an article or something and have it be applicable in lots of different places You get multiple credit for you know doing one piece of work Unfortunately, I don't think this is showing up terribly well on the screen so I won't spend much time on it, but When you go to fly big rockets, you need big wide open spaces to do it I will admit that on the bus coming down from the airport in Buenos Aires I kept looking at the window going this would be a great place to fly rockets Nice big open spaces relatively flat not hard to find things This is a photo taken in the evening in the northeast corner of the state of Colorado in the United States in an area Called the Pawnee National Grasslands We have a launch site there where we routinely have what's called a standing waiver with the Federal Aviation Administration allowing us to launch During the organized launches there to up to 20,000 feet above ground level and with some advanced notice We can call in short windows where we can have permission to go up to 34 or 35,000 feet And this is the big grassy area. That's you know many many Square miles of territory and so when we go there, it's really far from anything else We tend to camp out and spend the night sleeping out on the prairie and that's also a fun thing to do with your with your son Some of the launch sites aren't quite so remote though. This is one that we go to down in the Southwestern part of Colorado, you'll notice in the background. We have an audience. That's a bunch of cattle And the only problem with the cows is they tend to like to come up around where the launch area is and they leave hazardous waste deposits on the ground that we have to look out for but I'll tell you that after you launch a rocket or two. They're gone. I mean They don't hang around for this on the other hand there are other animals in Colorado that aren't quite as skittish Another one of the launch sites up in the central part of the state Which for those of you who enjoy animated TV stuff, this is actually in South Park Which is you know the scene of the South Park adult cartoon thing It's a working buffalo ranch I mean these are American bison that are raised for as a meat animal in that part of the world and This one got a little frisky My son laughs every time I show this photo because at the back of my Chevrolet suburban I was standing in the back and I had the doors closed around me hoping the buffalo would go away before I You know got pronged or something But we're pretty successful at sharing the the grasslands with them and it is it it does make it sort of amusing sometimes when we go out So what about avionics? I'm the avionics are Sometime I've gotten very interested in and what what you initially think about using these four is control of the recovery system The way these rockets work is there's a motor that you ignite electrically on the ground that you know launches the rocket and takes it up I'll actually show you a video of one of those sometimes soon and when It's done when the motor burns out. There's then sort of a Period of time where the rocket's coasting and trading velocity for altitude and then at some time later You as it goes over at Apigee There's a point where the rocket's moving at sort of the minimum speed during the flight where it's slowed all the way down And it's about to start coming back and it's at that point that you'd like to do something to the rocket And normally what you want to do is have it separate maybe the nose cone pops off Which is typically what you do with the little SD size rockets Sometimes the larger ones we use an ejection charge to separate the rocket, you know further back in the body too But in any case you'd like to do something at Apigee to make the rocket not be Fully aerodynamic so that it doesn't just come back in ballistically the same way it went up I've had that happen. It's not a lot of fun So what you'd really like to be able to do is deploy a parachute or streamer or something up at Apigee With the little rockets what you do is you Anticipate how long it should coast after the motor burns out before that happens and you pick a motor that has a specified built-in Delay charge which burns for that period of time and then fires a little black powder ejection charge to separate the rocket But of course, they're not terribly predictable. We're talking about the burn rate of a delay material And it's really not a great idea with larger rockets So in the larger rockets, we actually use a little Altimeter board that I'll show you some photos of here in a minute that have a barometric pressure sensor and a microprocessor and can Determine when the air pressure has stopped going down and is starting to go back up because you've gone over at the top and use that as a trigger to eject the first part of the Recovery system. It's also interesting that if you're building big rockets that go really high You don't want to put a big parachute out up high because then the winds will carry them really far away from you before they come Back to the ground and in that case we like to do something called dual deployment where the rocket goes up You separate it at Apigee to keep it from being Totally aerodynamic and to keep it from coming back down ballistically But it's still dropping at a very high rate of speed and then it's some predetermined altitude above the ground The electronics detects that you're almost back to ground and initiates another event that deploys the big parachute And when you do that Even if there's a lot of wind up at altitude the rocket won't drift too far and you'll still be able to find it when it comes back To the ground It's also fun to collect Flight data the first thing people want to know is how high did it go? And in fact you'll notice that when I show you some of our flight information It's usually the first thing I'm interested in to another question. How fast did it go? Did we break mock? Did we go faster than the speed of sound and yes in high-powered model rocketry? It is very easy to build rockets that will go faster than the speed of sound on the way up as they go through there They're a maximum velocity on myself have broken mock at least once And then it's interesting that if you have electronics that has a memory Capacity and the ability to extract data after the flight you can record this information through the whole flight and get a flight profile to study later And then the really sophisticated systems include some kind of a radio link to the ground so that while it's still up there You can be seeing where it is what's going on Maybe even have a GPS receiver that's reporting on the position of the rocket as it floats down during the recovery phase But you know even the simple data is fun this as I mentioned is one of my son's rockets It uses motor-based ejection. It's a very simple rocket But we have put a tiny little altimeter in it that I'll show you a photo of in the next slide called pico alt And that allows us to get a peak altitude reading Which it's calculating by before launch it sits there sort of measuring the barometric pressure and Make some determination of what the current altitude is and then as the rocket flies It's sitting there sampling the air pressure and when it sees the lowest air pressure in the flight It records that and then after the flight it will flash out on a little LED What the difference between the starting point and the highest point were and using that technology? We know that the highest flight we've had on this rocket took it to 2001 feet on a motor called a g77 r The r is significant to my son because that's the type of propellant in this motor and r means Red line which is a propellant that puts out a long sort of laser bright red flame At the tail of the rocket as it's going up my son's gotten to be known in the local rocketry community is a real red line junkie Big big motors with big red flames coming out of the backup and hey dad can we do that again? This is what the pico out looks like again. I'm sorry the photo on the projector. It's a little bit dark But that's a pencil point over to the right to give you an idea of how big it is There are only a few components on here This is the little mems pressure sensor on one side of the board and the other side has the microprocessor An LED surface mount LED and a couple of resistors and that's it and this is something that's available commercially It's about $40 us I think And lots of folks in the local club have bought these since they saw us playing with them and we use them a lot I Mentioned dual deploy. This was a custom rocket that I designed. It was my first dual deployment design How it got named that's a long story you can buy me a beer and I'll tell you but this one flew with another commercial altimeter that's a little more sophisticated called a perfect flight mod and used The ejection charges are actually made You know the little lights that people put on Christmas trees or hang up at the holidays tiny little light bulbs By breaking the glass on one of those carefully You can use the filament of the bulb as an igniter for black powder charge And these were done each of the ejection charges was about a half a gram of black powder And this is a 75 millimeter diameter rocket. So it gives you some sense. It doesn't take much black powder to separate these things and on the first flight of this rocket on an h128 it went to 932 feet and I'll show you the altitude profile. I got back from the altimeter in just a minute at a Later launch about a month later. That was the smallest motor the h128 was the smallest motor This rocket was designed to fly on the J3 30 is the largest motor I had designed this rocket to fly on and on that one. I went to 6,597 feet Which is important because in the u.s You know getting at least a mile high, which is five thousand two hundred and eighty is sort of a big deal And this took me to you know, well over a mile. Unfortunately on the third flight I had what we call a bucket recovery This is when the ejection system fails to do the right thing and we watch this rocket go up and Comes straight back down. I could hear the ejection charge for the main parachute trying to fire at 800 feet But the rocket at that point was at terminal velocity headed down and Half gram of black powder wasn't enough to separate it So I picked the pieces up the only piece that's left where the blue fin can is just in front of the fins From there to the back is intact from there to the front was all completely destroyed And they're in fact a little pieces probably still out in the grasslands And this is what the mod looks like this is actually two photographs blown up a lot It's a tiny little circuit board Dip switches this one actually uses an audible Anunciator a little buzzer thingy so that after the flight it beeps at you just kind of cool You go to pick the rocket up from where it came down and it's beeping at you telling it telling you what its maximum altitude was And as you can see the wires over on the right into the screw terminals or how you hook up the wires going to the ejection charges This is another photo showing how it's mounted on a little piece of used printed circuit board material with a battery that Could slide into the payload bay and this is one of the receptacles for the little holiday light bulbs that I was using for ejection charges This is all cardboard tubing and wooden centering you know plywood centering rings and bulkheads nothing terribly sophisticated This is a picture of it under dual deployment. You can see the fin can at the bottom the little Red chute part way up was the droga parachute that it came down from Apogee on and the pink chute up there is the the main parachute Apologize for it being fuzzy, but it's the only cool picture of a dual deployment that I've flown that I have so That's blown up a lot from an otherwise lousy photo, so And this is what it looked like on the ground I love that shirt for two reasons that my son's wearing the first is it says as a matter of fact I am a rocket scientist And The other reason is that it's almost a bright fluorescent yellow-green and even if I'm a mile or two away Across the prairie I can look back and see where he is and what he's doing back at the flight line, so It's pretty handy So this is the biggest rocket that I've actually flown so far This one's called a vertical assault And it's a 75 millimeter airframe built from a kit that my wife and kids bought me for Christmas the airframe on this one is a Pheneloc cardboard ish tube that's been wrapped in fiberglass Soaked in epoxy, so it's an epoxy fiberglass over cardboard tube and the thin assembly on this is actually a fairly high-tech composite stuff high-pressure injection molded, so it's a very tough rocket and Relatively inexpensive relatively light On this one. We actually flew this on an L730, which is the largest L motors are currently the largest ones I'm personally certified to fly those are the largest motors that are part of a level two certification And the cool thing is this was up at the Pawnee National Grasslands. You get a sense there. It was dry This is in the first weekend in May this year, but it just it goes on for a very long distance I mean really is a lot like between here and Buenos Aires the just big wide open pretty flat space But the really neat thing is that day the sky was almost completely clear And so with binoculars I was able to watch this rocket go up and see the ejection event at Apogee at 14,141 feet Fortunately I had an RF beacon mounted in this one a tiny little transmitter with an antenna wire Sending out a beeping signal because without it I'm not sure we would have found it it landed about a mile and a half downrange and I had to be within a Hundred meters or so of it before I could even tell where it was the When it was laying down flat on the ground. It's almost invisible even out in all of this prairie But that's my most significant flight so far that's what the motor looked like That's an L-class motor. It's a 54 millimeter diameter and you know, you long And for those of you who are curious, that's about a hundred and sixty five dollars to fly that motor The metal casing is reusable the fuel elements and so forth or not And this was the altimeter that that particular rocket flew in so from a company called missile works This again, that's a force four squares to the inch Graph paper so you get some idea of how big that is This is another interesting altimeter because it records peak velocity in addition to peak altitude It does not give you a plot afterwards of what's going on though Very nice little unit though cost about seventy five dollars US. So what's the problem with all of this? Well All these altimeters are completely proprietary Now when we're talking about things like the little pico out that we fly in my son's rockets who cares You know, it's a little piece of hardware the fact that there's some firmware running in it You don't really think about very much. It's a single task single purpose thing It's only goal is to tell you what the maximum altitude of the flight was and if it does that and does it well And it's tiny and it's cheap. Nobody really cares whether it's an open design or not but It really irritates me when we get to the more sophisticated units that the documentation tends to be very Application-oriented. There's no discussion of how these things actually work Some of them are sort of fiddly to configure that missile works thing has two buttons and a tricolor LED and the at a very complex configuration menu structure You spent a lot of time leaning on this button while you power it on and then pushing them in exactly the right sequence to Get them programmed and configured and that's okay But it just you know, you keep thinking maybe I could do this better There's no schematics no source code for the firmware. Therefore. There's really no ability to modify their behavior and in fact To my intense frustration even the serial protocol for configuration and data extraction on the models that have a serial port is usually considered proprietary And the feature set that's provided by the you know the free Windows and Mac software they give you is pretty limited Here's an example From that first dual deploy flight that I ever had that went to 932 feet This is an altitude versus time plot that came from the software that extracted the recorded data from the altimeter Now you'll notice that you know the rocket goes up comes over There was actually an event at the top which put out a shoot which caused it to come down at a certain rate of speed And then at a pre-programmed altitude which here would have been Whatever it was six seven 500 feet is where it was set for so a little bit below 500 feet We see a change in the descent rate because the larger parachute came out and it slowed down But you'll notice that those descent rates are penciled in by hand Because in order to figure them out I had to do that stuff that they taught us all an elementary school to do with a ruler and a plot like this where you Kind of line it up and see the tick marks I made and you know, you do the math in your head and I don't know that just really irritated me So at the end of the day to put this on this slide This is a printout from a program running under windows and I've been told yes I could have done a screen snapshot and done something else with it But that would require knowing how to do that on a windows machine, which I don't know At the end of the day, it just sort of annoys me that you know The software doesn't have features that I'd like to have the first question I wanted to know on a dual deployment flight was well What were the descent rates under the two shoots? The first thing I noticed when I did the calculation is 37 feet per second It's too slow. I had too big a drogue on that. I'd like a smaller one So it's going maybe 90 or 100 feet per second So and in 23 and a half feet per second not bad But the rules for a level 3 certification flight say it has to be under 20 So these are the sorts of things that you know when I'm flying test flights. I'd like to learn quickly So, you know being the glutton for punishment that I am And after talking to some of the manufacturers and finding out they were generally sort of uninterested in my willingness to sort of volunteer to you know go under NDA and write Linux software and all those sorts of things I broke down and decided to design my own altimeter board and do it 100% Open hardware design open software design The original intent was for it to be able to fit in a 24 millimeter tube Primary operations based on that same kind of pressure sensor. I wanted to record Profiles and not just peak altitudes so that I could gather real flight data with it support for dual deployment I also noticed that all of the ones that had Computer interfaces had some goofy little connector on board that did TTL level RS 232 or something like that and then you then had to spend an extra pile of money to get the you know USB to RS 232 TTL or whatever interface now none of that's all of that crazy Except that there are plenty of microcontrollers available out there now that have built-in USB and the mass of a mini USB connector is less than the mass of the Dufus TTL serial connectors. They were using so I Decided the right thing to do is pick a microprocessor that had USB on it Then not only would it be easier to get the data out and to configure it But I could actually power it from that when I recovered it After-flight and the batteries were running down and so forth and then I wanted to put some cool stuff in there because I Wanted this to be fun. I wanted to it to be sort of interesting and educational for the local rocketry club many of whom Many of the members of which are families with kids about the same age as my son All of whom are sort of interested in learning as we go through this process I'm not going to spend a lot of time on the hardware design. It uses a an NXP arm 7 Embedded chip with a lot of IO on it a separate CMOS serial double e-prom for storing the the flight data and Some mem sensors I spent some time playing with a two-axis magnetic sensor Which I thought would be cool for you know detecting the orientation of the rocket relative to the Earth's magnetic field And people have done that made it work, but it turns out the circuit board surface area required for all the supporting electronics Just seemed completely unreasonable. So I left that out This is what the raw circuit boards look like when they came back The PC board house I'm using is a super cheap service that does two layer Circuit boards for very low cost, but they have a minimum dimension that was larger than my circuit board So I actually laid up two of them side-by-side on the board and After receiving them would go down and cut the two apart myself It would have helped if I had left a little bit more space between them The reality is that you know This is one of those putting two of them down on a panel too late at night doing the math and your head kind of Problems by the time I snapped the two boards apart. I get exactly one of two that's useful. So, you know One of my decisions for the next cut of the board is to just go to a slightly larger circuit board surface area So I can do one at a time use the board shops minimum spacing and Part of the reason for this is I've realized that with all of these features on here I'm not going to fly this in a tiny little rocket with a 24 millimeter tube. I didn't know what I was thinking This is the backside of the board. It is double-sided surface mount in order to get it to route in two layers I did put the processor and the oscillators and so forth in the back of the board Now that I've had many more less sleepless nights to stare at the design the next cut of the board wall Have all the parts on the top surface. I finally figured out what I was doing that was stupid that was making routing it So hard you'll also notice the pads those of you who know anything about this will notice that the pads for the passive components are downright huge-looking compared to What else is on here? That's because I happen to have a huge pile of 1206 surface mount passive parts. That's you know in surface mount lingo That's the big huge surface mount resistors and capacitors It would certainly be easier to route the circuit board if those were smaller parts And so I may switch to smaller ones on a future version But as those of you who have been paying attention have noticed I'm you know getting to that age and stage where Optical assistance is required so the smaller the parts get doesn't necessarily make me happier This is what the first board looked like Loaded up on the top and the bottom since I didn't really plan to talk about this here I didn't bring any of the boards along to show off which I certainly could have done The software is written mostly in C with some arm 7 assembler runs out of the on-chip flash Comes up with USB serial emulation to give you a console talking to the thing This is all done with GCC new lib a little operating system called free our toss How many of you run into that any anybody? For this kind of project. It's actually pretty cool. It's more or less just a you know set of functions that are available for doing things like Time slicing, you know multi-tasking application providing some fundamental resource allocation and so forth It's it's not all that great, but for this kind of thing. It seemed like a good fit And the USB stack that's been implemented for this processor family In free our toss the way things work is that people tend to take an evaluation board from a chip manufacturer and create a demo package of free our toss for that board in which they write an application that Exercises all of the device-drivery stuff that they've turned on on the free our toss port in this case Someone named JC Ren who I've never met and never communicated with Created the demo package for a particular olem X Eval board that used the same processor chip that I started with and that was amazingly helpful because a lot of the basic Things that you go through trying to light up an embedded Device we're sort of taken care of already. I Want to spend just a moment talking about the licenses because I think everyone in this room Understands open-source software licenses. In fact in this case. I chose GPL version two or later as the license for the firmware on the board but hardware is an interesting thing to try and do something GPL ish with and a Good friend of mine who's a former president of Tucson amateur packet radio, which is a well-known nonprofit engineering and design organization in the u.s. That did lots of the early work on packet radio over amateur radio links leading to a lot of the Wi-Fi and related protocols that we all sort of take for granted today Created a while back this license called the tapper open hardware license And I happen to think it's a really cool license to use for things like printed circuit board designs It's intended to be GPL like but the problem is that copyright law Is sort of not the right way to tackle the problem for a hardware design? And so if you go take a look at this there's a reference at the end of the talk. It's a TAPR org slash OHL. I think we'll get you there if you're impatient I think it's a fairly masterfully crafted license for solving this particular problem I and Bruce parents and various other folks that you've heard of who care about these sorts of things help to review it and There do it does seem to be a growing community of people designing Small hobby and related kinds of nonprofit hardware things under that license Are some of the tools that I use for hardware design? I actually use the G Eda suite the canoe Eda tools G-Schem and PCB are really quite capable quite functional tools. I have to tell you that In my history Back around the time I got started with the Debian project I was managing a team that maintained all of the CAD and CAW e-tools used by one of those more significant design R&D labs in old test and measurement Hewlett-Packard and so I got spoiled I mean when I sit down in front of these tools there are things that they don't do that tools We were using 10 or 15 years ago could do But at the same time I've been really impressed with the fact that there really haven't been any roadblocks with these tools to doing You know useful multi-layer Printed circuit board surface mount printed circuit board designs as has always been the case it seems like most of my time is spent creating library parts to Capture the physical characteristics of the different components. I'm using and much less time is spent hooking them all up and getting the board Routed than you'd think There are also some interesting tools if you're doing hardware designs a digic key Which is a distributor of parts in the US that's well known for taking orders over the web and sending things FedEx overnight Has grown their website immensely and they now have really good part selection Facilities and the data sheets for all their parts are online and so forth I think that's just really cool the printed circuit board service that I'm using bare bones PCB has a thing called free DFM comm that's a design for manufacturer ability service where you can upload your printed circuit board artwork and They will run it through all of their Rules checkers and tell you about all the things in your design that may pose problems during manufacturing You know copper being too close together Holes that are you know sizes that'll be hard for them to drill things like that and that's a free service that they provide Obviously hoping to hook you in and get you to spend money with them Which I don't really mind when somebody's providing that kind of a service From a software development standpoint, there's nothing here that any of you would think are terribly unusual I was using gdb via open OCD Very cool that that stuff's packaged in Debian I will admit that I built the arm cross tool chain myself In large part cuz I thought I knew how to do it and wanted to see if I still could But also because at least on the day I went looking the tool chain stuff pointed to from Mdebian and other places was frankly a little confusing to figure out and It's one of those I knew how to do it I just went and did it instead of finding the right people to talk about but there's nothing magic in this tool chain It's just a GCC cross compiler with debugger for an arm target This is the actual olem X board that JC run wrote his demo free RTOS package for I got one of them and over on the right in the prototyping area You can see a little Prototyping board with a three-axis accelerometer and the memory chip Stuck down up in the corner and actually did quite a bit of the initial software development on this board while I was waiting for To sort of get done with the printed circuit board design and get all the parts back into build That's an olem X arm debugging dongle nothing super great about it except this particular unit It's kind of cool because it hangs on a USB port on your notebook and in addition to providing the JTAG microprocessor programming and debugging interface it also has a USB serial port built-in and a Switching power supply that will give you power to power your target system off of the USB spigot So it eliminates having to carry all sorts of bits and pieces around Cable I had to make up to interface that oh Some people who have played around with building hardware things at home have this mental model that somehow Surface mount is impossible to do at home and I really beg to differ with that I think it's much easier to design and build things using surface mount components than through hole components Because you're not constantly having to flip the circuit board back and forth to get to both sides of the board all the time I mean it sounds kind of silly I just happen to think it's true the only challenge I had is that the accelerometer is in a leadless part called a QFN It's not quite as bad as a ball grid array that has lots of hidden pads on the bottom But these are just flat pads around the outside edge of the part on the bottom of the part There's no way to get to them with a soldering iron But I discovered that you can actually for a couple hundred dollars buy a hot air reflow setup that works pretty well You have to have good tweezers and at my age Having your wife and kids buy you a really nice Trinocular microscope for as a Christmas present totally winning As you can see there's a pair of tweezers in the circuit board down at the bottom with this It's really easy to do Close-up very fine-pitched work So yes, you have to get a few tools to play with really tiny circuitry stuff at home But I don't consider it hugely difficult. In fact, as I said, I think it's easier than some of the things I used to do with through-hole parts So where am I in the process? Well, I showed you some photos of the first article PC board That was completely assembled mostly tested and I talked about it in Australia at the beginning of the year. There are enough problems with that board With this sort of VO dot one board that I have not actually released all of the design files For it. The reason for that is I really don't want anybody to waste time trying to duplicate that Because I do think it would be a waste of time and unfortunately the first article PC board was damaged before I actually got to fly it I keep debating with myself whether it's worth loading up another one and redoing all of the various hacks I had to do the board to make it work or if I should just Get on with doing the version of dot twos if I ever stay home long enough I'll get some more printed circuit boards cut and try and get some of these flown but as I pointed out at the beginning of my talk at the Earlier in the week with a couple photos this project that I'm involved in right now to build an airframe for getting the last and Final level of high power certification the level three certification is kind of got my mental priority at the moment There are a bunch of changes. I'm making the board design I'm going to fix all the known Issues the only one that's lingering is I've got a goofiness with the LED that tells you whether the USB is happy or not I need to figure out I may go to smaller passive parts that I haven't done that yet Making the board a little bit bigger so it's easier to get the boards made I Oopst on some pinouts on connectors that in ways that I think are worth fixing I also had a suggestion in an Australia Linux Conf Australia in January that I go investigate the lipo Lithium polymer battery technology I didn't realize that People were making good little charging chips for lipo cells and so forth, but they are there's a single chip solution That's five pins and trivial to use that will charge a lipo battery One of the nice things about lipo batteries is that they're all 3.7 volts and they only differ in the total charge Capacity so you can design for a small battery and if you need more current you just put a bigger one in And a bunch of the guys in local clubs that um, this is really cool But where's the LED or the buzzer that tells you how high it went after the flight? And I said oh, he's plugging to the USB and they said no, no, no If it doesn't beep, it's not an altimeter and so I've gotten some sample parts and I'll figure out what to do about that before long I Think it would be fun at some point to do some variants I've got a friend right now helping me to integrate some cheap surplus GPS receiver boards that we came across cheap as in under $20 for a nice integrated core and We're hoping actually to fly that as a payload during my level three certification flight in about a month But we'll see how that goes. I do think it would be fun to do some simpler versions of this board once this one's working That would make it easier for some of the kids to fly them in smaller rockets and so forth At this point, you know, I've probably got a minute or two left for some questions the information on this More can be found at these places altos metrum org. I registered to be the place for the altimeter It's a pretty bare website right now because as I mentioned I really don't want anyone wasting time trying to duplicate the view dot one work the couple people who are directly collaborating with me on Getting version O dot two together or doing so directly and not using the website at the moment But that is where all of this stuff will get published as soon as I have something that I think it's worth people duplicating gag comm slash rockets has the Status updates and design file for my level three rocket if any of you are curious about that As I already mentioned the open hardware licenses out on the Tapper website. So that I'll thank you for your time and attention I think we're more or less out of time, but if anybody has a burning question feel free Okay, well, I'll be around the rest of the evening and a little bit tomorrow morning So if this is fun stuff and you'd like to ask questions about it feel free to catch me. Thank you very much