(Original thread started on 12-18-08 by DonnyRay Jones) So you're gonna build a Learjet are you? Good for you. It's great fun and you'll learn a lot about things you never thought you'd need to know in life. There are many people here at Hangar45 who will eagerly help you. Welcome aboard. At NASA we practice something called "Lessons Learned". This is the study of projects, events, conversations, whatever, from which we can often improve our understanding of what went right, and what went wrong. Lessons learned help us avoid making the same mistakes again. They help us improve on the things we did correctly. But most important, those Lessons can be shared with others. Almost every sim builder has some experience to share with others. Some of us have built one or more sims, cockpits, models, components, software, all of which are necessary parts of the process. My present Lear 45 is the prototype of this particular model, but it's not the first cockpit I've built. I'm not the smartest guy in the sim community - by a long shot - but I have done my share of design and fabrication. And boy have I learned some lessons along the way. Over the next few weeks I'm going to try and post some of the lessons that I learned. Some good, some not-so-much, some of which the results are yet to be known. I haven't quite figured out how to organize this effort yet. We'll see where it goes. "Experience is what you get when things didn't go like you planned." Let's see if collectively we can improve on that truism. Part One: Can I really build a Learjet? Well....."probably". You're going to need some things. "Tools" fairly immediately comes to mind, along with "workshop", and some place to hangar your creation. You're gonna need mechanical and electrical construction skills, some knowledge of modern computers and the software that runs them, and you're gonna need a budget. "A budget?" Yeah. A budget. Regardless of what sim you construct or how fancy you decide to make it, there is some cost involved. Most of us don't have an unlimited budget, access to a full machine shop, or a staff of software coders. But you don't need all those things. You just need to realize that building a sim is a pretty significant undertaking, and unless you have an unlimited budget, you may have to learn how to make some of the items you need. "How long does it take to build one?" Answer: "How long do you HAVE?" I started work on my present Lear sim in late 2005. I'm still working on it. I don't know anybody who thinks they have "completed" a cockpit, but I read great stories from other builders who are somewhere along the way with their own design. Think about how complex you want yours to be. Do you just want to fly the Lear, or do you want the experience of sitting in a working cockpit? Those are two very different objectives. They will have a great impact on the sim you build. "Make a plan, then go make it work." Every firefighter knows this mantra. It is the basis for good fireground management and for successfully fighting fires. 90% of the critical decisions which must be made on the fireground must be made within the first minute or so of arrival, and they must be made when only about 10% of the circumstances are known to the responders. If you are just beginning to think about building a sim, learn a lesson from firefighters: THINK about what you want to accomplish FIRST, then figure out a PLAN to get it done, and finally put that plan into motion. The time you spend in productive thought about building a sim will pay you back many times over in success and enjoyment. Most people with average skills and a modest collection of basic tools can build a Learjet, but the one thing you absolutely must have is enthusiasm! I'd be lying if I told you it is "easy". It's not easy, at least in the sense that there is no "cookbook" with color glossy photographs and step by step instructions that explains exactly what you need and how to proceed. That's why Hangar45 exists - to serve the community of builders who are working on a Lear 45. We are bound by this common interest. We should all be grateful to Eric and Ron for the hard work they have done to give us this opportunity to work together. "But really, can I build a Learjet?" That's up to you. Personally, I believe that many people can do this. But if you're a disaster with hand tools, can't tell the red wire from the green wire, and expect to find all the parts and materials you need without any difficulty, maybe you should talk to other sim builders before getting in too deep. I'll help you, where I can, and so will others. But that decision is yours alone. In future parts of this thread I'll describe some of the success and failures I experienced while building my Lear. Others will post their experiences as well. It'll be fun. Welcome aboard. Part Two: Where are you gonna put it? The selection of hangar space for your Lear depends on several variables, the most important of which is likely to be "where-ever my wife says I can put it". <G> But if you've solved that one, here are some other things to consider. 1. Size and weight. A Lear cockpit will completely fill most spare bedrooms, so before you start building it in there take some time and lay out the design on the floor with painter's tape. Walk around it on all sides, considering that you will need space and access during the construction phase. If you are using an upstairs room, consider the weight loading on the floor system. A fully loaded 45 easily weighs more than your average set of bedroom furniture. 2. Build it outside - carry the parts indoors. The reason for this is that sooner or later most of us will have some occasion to dismantle our Lear for relocation. Perhaps you are moving to a new house, perhaps your family expands and you need the room for a nursery, whatever. So when you build your Lear, make the large parts outdoors and then carry them inside. Simple rule: If you can't get them IN the house, you can't get them OUT either. 3. Consider ceiling height, particularly if you intend to add a motion base at some point in the future. The Lear fits fine in a room with a standard eight foot ceiling - but only when it's level in the pitch axis. At high angles of pitch down, it's possible for the top of the fuselage to hit an eight foot ceiling. 4. Consider throw distance if you plan on using a projector. Better projectors with short-throw lenses are getting less expensive, but before you buy one go up on the manufacturer's web site and use their lens throw distance calculator to SEE how your image size will work out in the space you have. It's disappointing to get a shiny new projector installed and discover it will only make a 4-foot wide image in the ten feet of throw distance you have. So, as in carpentry, "Measure twice, cut once". Part Three: Be patient. This ain't no "weekend" project. I started working on my Lear in late 2006. *HAD I KNOWN THEN* that I would still be adding things to it three years later I might not have even started! But ignorance is bliss, or so they say, and I'm happy to report that I'm pretty happy with the outcome. Lessons learned: 1. "It takes as long as it takes, and it costs more than you have." But it is indeed a labor of love and like anything else important to humans, you will find a way to spend the time and afford the parts if you really want to build a Lear. 2. Keep a PC set up somewhere else in the house where you can go off and FLY from time to time. When I built my first prototype, I flew it nearly every day for about a year. And then I decided to build the fully functional cockpit and I needed the space for it. So I dismantled the prototype, leaving me with nothing to FLY during the long construction task. That was a mistake. For me, it's the love of FLYING that keeps me working on my Lear, not necessarily the love of *building it*. 3. Expect your plans for software, hardware, switches, and a hundred other things to change over time. Vendors in this hobby space come and go. Consider using a design that has some flexibility so that you are not locked into one vendor's products. Should that vendor go bust, you may be stuck with items that can't be supported. 4. Build your Lear so that it can be disassembled and relocated. We recently sold our home near Houston and are relocating. Packing up the house became the ultimate test of my Lear's modular design. It's in storage as I write this, packed safely away in about two hundred boxes and cartons. The "dismantling" part of the exercise actually went better than I had hoped. The "putting it back together" part comes after our relocation. Stay tuned. I'll let you know how that part worked. 5. TAKE PHOTOS as you build your Lear. You will find them useful for many purposes beyond simply sharing with others here on Hangar 45. I tried to take a set of photos after each major assembly or item was installed. I wish I had taken more, particularly close up photos of how I built the control column heads and some of the dual rudder mechanisms. While my Lear is in storage I have been trying to continue work on some of the design details, but I find that I cannot recall the exact dimensions or physical arrangements for some of the stuff without good photos. I wish I had taken MORE. Part Four: You’re gonna need connectors. If you look through the projects described by the members of Hangar45 you’ll notice something. We’re all building “prototypes”. Every one of us has a unique build. The people using Ron’s shell are closer than anybody to what might be called a production version, but even among those builders each one is unique. Don’t think so? Go back and look at the photos again. Even something which should easily be possible to do in a uniform manner, like mounting LCDs behind the panel, are not standardized. If you look deeper than that, to things such as panels, knobs, switches, etc., it’s easy to see how each builder has gone his own way based on availability of materials, skills, and budgets. Lesson learned: This is known as “concurrent development” in the aerospace world, and it’s a dirty word. It often results in high program cost, reduced capabilities, oddball hardware and software, things that don’t fit right, expensive rework, and a generally unhappy end result. If you want to read a good history on this look up the program management study on the F-111 aircraft. Big glaring examples of things that will jump up and bite you later in the development. The nature of prototype work is that you often find yourself having to take something apart for revision, rework, or even redesign. In something as complex as a simulator there are wires – LOTS of wires – all of which must be connected to something on both ends. You don’t want to have to disconnect all those wires from terminal strips, or switch contacts, or lamps, or throttle pots, or whatever unless you can do so quickly and easily. And for that you’re gonna need connectors. The wiring in real aircraft is equipped with connectors at both ends of the various harness assemblies so that parts can easily be removed. Control heads all have multipin connectors. Avionics and other boxes are connectorized. You pop the connector and remove the box without hassle. I have put a connector on the FAR end of every wire in my Lear. The other end of nearly all of my wiring goes back to a common interface box where logic and interface cards are located. This is a large shallow chassis intended to allow sufficient room in which to change or add to the design as necessary. I built my Lear this way because I knew when I started that I would eventually have to dismantle it when we relocated out of Texas. Every panel, control, or assembly with wiring is connectorized on the *device* end of the wire. None of the wiring at the interface end is connectorized. It was my intention that one could disconnect all the *end devices* and simply pull the wiring back toward the interface box, and pile it all up in one good stack without having to disconnect anything in the interface box. Having gone through that experience of dismantling my Lear, I can tell you this scheme worked well. I can also tell you that the presence of connectors on everything made it FAR easier to work on small sections of the sim at a time. I bet I had the throttle quadrant in and out of my airframe fifty times. And every time it took less than a minute to disconnect or reconnect all that wiring, because I had used connectors for everything in the quadrant. My prototype contains a variety of connector types, mostly because I used what I had in the trusty Junque box. (Junque = “high class” junk.) All of my knee panels are wired with multipin aircraft connectors, because I needed a high pin count and I had a good supply of them in hand. The left and right hand outside panels (the dimmer panels) are wired with conventional 15 pin D-sub connectors. I used these because you can buy pre-made cables with 15 pin D-subs on each end at any decent computer store. So I bought the pre-made cables and cut one end off to whatever length I needed. Saves a ton of labor, and they’re cheap. In other places I’m using USB connectors, RJ45 connectors, audio connectors, mic connectors, power connectors, multipin Molex connectors, even a couple of parallel printer cables with the 25-pin connector. There are many different connector requirements in a sim. I have used the following rule to select the appropriate connector based on – in this order – (1) pin count, (2) size (3) ready availability of cheap pre-made cables, and (4) availability and cost. Your mileage may vary, but you’re gonna need connectors on your wiring. Be thinking about it NOW before you run into that circumstance where you have to remove something that you’ve hard wired at both ends. Save yourself a lot of hassle. Part Five: Which pots should I use? A good amount of bandwidth seems to be used up in the sim community with discussions of pots. Which pot should I use? What value should it be? Linear or rotary movement? Linear or log taper? The term “pot” is an abbreviation for “potentiometer”. I’ll use “pot” here, and my typing fingers and spelling genes will appreciate it. A pot is in effect a tapped resistor. It’s a FIXED VALUE resistor, to which is provided a means for making an electrical connection anywhere along the physical length of the resistive material. That connection is called a “wiper” or sometimes “slider”. The wiper moves along the resistive material as you turn the shaft of a rotary, or move the slider on a linear pot. It’s important to get a clear understanding of how the pot works electrically if you want to better understand how to select and use them. Variable resistors (pots) can be connected in multiple ways, but the two most common methods are known as a “rheostat” and a “divider” connection. A rheostat connection forms a variable resistance between the slider and ONE end of the pot. In fact, a true rheostat only HAS two connections, one to the end of the resistive element and the other to the wiper. A pot has three connections, one to each end of the resistive element, and one to the wiper. If you need to vary the RESISTANCE in a circuit, as is often done in lamp dimmers, you need a rheostat connection. As you turn the shaft on the rheostat the wiper moves up and down the resistive element. This causes a change in the resistance observed between the two connection terminals. So if you make a RHEOSTAT connection in a circuit, the pot will produce a change in resistance as you operate it. Lesson learned: Changing the RESISTANCE in a series circuit affects the CURRENT that flows in that circuit. If you need to vary the VOLTAGE in a circuit, as is done with flight controls and other sensors, you want a divider connection. This connection forms a voltage “divider” across the resistive element of the pot. Think of it this way. If you connect two resistors in series and apply 5 volts across them, the voltage observed AT THE CONNECTION IN THE MIDDLE will be some portion of the applied 5 volts. If both resistors are of the identical value, that voltage will be exactly half of the applied voltage. What you get in the middle is divided by the two series resistors in the circuit. Hence – someone long ago called it a voltage divider and the name stuck. In a pot the wiper is free to slide along the length of the resistive element. The wiper IS this “connection in the middle”, and it moves along the resistive element as you operate the pot. If you think about that for a minute you’ll realize that what you are really changing is the POINT along the resistive element where the wiper makes connection. And when you change that physical point, you change the relative values of the two resistors in the divider. Change the relative values of the resistors, and you will change the VOLTAGE observed at the wiper. Lesson learned: Changing the ratio of two series connected resistors in a circuit affects the VOLTAGE observed at the common connection (the wiper of the pot). Game inputs and most other inexpensive analog input cards do not measure CURRENT. They measure VOLTAGE, and that’s why you need a classic voltage divider circuit for your flight controls. You need a pot. Ok-all that’s good history, but WHICH pot do I need? It depends. When pots are used as position sensors best performance is obtained by flowing a few milliamperes (ma) of current through the element. 5 ma has been found to work well with many different types of pots. So the first thing to consider when selecting a pot is “What is the supply voltage in the circuit to which the pot is connected?” Many analog input cards run on +5 VDC. For a 5V analog card which applies 5V to the pot, a 1K value will flow 5ma in the pot. If you have a card which supplies 12V to the pot, you’d want a 2400 ohm pot. The nearest standard value pot would be a 2.5K, and that will flow a tad less than 5ma, but perfectly workable. Lesson learned: Pick the pot VALUE to get about 5ma of current flow in the pot, based on the voltage supplied to the pot by your analog input card. How is that determined? Ohm’s Law. E = I X R (or) I=E/R (or) R=E/I Linear or log taper? The “taper” of a pot refers to the manner in which the resistance changes along the length of the resistive element. “Linear” means that the incremental change in resistance is equal along the length of the element. “Log” means that the incremental change in resistance varies at a logarithmic rate along the length of the element. An “audio” taper pot is a form of log taper, selected to provide a satisfying change in audio volume as a person “turns the volume control”. (There is no electrical reason for this, it is purely to accommodate the human tendency to rotate controls approximately 90 degrees in one smooth motion.) For flight controls, we want as LINEAR a response as we can GET in the entire system, and for that we need pots with a linear taper. Rotary or slide pots? It depends. The maximum amount of linearity is obtained when the pot is operated over the maximum physical travel of the wiper. This means that we want to arrange our mechanical linkages and such so that we turn a rotary from “stop to stop”. If using a slide pot, you want the wiper to run “end to end”. This provides the widest range of input voltage to your analog card, and thus reduces the effect of non-linearity and quantization errors in the analog-to-digital (A/D) converters on the card. If you’re running a 5V card, you want the pot to provide the full zero-to-5-volts of input to the card as the pot is operated through it’s normal range. Here’s why: An A/D “reads” the input voltage in increments. If you divide 5 volts by 256 increments, each increment equals approximately .02 volts. This is an acceptably “small” step to detect tiny physical movements of the flight control. But if your pot linkage only moves the pot for 50% of it’s range, then the input card only has 2.5V of range to digitize. 2.5 V divided by 256 increments is only .01 volts. Thus the range of movement offered to your analog card is WORSE by a factor of DOUBLE. If you’ve had to adjust the gain of your flight control inputs to reduce the elevator sensitivity, for example, or tinker with the flight model in MSFS to smooth out the hand flying qualities of the Lear, this is likely the reason why. Lesson learned: Use any style pot you wish, but make sure that your mechanical linkage operates the pot over it’s full range. You will find this much easier to do with slide pots than rotary pots. Should I use shielded wire on my pots? Yes. Even if I am using low-impedance pots? Yes. Here’s why: The input impedance of the analog card is the limiting factor here – not the value of the pot. Most analog card inputs are relatively high impedance, so if you use say….a 10K or 100K pot, the entire circuit is more likely to be affected by electrical noise. So use shielded cable. It’s cheap, and it’s good insurance against such things as the cheap noisy power supplies found in nearly all computers. But wait! There’s more! There’s another reason why you don’t want to use high-value pots on an analog input. Let’s say you’re using a 100K pot on your elevator, and your analog card input has an input impedance of 50K. For the purposes of this example, park the elevator at exactly neutral pitch. Now consider the elevator pot. At the middle position, that pot should be sitting at half travel. For a 100K pot, that is the equivalent of two 50K resistors in series, with the wiper sitting in the middle. So that’s a 50K resistor from 5V down through the top “half” of the pot, out the wiper, and into the 50K analog input, which is in parallel with the “bottom” half of the pot. The combination of the analog input impedance and the bottom half of the pot effectively produces a resistance of 25K between the wiper and ground. NOW, think about that voltage divider. NOW it’s a 50K pot from 5V down to the wiper, and a 25K pot from the wiper to ground. What VOLTAGE would you observe at the wiper? It SHOULD BE exactly 2.5V with the elevator at neutral pitch, but in this example you will find it to be much less. How much less? Ohm’s Law will tell you, and once you figure that out you’ll see the point here. If your pot value isn’t compatible with your card input impedance, you will have non-linearity problems over the mechanical range of the flight control. Lesson learned: The input impedance of your analog card should be AT LEAST 10X the value of your pot to eliminate loading effects of the input. You can’t do anything about the input impedance of your card, but you can select a good low value pot to connect to it. For our example above, if your card has an input impedance of 50K, simply divide by 10. A 5K pot would meet THIS requirement, but 5K may not be a low enough value to flow 5ma in the circuit. You can always use a LOWER value pot to get the 5ma, and this will provide additional loading margin against the card input impedance. Perhaps the best piece of advice I can offer here is this: BUY QUALITY POTS! Cheap pots are notorious for poor repeatability, noisy wiper connections, temperature drift issues, and even simple mechanical failure. There are many occasions in simulator building when cheeeeeep components will work well. This ain’t one of them. BUY QUALITY POTS! Who knew pots were so much trouble? And we haven’t even discussed trim or autopilot inputs, yaw dampers, rudder boost, or ways to implement summing nodes for proportional toe brakes on both sides of the cockpit. Part Six: Floobydust Floobydust: “A collection of important but otherwise unrelated things.” • Transport category aircraft panel lighting is a color known in the biz as “Blue-White”. White LEDs are very near that color. Incandescent bulbs appear as what people call “warm white” (nearer the red end of the spectrum). Bulbs are used in switches, except in new aircraft like the 787, and those are LED-illuminated. • Can I use automobile seats in my sim? I did, but consider that auto seats are not “flat” on the bottom. They have all sorts of weird curves under them because the floor of an auto is not flat either. Also, the bottom cushion has no real structure to it so it’s impractical to try and cut out the “notch” for the control column. • How are you going to couple the autopilot servos to the physical controls? Cables? Mechanical linkages? If you want your controls to physically move under command of the autopilot you’re gonna need to plan ahead for this. Here’s a hint: Put your sensors on the crossover linkage (or cables), but not in the yoke(s). • We need a working replica of the standby compass to mount in the windscreen center structure. I think you can buy a replica compass from one of the sim community suppliers, but to my knowledge nobody has successfully mated one to the Lear. • Landing gear switch knob. I need one, and I imagine that many others do too. • As much as is possible try to keep the top of yoke columns clear of mechanical parts. A lot of wiring has to go in there and it takes great delight in getting itself tangled in the moving parts. This is not good, particularly at FL-290 in the dark. • How do you interface the trim switch on the control yoke? You use it to operate small motor-driven pots, the output of which is summed with your primary flight control sensors. • A couple of folks have asked, “I’m using a bellcrank at the top of my yoke to connect the crossover cables to the other side, but they don’t turn satisfactorily. The further you turn one yoke, the less movement is observed in the other yoke.” That’s because a bellcrank does not present a constant radius turn from the perspective of the attached cable. Use a pulley instead and it will work fine. • USE SCHEDULE 80 PVC PIPE! That’s the thick wall stuff. Home Depot, Lowe’s, etc., stock it in the plumbing section. And DON’T GLUE IT! USE SCREWS! You may need to take it apart sooner than you imagine. • You know how computers hate static electricity? You ARE using static-resistant carpet in your cockpit aren’t you……ahem. • And finally, “Please confine your crashes to the far end of the field. Thank You.” Part Seven: The Advantages of a Standard Design, (or) “Let’s Have Some Controversy Here People!” Question: Do you know how Henry Ford sold a lot of automobiles to people of ordinary financial means? Answer: He figured out how to manufacture those automobiles and sell them at prices that buyers could afford. You know this story. Henry Ford perhaps didn’t invent the concept of mass production using standardized parts, but he was the first to perfect it commercially in the automobile industry. The use of standardized parts and construction enabled Ford to sell reasonably priced automobiles to far more buyers than could otherwise afford them. The community of people who are building aircraft simulators presently suffers from the same “mostly-built-by-hand” costs and issues as automobile makers did prior to Henry Ford. It is not a sustainable model. I propose that we, as members of Hangar45, consider the means and methods that are known to reduce cost, effort, and ad-hoc engineering failures. One of our primary objectives is to reduce the cost of panels, wiring, hardware, and other specialized components. Have you added up the cost of a full set of panels, switches, pedals, throttles, cup holders, software, interface cards, and whatever else is needed to make them function? Protip: DON’T! It’ll discourage you. Another primary objective is to dramatically improve the success ratio of builders. It’s hard to build a functional sim cockpit in the present day. Think about all the tools and equipment and time and money and everything that you have invested in your aircraft. Dozens and dozens of builders have come and gone in our group, and yet Hangar45 has not a single example of a working sim that could be called complete and flyable. No we don’t. Not mine, not yours. Think about the barriers faced by builders. I’ll list a few. I’ll bet you can add to this list. • Builder skills. Not every builder can solder wiring or run a CNC machine. • Budgets. Even if you build a stripped-down cockpit, the cost is not trivial. • Lack of parts. Home Depot, Lowe’s, and e-Bay ain’t got everything we need. • Time. How much longer do you want to spend BUILDING your sim? • Finished product. Builders want something that will be a source of pride and achievement, but many don’t know how to achieve that result. We can do better than this. Think about the incredible variety and number of talented members of Hangar45. What if someone brought them to one location, equipped with tools and machines, supplied them with materials and parts, and said, “Go figure out how to mass-produce a Lear 45 simulator.” Think about the outcome. Think about what we could achieve, working together on a standardized design. Allow your imagination to run a bit. Why can’t we do this anyway, even though we’re scattered all over the globe? Bombardier does it. Boeing does it. Airbus does it. Even General Motors, Ford and Nissan does it. So can we. Back in December 2009, just before we packed up to move to South Carolina, I promised to write some words about connectors and wiring techniques and related topics. I’m still planning to do that, but there are a great variety of issues that should be addressed before a workable design is realized. Here are a few examples to get you thinking: • Where should the interface cards be located, and must we all use the same type? I built some panels with multipin connectors and wiring that connect to interface cards located in the avionics bay. I built other panels with the interface cards *in* the panel so that they only need a USB cable for a connection. Are either of these methods better? Why? • What kind of cable (wire) is appropriate for a wire harness? Some builders use wiring harnesses built by hand from separate pieces of wire. Some are using flat ribbon cable. Are you using shielded wire? Why or why not? Which of these are good design decisions? • How do you decide the pinouts to use for each connector? (Which wire goes on which pin.) If we all agreed on a common schematic and connector, we gain no benefit whatsoever unless we’re using a standardized pinout. • How is your panel wired? This means “schematic diagram”. Are you wiring your panel switches as common ground or common +Vcc? (Active high, or active low?) Did you use pull-up / pull-down resistors? Are you bringing each contact out to the connector, or just the one *you* think you’ll need? How about lamp inputs for the dimmers? Is yours a 5 volt or 28 volt dimmer panel? • Momentary or push-latch-push-unlatch switches? If you use a mechanically latched switch, MSFS cannot change the state of the switch when an internal variable changes how the sim is running. It ain’t good when your switch/indicator status doesn’t match what the airplane is doing. In complex systems design we use something called an Interface Control Document (ICD). The ICD is a document that describes everything necessary to build wiring, equipment, or systems that must connect to either side of the interface. Suppose I said to you, “I need a spare tire for my car, but it must be a 15-inch wheel with 5 lug nut holes.” That’s a simple example of an ICD specification. It tells you two of the most important parameters that one must know to obtain a spare tire that will fit and work on my car. Beyond that, ANY brand, color, price, etc. unit would work provided it meets that spec. We should be using some form of the ICD method for panels, wiring, linkages and such. This frees builders and suppliers from the task of having to do an original design for nearly every piece used in our cockpits. If we had an ICD for the electrical panel, for example, both builders AND manufacturers would know exactly how to fabricate and wire that panel, or harness, in a way that guarantees it will work when installed. The ICD specifies a connector type, the pinouts, and any mechanical details needed to locate and orient the connector on the rear of the panel. Another ICD will specify how the panel should be wired internally. This means, “a schematic”. If you build your panel to the connector ICD and wire it to the panel ICD, you gain the assurance that it will function as intended when connected to a wiring harness (also built to an ICD spec). The use of the ICD method makes it possible for builders or manufacturers to produce nearly any kind of panel they wish, yet it will still function correctly in the airplane so long as it conforms to the ICD spec. Can’t afford AML switches? No problem. Use whatever you can afford but wire them per the ICD and they’ll work in the airplane. Can’t afford backlighting? No problem. Make your panel face out of cardboard if you wish, and so long as you follow the ICD specs it will fit in the panel and function as intended. If you’re short on cash, you may elect to buy or build the “entry level” products. So long as they conform to the ICD spec, you gain the ability to upgrade to a nicer panel, TQ, or set of rudder pedals, by simply removing the “basic” and installing the “better”. The ICD method provides an amazing degree of flexibility for everyone, at many skill levels and costs. The use of ICDs would enable builders to get panels from Big Bird, wiring from Elmo, interface cards of their choice from Elmo, and software that runs it, and it would all function together as intended. All of these sub-systems become part of the standard design, with each element described and specified by one or more ICDs. This is a complex topic. To my knowledge no other (hobbyist) aviation simulator group has standardized a design across multiple suppliers and builders. As members of Hangar45, we have an opportunity here to demonstrate how it should be done. It is my intention to stir useful conversation and discussion. Given the talent we have here at Hangar45, I’m asking the question “Why haven’t we done this?” Comments encouraged. Ideas welcomed. Good ideas may get you 15 minutes of fame. Or an ICD Standards Committee assignment. DonnyRay Jones (Original thread started on 12-18-08 by DonnyRay Jones) So you're gonna build a Learjet are you? Good for you. It's great fun and you'll learn a lot about things you never thought you'd need to know in life. There are many people here at Hangar45 who will eagerly help you. Welcome aboard. At NASA we practice something called "Lessons Learned". This is the study of projects, events, conversations, whatever, from which we can often improve our understanding of what went right, and what went wrong. Lessons learned help us avoid making the same mistakes again. They help us improve on the things we did correctly. But most important, those Lessons can be shared with others. Almost every sim builder has some experience to share with others. Some of us have built one or more sims, cockpits, models, components, software, all of which are necessary parts of the process. My present Lear 45 is the prototype of this particular model, but it's not the first cockpit I've built. I'm not the smartest guy in the sim community - by a long shot - but I have done my share of design and fabrication. And boy have I learned some lessons along the way. Over the next few weeks I'm going to try and post some of the lessons that I learned. Some good, some not-so-much, some of which the results are yet to be known. I haven't quite figured out how to organize this effort yet. We'll see where it goes. "Experience is what you get when things didn't go like you planned." Let's see if collectively we can improve on that truism. Part One: Can I really build a Learjet? Well....."probably". You're going to need some things. "Tools" fairly immediately comes to mind, along with "workshop", and some place to hangar your creation. You're gonna need mechanical and electrical construction skills, some knowledge of modern computers and the software that runs them, and you're gonna need a budget. "A budget?" Yeah. A budget. Regardless of what sim you construct or how fancy you decide to make it, there is some cost involved. Most of us don't have an unlimited budget, access to a full machine shop, or a staff of software coders. But you don't need all those things. You just need to realize that building a sim is a pretty significant undertaking, and unless you have an unlimited budget, you may have to learn how to make some of the items you need. "How long does it take to build one?" Answer: "How long do you HAVE?" I started work on my present Lear sim in late 2005. I'm still working on it. I don't know anybody who thinks they have "completed" a cockpit, but I read great stories from other builders who are somewhere along the way with their own design. Think about how complex you want yours to be. Do you just want to fly the Lear, or do you want the experience of sitting in a working cockpit? Those are two very different objectives. They will have a great impact on the sim you build. "Make a plan, then go make it work." Every firefighter knows this mantra. It is the basis for good fireground management and for successfully fighting fires. 90% of the critical decisions which must be made on the fireground must be made within the first minute or so of arrival, and they must be made when only about 10% of the circumstances are known to the responders. If you are just beginning to think about building a sim, learn a lesson from firefighters: THINK about what you want to accomplish FIRST, then figure out a PLAN to get it done, and finally put that plan into motion. The time you spend in productive thought about building a sim will pay you back many times over in success and enjoyment. Most people with average skills and a modest collection of basic tools can build a Learjet, but the one thing you absolutely must have is enthusiasm! I'd be lying if I told you it is "easy". It's not easy, at least in the sense that there is no "cookbook" with color glossy photographs and step by step instructions that explains exactly what you need and how to proceed. That's why Hangar45 exists - to serve the community of builders who are working on a Lear 45. We are bound by this common interest. We should all be grateful to Eric and Ron for the hard work they have done to give us this opportunity to work together. "But really, can I build a Learjet?" That's up to you. Personally, I believe that many people can do this. But if you're a disaster with hand tools, can't tell the red wire from the green wire, and expect to find all the parts and materials you need without any difficulty, maybe you should talk to other sim builders before getting in too deep. I'll help you, where I can, and so will others. But that decision is yours alone. In future parts of this thread I'll describe some of the success and failures I experienced while building my Lear. Others will post their experiences as well. It'll be fun. Welcome aboard. Part Two: Where are you gonna put it? The selection of hangar space for your Lear depends on several variables, the most important of which is likely to be "where-ever my wife says I can put it". <G> But if you've solved that one, here are some other things to consider. 1. Size and weight. A Lear cockpit will completely fill most spare bedrooms, so before you start building it in there take some time and lay out the design on the floor with painter's tape. Walk around it on all sides, considering that you will need space and access during the construction phase. If you are using an upstairs room, consider the weight loading on the floor system. A fully loaded 45 easily weighs more than your average set of bedroom furniture. 2. Build it outside - carry the parts indoors. The reason for this is that sooner or later most of us will have some occasion to dismantle our Lear for relocation. Perhaps you are moving to a new house, perhaps your family expands and you need the room for a nursery, whatever. So when you build your Lear, make the large parts outdoors and then carry them inside. Simple rule: If you can't get them IN the house, you can't get them OUT either. 3. Consider ceiling height, particularly if you intend to add a motion base at some point in the future. The Lear fits fine in a room with a standard eight foot ceiling - but only when it's level in the pitch axis. At high angles of pitch down, it's possible for the top of the fuselage to hit an eight foot ceiling. 4. Consider throw distance if you plan on using a projector. Better projectors with short-throw lenses are getting less expensive, but before you buy one go up on the manufacturer's web site and use their lens throw distance calculator to SEE how your image size will work out in the space you have. It's disappointing to get a shiny new projector installed and discover it will only make a 4-foot wide image in the ten feet of throw distance you have. So, as in carpentry, "Measure twice, cut once". Part Three: Be patient. This ain't no "weekend" project. I started working on my Lear in late 2006. *HAD I KNOWN THEN* that I would still be adding things to it three years later I might not have even started! But ignorance is bliss, or so they say, and I'm happy to report that I'm pretty happy with the outcome. Lessons learned: 1. "It takes as long as it takes, and it costs more than you have." But it is indeed a labor of love and like anything else important to humans, you will find a way to spend the time and afford the parts if you really want to build a Lear. 2. Keep a PC set up somewhere else in the house where you can go off and FLY from time to time. When I built my first prototype, I flew it nearly every day for about a year. And then I decided to build the fully functional cockpit and I needed the space for it. So I dismantled the prototype, leaving me with nothing to FLY during the long construction task. That was a mistake. For me, it's the love of FLYING that keeps me working on my Lear, not necessarily the love of *building it*. 3. Expect your plans for software, hardware, switches, and a hundred other things to change over time. Vendors in this hobby space come and go. Consider using a design that has some flexibility so that you are not locked into one vendor's products. Should that vendor go bust, you may be stuck with items that can't be supported. 4. Build your Lear so that it can be disassembled and relocated. We recently sold our home near Houston and are relocating. Packing up the house became the ultimate test of my Lear's modular design. It's in storage as I write this, packed safely away in about two hundred boxes and cartons. The "dismantling" part of the exercise actually went better than I had hoped. The "putting it back together" part comes after our relocation. Stay tuned. I'll let you know how that part worked. 5. TAKE PHOTOS as you build your Lear. You will find them useful for many purposes beyond simply sharing with others here on Hangar 45. I tried to take a set of photos after each major assembly or item was installed. I wish I had taken more, particularly close up photos of how I built the control column heads and some of the dual rudder mechanisms. While my Lear is in storage I have been trying to continue work on some of the design details, but I find that I cannot recall the exact dimensions or physical arrangements for some of the stuff without good photos. I wish I had taken MORE. Part Four: You’re gonna need connectors. If you look through the projects described by the members of Hangar45 you’ll notice something. We’re all building “prototypes”. Every one of us has a unique build. The people using Ron’s shell are closer than anybody to what might be called a production version, but even among those builders each one is unique. Don’t think so? Go back and look at the photos again. Even something which should easily be possible to do in a uniform manner, like mounting LCDs behind the panel, are not standardized. If you look deeper than that, to things such as panels, knobs, switches, etc., it’s easy to see how each builder has gone his own way based on availability of materials, skills, and budgets. Lesson learned: This is known as “concurrent development” in the aerospace world, and it’s a dirty word. It often results in high program cost, reduced capabilities, oddball hardware and software, things that don’t fit right, expensive rework, and a generally unhappy end result. If you want to read a good history on this look up the program management study on the F-111 aircraft. Big glaring examples of things that will jump up and bite you later in the development. The nature of prototype work is that you often find yourself having to take something apart for revision, rework, or even redesign. In something as complex as a simulator there are wires – LOTS of wires – all of which must be connected to something on both ends. You don’t want to have to disconnect all those wires from terminal strips, or switch contacts, or lamps, or throttle pots, or whatever unless you can do so quickly and easily. And for that you’re gonna need connectors. The wiring in real aircraft is equipped with connectors at both ends of the various harness assemblies so that parts can easily be removed. Control heads all have multipin connectors. Avionics and other boxes are connectorized. You pop the connector and remove the box without hassle. I have put a connector on the FAR end of every wire in my Lear. The other end of nearly all of my wiring goes back to a common interface box where logic and interface cards are located. This is a large shallow chassis intended to allow sufficient room in which to change or add to the design as necessary. I built my Lear this way because I knew when I started that I would eventually have to dismantle it when we relocated out of Texas. Every panel, control, or assembly with wiring is connectorized on the *device* end of the wire. None of the wiring at the interface end is connectorized. It was my intention that one could disconnect all the *end devices* and simply pull the wiring back toward the interface box, and pile it all up in one good stack without having to disconnect anything in the interface box. Having gone through that experience of dismantling my Lear, I can tell you this scheme worked well. I can also tell you that the presence of connectors on everything made it FAR easier to work on small sections of the sim at a time. I bet I had the throttle quadrant in and out of my airframe fifty times. And every time it took less than a minute to disconnect or reconnect all that wiring, because I had used connectors for everything in the quadrant. My prototype contains a variety of connector types, mostly because I used what I had in the trusty Junque box. (Junque = “high class” junk.) All of my knee panels are wired with multipin aircraft connectors, because I needed a high pin count and I had a good supply of them in hand. The left and right hand outside panels (the dimmer panels) are wired with conventional 15 pin D-sub connectors. I used these because you can buy pre-made cables with 15 pin D-subs on each end at any decent computer store. So I bought the pre-made cables and cut one end off to whatever length I needed. Saves a ton of labor, and they’re cheap. In other places I’m using USB connectors, RJ45 connectors, audio connectors, mic connectors, power connectors, multipin Molex connectors, even a couple of parallel printer cables with the 25-pin connector. There are many different connector requirements in a sim. I have used the following rule to select the appropriate connector based on – in this order – (1) pin count, (2) size (3) ready availability of cheap pre-made cables, and (4) availability and cost. Your mileage may vary, but you’re gonna need connectors on your wiring. Be thinking about it NOW before you run into that circumstance where you have to remove something that you’ve hard wired at both ends. Save yourself a lot of hassle. Part Five: Which pots should I use? A good amount of bandwidth seems to be used up in the sim community with discussions of pots. Which pot should I use? What value should it be? Linear or rotary movement? Linear or log taper? The term “pot” is an abbreviation for “potentiometer”. I’ll use “pot” here, and my typing fingers and spelling genes will appreciate it. A pot is in effect a tapped resistor. It’s a FIXED VALUE resistor, to which is provided a means for making an electrical connection anywhere along the physical length of the resistive material. That connection is called a “wiper” or sometimes “slider”. The wiper moves along the resistive material as you turn the shaft of a rotary, or move the slider on a linear pot. It’s important to get a clear understanding of how the pot works electrically if you want to better understand how to select and use them. Variable resistors (pots) can be connected in multiple ways, but the two most common methods are known as a “rheostat” and a “divider” connection. A rheostat connection forms a variable resistance between the slider and ONE end of the pot. In fact, a true rheostat only HAS two connections, one to the end of the resistive element and the other to the wiper. A pot has three connections, one to each end of the resistive element, and one to the wiper. If you need to vary the RESISTANCE in a circuit, as is often done in lamp dimmers, you need a rheostat connection. As you turn the shaft on the rheostat the wiper moves up and down the resistive element. This causes a change in the resistance observed between the two connection terminals. So if you make a RHEOSTAT connection in a circuit, the pot will produce a change in resistance as you operate it. Lesson learned: Changing the RESISTANCE in a series circuit affects the CURRENT that flows in that circuit. If you need to vary the VOLTAGE in a circuit, as is done with flight controls and other sensors, you want a divider connection. This connection forms a voltage “divider” across the resistive element of the pot. Think of it this way. If you connect two resistors in series and apply 5 volts across them, the voltage observed AT THE CONNECTION IN THE MIDDLE will be some portion of the applied 5 volts. If both resistors are of the identical value, that voltage will be exactly half of the applied voltage. What you get in the middle is divided by the two series resistors in the circuit. Hence – someone long ago called it a voltage divider and the name stuck. In a pot the wiper is free to slide along the length of the resistive element. The wiper IS this “connection in the middle”, and it moves along the resistive element as you operate the pot. If you think about that for a minute you’ll realize that what you are really changing is the POINT along the resistive element where the wiper makes connection. And when you change that physical point, you change the relative values of the two resistors in the divider. Change the relative values of the resistors, and you will change the VOLTAGE observed at the wiper. Lesson learned: Changing the ratio of two series connected resistors in a circuit affects the VOLTAGE observed at the common connection (the wiper of the pot). Game inputs and most other inexpensive analog input cards do not measure CURRENT. They measure VOLTAGE, and that’s why you need a classic voltage divider circuit for your flight controls. You need a pot. Ok-all that’s good history, but WHICH pot do I need? It depends. When pots are used as position sensors best performance is obtained by flowing a few milliamperes (ma) of current through the element. 5 ma has been found to work well with many different types of pots. So the first thing to consider when selecting a pot is “What is the supply voltage in the circuit to which the pot is connected?” Many analog input cards run on +5 VDC. For a 5V analog card which applies 5V to the pot, a 1K value will flow 5ma in the pot. If you have a card which supplies 12V to the pot, you’d want a 2400 ohm pot. The nearest standard value pot would be a 2.5K, and that will flow a tad less than 5ma, but perfectly workable. Lesson learned: Pick the pot VALUE to get about 5ma of current flow in the pot, based on the voltage supplied to the pot by your analog input card. How is that determined? Ohm’s Law. E = I X R (or) I=E/R (or) R=E/I Linear or log taper? The “taper” of a pot refers to the manner in which the resistance changes along the length of the resistive element. “Linear” means that the incremental change in resistance is equal along the length of the element. “Log” means that the incremental change in resistance varies at a logarithmic rate along the length of the element. An “audio” taper pot is a form of log taper, selected to provide a satisfying change in audio volume as a person “turns the volume control”. (There is no electrical reason for this, it is purely to accommodate the human tendency to rotate controls approximately 90 degrees in one smooth motion.) For flight controls, we want as LINEAR a response as we can GET in the entire system, and for that we need pots with a linear taper. Rotary or slide pots? It depends. The maximum amount of linearity is obtained when the pot is operated over the maximum physical travel of the wiper. This means that we want to arrange our mechanical linkages and such so that we turn a rotary from “stop to stop”. If using a slide pot, you want the wiper to run “end to end”. This provides the widest range of input voltage to your analog card, and thus reduces the effect of non-linearity and quantization errors in the analog-to-digital (A/D) converters on the card. If you’re running a 5V card, you want the pot to provide the full zero-to-5-volts of input to the card as the pot is operated through it’s normal range. Here’s why: An A/D “reads” the input voltage in increments. If you divide 5 volts by 256 increments, each increment equals approximately .02 volts. This is an acceptably “small” step to detect tiny physical movements of the flight control. But if your pot linkage only moves the pot for 50% of it’s range, then the input card only has 2.5V of range to digitize. 2.5 V divided by 256 increments is only .01 volts. Thus the range of movement offered to your analog card is WORSE by a factor of DOUBLE. If you’ve had to adjust the gain of your flight control inputs to reduce the elevator sensitivity, for example, or tinker with the flight model in MSFS to smooth out the hand flying qualities of the Lear, this is likely the reason why. Lesson learned: Use any style pot you wish, but make sure that your mechanical linkage operates the pot over it’s full range. You will find this much easier to do with slide pots than rotary pots. Should I use shielded wire on my pots? Yes. Even if I am using low-impedance pots? Yes. Here’s why: The input impedance of the analog card is the limiting factor here – not the value of the pot. Most analog card inputs are relatively high impedance, so if you use say….a 10K or 100K pot, the entire circuit is more likely to be affected by electrical noise. So use shielded cable. It’s cheap, and it’s good insurance against such things as the cheap noisy power supplies found in nearly all computers. But wait! There’s more! There’s another reason why you don’t want to use high-value pots on an analog input. Let’s say you’re using a 100K pot on your elevator, and your analog card input has an input impedance of 50K. For the purposes of this example, park the elevator at exactly neutral pitch. Now consider the elevator pot. At the middle position, that pot should be sitting at half travel. For a 100K pot, that is the equivalent of two 50K resistors in series, with the wiper sitting in the middle. So that’s a 50K resistor from 5V down through the top “half” of the pot, out the wiper, and into the 50K analog input, which is in parallel with the “bottom” half of the pot. The combination of the analog input impedance and the bottom half of the pot effectively produces a resistance of 25K between the wiper and ground. NOW, think about that voltage divider. NOW it’s a 50K pot from 5V down to the wiper, and a 25K pot from the wiper to ground. What VOLTAGE would you observe at the wiper? It SHOULD BE exactly 2.5V with the elevator at neutral pitch, but in this example you will find it to be much less. How much less? Ohm’s Law will tell you, and once you figure that out you’ll see the point here. If your pot value isn’t compatible with your card input impedance, you will have non-linearity problems over the mechanical range of the flight control. Lesson learned: The input impedance of your analog card should be AT LEAST 10X the value of your pot to eliminate loading effects of the input. You can’t do anything about the input impedance of your card, but you can select a good low value pot to connect to it. For our example above, if your card has an input impedance of 50K, simply divide by 10. A 5K pot would meet THIS requirement, but 5K may not be a low enough value to flow 5ma in the circuit. You can always use a LOWER value pot to get the 5ma, and this will provide additional loading margin against the card input impedance. Perhaps the best piece of advice I can offer here is this: BUY QUALITY POTS! Cheap pots are notorious for poor repeatability, noisy wiper connections, temperature drift issues, and even simple mechanical failure. There are many occasions in simulator building when cheeeeeep components will work well. This ain’t one of them. BUY QUALITY POTS! Who knew pots were so much trouble? And we haven’t even discussed trim or autopilot inputs, yaw dampers, rudder boost, or ways to implement summing nodes for proportional toe brakes on both sides of the cockpit. Part Six: Floobydust Floobydust: “A collection of important but otherwise unrelated things.” • Transport category aircraft panel lighting is a color known in the biz as “Blue-White”. White LEDs are very near that color. Incandescent bulbs appear as what people call “warm white” (nearer the red end of the spectrum). Bulbs are used in switches, except in new aircraft like the 787, and those are LED-illuminated. • Can I use automobile seats in my sim? I did, but consider that auto seats are not “flat” on the bottom. They have all sorts of weird curves under them because the floor of an auto is not flat either. Also, the bottom cushion has no real structure to it so it’s impractical to try and cut out the “notch” for the control column. • How are you going to couple the autopilot servos to the physical controls? Cables? Mechanical linkages? If you want your controls to physically move under command of the autopilot you’re gonna need to plan ahead for this. Here’s a hint: Put your sensors on the crossover linkage (or cables), but not in the yoke(s). • We need a working replica of the standby compass to mount in the windscreen center structure. I think you can buy a replica compass from one of the sim community suppliers, but to my knowledge nobody has successfully mated one to the Lear. • Landing gear switch knob. I need one, and I imagine that many others do too. • As much as is possible try to keep the top of yoke columns clear of mechanical parts. A lot of wiring has to go in there and it takes great delight in getting itself tangled in the moving parts. This is not good, particularly at FL-290 in the dark. • How do you interface the trim switch on the control yoke? You use it to operate small motor-driven pots, the output of which is summed with your primary flight control sensors. • A couple of folks have asked, “I’m using a bellcrank at the top of my yoke to connect the crossover cables to the other side, but they don’t turn satisfactorily. The further you turn one yoke, the less movement is observed in the other yoke.” That’s because a bellcrank does not present a constant radius turn from the perspective of the attached cable. Use a pulley instead and it will work fine. • USE SCHEDULE 80 PVC PIPE! That’s the thick wall stuff. Home Depot, Lowe’s, etc., stock it in the plumbing section. And DON’T GLUE IT! USE SCREWS! You may need to take it apart sooner than you imagine. • You know how computers hate static electricity? You ARE using static-resistant carpet in your cockpit aren’t you……ahem. • And finally, “Please confine your crashes to the far end of the field. Thank You.” Part Seven: The Advantages of a Standard Design, (or) “Let’s Have Some Controversy Here People!” Question: Do you know how Henry Ford sold a lot of automobiles to people of ordinary financial means? Answer: He figured out how to manufacture those automobiles and sell them at prices that buyers could afford. You know this story. Henry Ford perhaps didn’t invent the concept of mass production using standardized parts, but he was the first to perfect it commercially in the automobile industry. The use of standardized parts and construction enabled Ford to sell reasonably priced automobiles to far more buyers than could otherwise afford them. The community of people who are building aircraft simulators presently suffers from the same “mostly-built-by-hand” costs and issues as automobile makers did prior to Henry Ford. It is not a sustainable model. I propose that we, as members of Hangar45, consider the means and methods that are known to reduce cost, effort, and ad-hoc engineering failures. One of our primary objectives is to reduce the cost of panels, wiring, hardware, and other specialized components. Have you added up the cost of a full set of panels, switches, pedals, throttles, cup holders, software, interface cards, and whatever else is needed to make them function? Protip: DON’T! It’ll discourage you. Another primary objective is to dramatically improve the success ratio of builders. It’s hard to build a functional sim cockpit in the present day. Think about all the tools and equipment and time and money and everything that you have invested in your aircraft. Dozens and dozens of builders have come and gone in our group, and yet Hangar45 has not a single example of a working sim that could be called complete and flyable. No we don’t. Not mine, not yours. Think about the barriers faced by builders. I’ll list a few. I’ll bet you can add to this list. • Builder skills. Not every builder can solder wiring or run a CNC machine. • Budgets. Even if you build a stripped-down cockpit, the cost is not trivial. • Lack of parts. Home Depot, Lowe’s, and e-Bay ain’t got everything we need. • Time. How much longer do you want to spend BUILDING your sim? • Finished product. Builders want something that will be a source of pride and achievement, but many don’t know how to achieve that result. We can do better than this. Think about the incredible variety and number of talented members of Hangar45. What if someone brought them to one location, equipped with tools and machines, supplied them with materials and parts, and said, “Go figure out how to mass-produce a Lear 45 simulator.” Think about the outcome. Think about what we could achieve, working together on a standardized design. Allow your imagination to run a bit. Why can’t we do this anyway, even though we’re scattered all over the globe? Bombardier does it. Boeing does it. Airbus does it. Even General Motors, Ford and Nissan does it. So can we. Back in December 2009, just before we packed up to move to South Carolina, I promised to write some words about connectors and wiring techniques and related topics. I’m still planning to do that, but there are a great variety of issues that should be addressed before a workable design is realized. Here are a few examples to get you thinking: • Where should the interface cards be located, and must we all use the same type? I built some panels with multipin connectors and wiring that connect to interface cards located in the avionics bay. I built other panels with the interface cards *in* the panel so that they only need a USB cable for a connection. Are either of these methods better? Why? • What kind of cable (wire) is appropriate for a wire harness? Some builders use wiring harnesses built by hand from separate pieces of wire. Some are using flat ribbon cable. Are you using shielded wire? Why or why not? Which of these are good design decisions? • How do you decide the pinouts to use for each connector? (Which wire goes on which pin.) If we all agreed on a common schematic and connector, we gain no benefit whatsoever unless we’re using a standardized pinout. • How is your panel wired? This means “schematic diagram”. Are you wiring your panel switches as common ground or common +Vcc? (Active high, or active low?) Did you use pull-up / pull-down resistors? Are you bringing each contact out to the connector, or just the one *you* think you’ll need? How about lamp inputs for the dimmers? Is yours a 5 volt or 28 volt dimmer panel? • Momentary or push-latch-push-unlatch switches? If you use a mechanically latched switch, MSFS cannot change the state of the switch when an internal variable changes how the sim is running. It ain’t good when your switch/indicator status doesn’t match what the airplane is doing. In complex systems design we use something called an Interface Control Document (ICD). The ICD is a document that describes everything necessary to build wiring, equipment, or systems that must connect to either side of the interface. Suppose I said to you, “I need a spare tire for my car, but it must be a 15-inch wheel with 5 lug nut holes.” That’s a simple example of an ICD specification. It tells you two of the most important parameters that one must know to obtain a spare tire that will fit and work on my car. Beyond that, ANY brand, color, price, etc. unit would work provided it meets that spec. We should be using some form of the ICD method for panels, wiring, linkages and such. This frees builders and suppliers from the task of having to do an original design for nearly every piece used in our cockpits. If we had an ICD for the electrical panel, for example, both builders AND manufacturers would know exactly how to fabricate and wire that panel, or harness, in a way that guarantees it will work when installed. The ICD specifies a connector type, the pinouts, and any mechanical details needed to locate and orient the connector on the rear of the panel. Another ICD will specify how the panel should be wired internally. This means, “a schematic”. If you build your panel to the connector ICD and wire it to the panel ICD, you gain the assurance that it will function as intended when connected to a wiring harness (also built to an ICD spec). The use of the ICD method makes it possible for builders or manufacturers to produce nearly any kind of panel they wish, yet it will still function correctly in the airplane so long as it conforms to the ICD spec. Can’t afford AML switches? No problem. Use whatever you can afford but wire them per the ICD and they’ll work in the airplane. Can’t afford backlighting? No problem. Make your panel face out of cardboard if you wish, and so long as you follow the ICD specs it will fit in the panel and function as intended. If you’re short on cash, you may elect to buy or build the “entry level” products. So long as they conform to the ICD spec, you gain the ability to upgrade to a nicer panel, TQ, or set of rudder pedals, by simply removing the “basic” and installing the “better”. The ICD method provides an amazing degree of flexibility for everyone, at many skill levels and costs. The use of ICDs would enable builders to get panels from Big Bird, wiring from Elmo, interface cards of their choice from Elmo, and software that runs it, and it would all function together as intended. All of these sub-systems become part of the standard design, with each element described and specified by one or more ICDs. This is a complex topic. To my knowledge no other (hobbyist) aviation simulator group has standardized a design across multiple suppliers and builders. As members of Hangar45, we have an opportunity here to demonstrate how it should be done. It is my intention to stir useful conversation and discussion. Given the talent we have here at Hangar45, I’m asking the question “Why haven’t we done this?” Comments encouraged. Ideas welcomed. Good ideas may get you 15 minutes of fame. Or an ICD Standards Committee assignment. DonnyRay Jones So You're Gonna Build a Learjet By DonnyRay Jones
2017-10-10