This is the first in a series of five articles that will provide an overview of an electronics system design process, using the example of a solar oven controller. The process described here works for both simple and complex designs. Complex designs usually merit more iterations and more layers of optimization, but the steps are about the same.
This first article will run through the process of defining the early requirements for a system — figuring out what to build and being clear about why. We’ll tackle that process twice: first, in a general way, at the highest thousand-miles-away level, and second, close-in, running through an example electronic system architecture design.
The second article will provide a summary of the process of designing a system architecture.
The third will delve into creating an actual layout for a real example design.
The fourth article will describe test and evaluation planning.
The final article will describe assembly and test of the first build of the example system, from fabrication, through initial inspection, and ending in testing.
Let's get started!
Step 1: Define the problem.
The first step is to fully understand the problem you’re proposing to solve with your design. That is:
- What need does the world have for your system?
- Who will use your solution?
- Who will be affected by it? and
- Who will affect it?
- Why are the approaches others have taken to solving this problem not already good enough?
If you’re making something for the purpose of entertaining yourself, this last question might be trivially easy to answer, but if you’re making a system you want to send out into the world to be useful in a sustainable way, you should really master the answers to all of these questions, with as much verifiable, trustworthy data as you can muster.
In most cases, this will be an iterative process: you will sketch in a set of requirements that rely on what you know. As you compose your set of initial requirements, you’ll identify things you need to find out. You’ll be able to research some of those questions yourself from existing pools of data or be able to pull together answers from appropriate experts.
But you’ll probably find that some of the questions can be best answered by simply trying things out. Those questions may lead to experiments with simple trial hardware solutions. You’ll need to define the requirements for your experiments and trial solutions as well. Then, as you experiment with your simple trials, you’ll realize you should refine some of the initial requirements. You’ll probably go through several rounds of experimentation and refinement of your requirements.
Let’s run through the first iteration of answering these critical questions for an example system. Our example system will be a solar oven controller. The discussion below will cover why we might want to add controllers to solar ovens. The second half of this post delves into why we think the world might need more solar ovens in the first place.
What is the problem we are trying to solve? The core problem being tackled by the solar oven controller is making solar cooking convenient relative to other cooking technologies.
The specific inconveniences of un-instrumented solar ovens that we want to ameliorate are:
- Uncertainty about the safety of food cooked in our solar cookers.
One of the easy ways to use a solar cooker is to set up a slow-cooked meal from recently refrigerated ingredients in the morning, leave the cooker unattended, and then retrieve a cooked meal from the cooker upon return many hours later. This food might have lingered too long at a temperature conducive to the growth of dangerous microbes.
Food kept cool (40°F or below) or warm (above 140°F) is pretty safe, but temperatures between these bounds can be welcoming to fast-growing populations of bacteria that cause disease, like salmonella. This is particularly challenging for unattended solar cooking. Without knowing how long the meal lingered in that “Danger Zone” between 40°F and 140°F, the slow-cooked meal is inherently risky.
Furthermore, the same food safety requirements for minimum cooking temperatures for meats, eggs, and seafood that apply with conventional cooking techniques apply to solar ovens too. Just as when cooking on a barbecue or in an oven, the cook must measure the internal temperature of the chicken or steak or other meaty thing to be really confident that it’s safe.
Without some electronics to ease the measurement, this can be even more inconvenient with a solar cooker than it is with a barbecue or oven because the range of times over which the food must be checked may be longer, the food may be harder to get to, and the cooker cools off each time it’s opened to check on the food inside.
- Inconvenience of knowing when the food is “done.”
Solar cooking is often much slower and more unpredictable than cooking in a microwave, a stovetop burner, or a gas or electric oven. Because the temperature of the oven is not as controlled as it would be in a regular oven, the food might reach its ideal state in as little as 30 minutes in the oven or as long as 4 hours after the user sets it up in the oven. Because of all this variability in cooking temperature, it’s easier to overcook some kinds of foods, especially baked goods. Avoiding overcooking might require a lot of labor. And to make the problem worse, when the cook opens the oven chamber to check on the food, heat escapes from the oven cavity, possibly significantly slowing the cooking.
Step 2: Define the users.
Who will use our solution, who will be affected by it, and who will affect it?
The target users for the solar oven controller will be people who:
- Want to cook with solar cookers
- Have any of a wide range of solar panel and box cookers (but not parabolic dish cookers, which this solution avoids due to safety concerns)
- Are cost-sensitive but are willing to make an outlay of several hundred dollars for a gas oven, a microwave, a toaster, etc.
- Are accustomed to programming microwave ovens, reading feedback like timer countdowns from their household devices, and can program or set an oven to cook at a specified temperature
- Have schedules with only a bit of flexibility for when food preparation and consumption will occur
Notably, although many of the users of solar ovens are people in areas where natural disasters, drought, or war have wiped out alternative infrastructure for cooking, the solar oven controller here is not aimed at them. We can imagine a different set of requirements for a controller for that population: it should be built out of locally available components that can be assembled without skilled labor, optimized for cost in the pennies per oven, and designed for a user who might be lacking familiarity with household devices equipped with embedded electronics. That’s a harder design problem and, worthy as it might be, is not within the scope of this design effort.
If we were making a commercial solar oven, we would now try to answer the remaining questions of who else will affect or be affected by our design by sketching in the community around our controller. Who will do the controller and oven assembly, distribution, installation, and configuration? Who will perform training? What will the community look like? If we were making a controller for charitable purposes, we’d probably have to figure out how to integrate it into the training and distribution programs of non-profits. But this particular controller is a free, open-source design for electronics and maker hobbyists and solar oven enthusiasts. We can, in this particular example, skip making plans for assembly or distribution of actual hardware, plans for training, and plans for cultivating a community of users.
Step 3: Look at other solutions.
Why are the approaches others have taken to solving this problem not already good enough?
Let’s separate this question into three parts:
- Why are common household devices for cooking food not already good enough?
- Why are solar ovens without electronic controllers or sensors not already good enough?
- Why are sensors or controllers that already exist not already good enough?
The preliminary answers to questions 1 and 2 are covered in the second half of this post. But why are sensors used with ovens and barbecues and other household cooking devices not good enough? The passive oven thermometers that measure the temperature of the air inside an oven don’t do a reliable enough job of reporting the temperature of the food being cooked in the oven and they also don’t provide history, both of which are essential to our food safety concerns.
Meat thermometers plunged into the food when it has been removed from the oven can answer the question of whether the internal temperature of a chicken or a soufflé has reached the minimum safe temperature, but they do little to help address the temperature Danger Zone and doneness issues. Remote digital thermometers are close to what we need: They can track the temperature, report when an internal temperature reaches the minimum safe temperature, and tell us when a meal has been within a temperature range for a programmed period. But we haven’t found any that track the time in the temperature Danger Zone and allow us to build models for doneness for food cooked over a wide and unknown temperature range.
So now that we have a description of the problem, it’s time to envision the requirements for a solution.
Step 4: Establish requirements.
Our top guiding requirement is for the solar oven controller to make the solar oven as appealing and convenient as competing household cookers with respect to safety of the food and understanding the status of the food. The solution will need to provide a way for the controller to determine what the user wants from the cooker. First, what will “safely cooked” require? Second, what will “done” mean? Since both of these targets vary depending on the food being cooked, the oven controller must easily adapt to the user’s needs each time the oven is used.
The solution will also have to provide a way for the controller to notify the user of the status of the food, specifically:
- When the food is nearing a hazardous condition (too long with the temperature in the Danger Zone) while the user still has time to intervene and save the food
- When the food has been exposed too long to the Danger Zone temperature range and should be discarded
- When the food is done
- How far along the food is in the process of being cooked
Further requirements that follow from our guiding convenience requirement are:
- Any burden placed on the user for specifying what they want done with the food should be no more onerous than the process for programming a microwave, pressure cooker, or sous vide system.
- Using the controller should require no training and leave little room for confusion.
- The controller should require no burdensome storage, setup, cleaning, or maintenance.
The controller must operate over a set of environmental conditions that is convenient for our expected user: It works outdoors, in bright sun or shade, across a reasonably wide range of outdoor temperatures, and is not dangerous or damaged by heavy rains. A controller for a solar oven that is permanently installed in a kitchen would not need to meet these last environmental requirements, but for now, the controller targeted here will work for both in-wall solar ovens and outdoor ovens, so we’ll tackle the more difficult of each set of environmental requirements.
We can also tentatively add some requirements that address our initiative to make solar ovens more appealing. We can target the appeal that comes from better understanding of how food is cooking. The new system can provide insight into what parts of the food are cooking at what temperature and see how temperature patterns affect the result.
The table below lists a set of preliminary solar oven controller requirements that translate the single “convenience and appeal” requirement into design requirements. For now, the table omits several categories of requirements that would have to be established if we were designing for commercial distribution, such as safety and regulatory requirements, consideration of reliability, upgrade paths, obsolescence, and most other market requirements.
This table does include one sort of market requirement: a fairly arbitrary parts cost target. The parts cost goal comes from the combination of wanting to make a solar oven that’s comparable in appeal to a microwave or stove and an estimate that if the parts cost is between $10 and $20, the controller parts/assembly and packaging labor/design/maintenance cost will be in the realm of a few multiples of $10 if making many controllers. A starting estimate is that a few multiples of $10 should be a reasonable cost for the controller of a household cooker that is comparable in appeal to its competition. If we intended to make the cooker commercially, we’d want to do a little more research into what the market seems likely to bear.
This table of preliminary requirements contains several elements that are based on guesses and assumptions. Some of these guesses can be turned into data through practical experiments, noted by the “needs experimentation” column. Most of these have to do with users’ perceptions of convenience.
We don’t know exactly what status information users will actually find useful. Finding out will require experimenting with representative hardware, real food, and real people. One assumption underlying the entire set of requirements is that such a system might make solar cooking more attractive to cooks. We’ll test that with the first experiment, too.
We want realistic, useful results, but not at the cost of years of experimentation and design iterations. We’ll aim for a balance by requiring that a round of experiments in the design cycle shouldn’t take more than a week or two — but we’ll temper that requirement by making sure we don’t cut off useful avenues of design only because we didn’t want to put the labor into checking on them. If we find something we really want to consider that is not practical to develop and test in a short period, we can then consider other ways to evaluate it.
Ideally, the design experiments will help answer these questions:
- In practice, how is it most convenient and effective for the user to indicate how they’d like the food cooked? We can experiment with a few ways of doing this: We’ll start with a pretty rich user interface with a graphic display that lets us try both very simple interfaces like those of most microwaves and more graphically rich displays. We’ll also make a rich user input interface to start with that allows the cook to enter times with individual digits, make selections from menus, and increase and decrease times and temperatures.
- Where should the interactions between the user and the controller take place? Is it convenient to provide the user interface on a user’s tablet or smartphone? Do users prefer an oven-mounted display? Do we need something hand-held that the cook carries around, like some commercial remote digital thermometers? Do we need several duplicate interfaces? We’ll take the opposite approach for this one: We’ll start with the simplest oven-mounted interface and require a second set of experiments if it proves inadequate.
- What kinds of notifications are convenient? Is a buzzer or beeper (like the timer on a microwave) sufficient for notifying a user that the food needs attention? Can it be oven-mounted? We’ll start with an oven-mounted beeper and consider a second set of experiments if it proves inadequate.
- What kind of status of the food is actually useful, appealing, and actionable? We’ll provide a rich graphical interface to enable experimentation with various types of food status.
- We’re also testing the overarching question: Can we make solar cooking more convenient and attractive by adding electronics?
Now that we know what we want to test first, we’ll make a new set of requirements: for a first prototype system that will be the subject of the remainder of the articles in this series. The function of the first prototype system will be to provide a platform we’ll use to try and answer our first round of questions. The table below lists the requirements for this one-off experimental prototype.
One of the critical design features may turn out to be how we handle power. A user study of the acceptability of having to plug the controller into a power outlet, having to replace or recharge batteries, etc., will not be part of this effort but likely would find a place in a second design round.
Introduction to Solar Ovens
If you haven’t run into solar ovens before, they’re devices for cooking with sunlight. Basically, they heat up food and liquids with the greenhouse effect. A solar oven is a sealed, insulated chamber where at least part of the surface is transparent to a wide spectrum of visible and ultraviolet light but that does not pass infrared light well.
A solar oven can be as simple as a black pot enclosed in a sealed plastic bag or as fancy as an evacuated tube surrounded by a sun-tracking solar concentrator. Decent solar ovens can boil and sterilize water, cook most kinds of food, and keep cooked food warm for many hours after cooking it. The best ones are well insulated and take in a lot of sunlight. They heat up well on sunny days, even out in the snow!
Below are four examples of solar ovens:
A commercially available lightweight box solar oven with reflector, the All American Sun Oven from Sun Ovens International.
A commercially available evacuated-tube solar oven with parabolic reflector, the GoSun Sport from GoSun.
A loaf of bread cooking in the author’s solar oven, which is made out of a square of glass, aluminum foil, Elmer’s glue, cardboard boxes, and black paint. It’s built from instructions on the Backwoods Home magazine website.
A solar oven made from a car windshield shade, a plastic bag, and a cookie cooling rack, from the Solar Cookers International Network.
A well-built solar cooker can be convenient to use for slow-cooking: A cook can put the ingredients for a single-dish meal into a solar oven in the morning, go away to work for many hours, and then come home to a slow-cooked meal. The food will stay warm until the diners are ready for it — which can even be long after dark for well-insulated cookers.
Solar cookers are incredibly useful in developing countries or disaster areas where supplying cooking fuel is problematic. Similarly, they can be handy on long hiking or camping trips. They can also be useful in the developed world for cooking on hot days when using an indoor stove or oven would be unpleasant. But for the most part, solar ovens in the developed world are niche devices, used only by a small population of solar-cooking enthusiasts, hard-core environmentalists, campers, and survivalists.
Basically, they solve a problem — how to cook appealing food conveniently while consuming minimal resources — which is already pretty well-solved by other technologies. Most of us have microwave ovens and gas or electric ranges that cook quickly, whenever we wish (or at least, any time that the electricity or fuel they require is available). Solar cookers are not as convenient, but we think that with some improvements, solar ovens can cook as well or better than more conventional cookers in many cases, and with equal or greater convenience.
Imperfections in Common Cooking Tools
Gas and electric ranges, microwave ovens, barbecue grills, and our other common household cooking devices have a lot to offer, but they also some critical imperfections that are worth pointing out
- Burdensome infrastructure needs:
- Gas ranges require gas to be piped in, with all the attendant materials, holes in walls and floors, and code requirements that necessitate inspection and testing.
- Electric and dual-fuel ranges may require a 220V outlet, the provisioning of which is similarly inconvenient.
- Ranges require a lot of space and often a built-to-fit opening in a countertop.
- Cooktops and gas ranges require vents that extend through the envelope of the building and often need unpleasantly noisy fans.
If you’re building a new or remodeled kitchen, all these needs can be very expensive!
- Availability or up-time isn’t perfect:
- When the power is out, electric ranges and microwaves can’t cook.
- And gas cookers also stop working when the gas runs out.
- Difficult to fine-tune:
- Our most common cooking devices (gas and electric cooktops and ovens, microwaves) don’t make it easy to cook to just the right level of doneness for some kinds of food, so our food is often over- or under-cooked. For example, a steak might be perfect if seared for two minutes at one temperature but overcooked at three minutes at the same temperature or two minutes at a slightly higher temperature.
- Sometimes, it’s really hard to get the cooking just right throughout the whole slab or batch.
- Don’t typically come with hardware that checks on the safety of cooked foods:
- The CDC estimates that about 48 million people in the U.S. suffer from foodborne illnesses each year — about one in every six of us. Illnesses acquired from food make up about 27% of all illnesses in the U.S.!
- The CDC also estimates that about half of foodborne illnesses are from meats, especially chicken. Some of those meats are hazardous because they weren’t brought to sufficiently high temperatures during cooking.
- Most cooking devices don’t provide easy-to-use tools that confirm when internal meat/fish/poultry/egg temperatures reach safe temperatures during and after cooking. Cooks can acquire and supply these themselves, but tools integrated with the cooker could be easier to use.
- The labor required for cooking, especially the timing, can be inconvenient and more than we want to dedicate to that task.
- Conventional cooking directly consumes fuel or electricity that costs money for every cooking attempt:
- In the U.S., cooking is directly responsible for about 2% of total energy consumed in the entire country.
- Just for cooking, we need the equivalent of the total energy output of about 77 coal power plants plus 207 petroleum plants, 303 natural gas plants, 6 nuclear plants, and 396 nuclear plants. (These details are all from the U.S. Department of Energy’s Buildings Energy Data Book, available online.)
- Stoves and ovens heat their surroundings as well as the food:
- The U.S. spends about 15% of its total energy consumption on cooling buildings, or about another seven times the numbers above. One survey concluded that about a quarter of cooling was needed because of things done within the building to further heat it up, like cooking.
- If we add in the additional energy needed to cool our homes after heating them up from cooking, we’d probably have another several hundred more power plants needed to support our domestic cooking needs.
- Conventional cooking contributes to greenhouse gases:
- S. residential cooking is directly responsible for contributing 31 million metric tons of CO2 to the environment each year, about 2.8% of the total residential contribution to CO2.
- Commercial cooking is responsible for another 13.8 million metric tons of CO2, or 1.4% of the commercial sector’s contributions.
Advantages of Existing Solar Cookers
In sharp contrast to this conventional way of doing things, solar ovens don’t need any additional energy to cook food. Nor do they generate any CO2 from fuel combusted to cook. They basically cook totally for free, once the cost of building the solar cooker is accounted for.
If we, in the U.S., used solar ovens for some ambitious but not wildly insane fraction of our total cooking needs, say 10%, we’d save 0.2% of our total U.S. residential energy, which would allow us to get by with 8 fewer coal plants, 51 fewer oil/gas plants, 40 fewer renewable energy plants, and a nuclear power or so fewer than we need right now!
That adoption rate isn’t necessarily a crazy one, either. Microwave ovens shine as an example of a cooking technology that was rapidly — and almost universally — adopted as soon as it became affordable and reasonably convenient. Fewer than 1% of U.S. households had a microwave oven in 1971, but 28 years later, more than 90% did, according to the U.S. Bureau of Labor Statistics. But microwaves weren’t adopted just because they cook food cheaply (they have the cheapest operating cost of any common food cooker, according to the State of California’s data). Microwaves appeal widely because they’re so fast.
And just like microwaves, in addition to cooking more cheaply than competing devices, solar cookers can do some things that are just better than other cooking devices. Most notably, solar cookers cook slowly and fairly steadily, just like electric slow-cookers and sous vide water-bath cookers. That means solar cookers have the same cooking advantages:
- Many kinds of foods just taste better when slow cooked. Inexpensive cuts of meat that would be too tough if cooked fast will become as tender as much more expensive meats. Stews and braises will take on more flavor. Vegetables can soak up the seasonings of their sauces. Slow-cooking at lower temperatures can also preserve some delicate flavors that higher temperatures would destroy.
- Prep-work for a solar-cooked meal can be shifted from the time just before eating to the morning before or even the evening before. This works best for stews, pot roasts, casseroles and soups — dishes that would take an hour or more to cook on the stovetop or in a gas or electric range.
Disadvantages of Existing Solar Cookers
In spite of all the advantages listed out above, solar cookers have some big disadvantages in their typical incarnation today:
- Most solar ovens are designed to be used outside and won’t work at all in most kitchens.
- Permanent through-wall or window solar ovens require an appropriately oriented, shade-free wall or window to be available and are therefore difficult for most kitchens to incorporate.
- Their performance depends on the weather, their location/orientation, and the time of day. They heat up only when bright, direct sun shines on them. This means that their performance can be hard to predict. Temperatures can vary widely depending on the brightness of the light falling on them and the orientation of the cooker relative to the sun’s angle. Also, depending on the cooker, they may be pretty sensitive to wind.
- They can’t be used for very-high-heat cooking, like stir frying. With a large parabolic dish reflector, solar cookers can reach high enough temperatures to flash-fry food, but the concentrated sunlight poses a safety risk, and none of the cookers of interest in this article series uses concentrators that are that strong. This restriction means the types of cooking that can be done are similar to that of a gas, electric, or microwave oven or a range top at moderate or low heat.
- Slow-cooked meals that are not monitored run a risk of lingering too long in the Danger Zone of temperatures, where hazardous populations of bacteria can grow. This is a much more significant hazard than it would be with a conventional oven, electric slow-cooker, sous vide cooker, or other conventionally powered/fueled cooker because the cooking process itself is often slow and unpredictable.
- With meaty dishes, solar cookers also run the risk of not bringing the meat to a high enough temperature to kill bacteria. This may be a greater risk than typically faced with more conventional cookers.
- For some kinds of solar cookers and some kinds of recipes, solar ovens require a lot of labor and inconvenience to cook food just the right amount..
More on Solar Cooking: Slow Cooking, Overcooking, and Convenience
Some recipes are ideal for unattended solar cooking: The cook can put the ingredients in the cooker in the morning, orient the cooker so it will be aligned with the sun at midday or a little closer to when they’ll want to consume the food, and then think no more of it. With a well-insulated cooker, food will stay cool until the sun shines brightly into the oven, will cook when the cooker is best aligned with the sun, and then will stay warm until the cook wants to take it out of the oven many hours later.
The easiest of these recipes are the ones that would be cooked slowly anyway — recipes that would take an hour or more if cooked on a range. They include stews, soups, pot roasts, and hard vegetables like potatoes and carrots. There might be one caveat: If the recipe calls for really long cooking, say a few hours on the range, the solar cooking version of it might need an adjustment of the cooker’s orientation part of the way through the cooking so that the food remains at the cooker’s higher temperature range for longer.
Some other recipes are also excellent for solar cooking with straightforward adaptations that are already commonly practiced for the steamy, moist cooking environment of slow cooking. These won’t overcook even after long periods in a typical solar oven.
Finally, some other recipes can be cooked in solar ovens, but not for long periods. These are things that will overcook if left in the solar oven too long: baked goods like bread, cookies, and cakes; tender green vegetables like peas, green beans, and asparagus; and fish. These can be especially inconvenient to cook in a solar oven. Cookies, for example, are typically cooked covered, so the only way to know if they’re ready in an un-instrumented solar oven is to open the oven and uncover them. Besides the inconvenience of repeatedly checking every few minutes when your cooker is outside in the glaring sun on a hot day, opening the oven lowers the temperature of your oven and makes the cooking time longer.
Making Solar Cookers Better
To compete successfully with widely accepting cooking technologies, solar cooking must be at least as good in safety, acquisition, setup cost, convenience, and cooked-food quality. Getting to that goal appears to require a lot of change:
- Re-orienting solar ovens must be far more convenient
- We need built-in solar cookers, with accompanying changes in design, skilled builders and installers, and code changes
- Cookers should provide more active control of insulation, pressure, and probably back-up heat sources
- Some culinary and cultural evolution will be required.
These are all reasonable problems to solve with some time and effort. But a few of them — notably, the safety concern arising from unattended, un-instrumented solar cooking and the labor and inconvenience related to knowing when food is done — appear easy to solve with a pretty simple electronics subsystem. That will be the main subject of this series of articles.