How to move something you can't even see
Suppose your chore for today is to empty a stable full of rank smelling horse manure. Not the nicest of jobs, so you'll want to do it as quickly as possible. You won't be able to move it all at once, because there's too much of it. To get the job done fast, you need to move as much manure as you can in one go. The best thing to do is use a wheelbarrow. Pile the manure up into the barrow, wheel the barrow outside, and then empty the manure into a pile in the stable yard. With a few of these trips, you can shift the manure from inside the stable to outside.
Moving something you can see is easy. But now let's give you a harder chore. Your new task is to move the heat from the inside of a refrigerator to the outside to keep your food fresh. How can you move something you can't see? You can't use a wheelbarrow this time. Not only that, but you can't open the door to get at the heat inside, or you'll let the heat straight back in again. Your mission is to remove the heat, continually, without opening the door even once. Tricky problem, eh? But it's not impossible—at least not if you understand the science of gases

How to move heat with a gas
Let's step sideways a moment and look at how gases behave. If you've ever pumped up the tires on a bicycle, you'll know that a bicycle pump soon gets quite warm. The reason is that gases heat up when you compress (squeeze) them. To make the tire support the weight of the bicycle and your body, you have to squeeze air into it at a high pressure. Pumping makes the air (and the pump it passes through) a little bit hotter. Why? As you squeeze the air, you have to work quite hard with the pump. The energy you use in pumping is converted into potential energy in the compressed gas: the gas in the tire is at a higher pressure and higher temperature than the cool air around you. If you squeeze a gas into half the volume, the heat energy its molecules contain fills only half as much space, so the temperature of the gas rises (it gets hotter).
What happens if you release a gas that's stored at high pressure? When you spray an aerosol air freshener, you've probably noticed that the spray is really cold—for exactly the opposite reason that a bicycle pump gets hot. When you release the gas, it is suddenly able to expand and occupy much more volume. The heat energy its molecules contain is now divided over a much bigger volume of space, so the temperature of the gas falls (it gets cooler).


The heating and cooling cycle
By compressing gases, we make them hotter; by letting them expand, we make them cooler. How can we use this handy bit of physics to shift heat from the inside of a refrigerator? Suppose we made a pipe that was partly inside a refrigerator and partly outside it, and sealed so it was a continuous loop. And suppose we filled the pipe with a gas. Inside the refrigerator, we could make the pipe gradually get wider, so the gas would expand and cool as it flowed through it. Outside the refrigerator, we could have something like a bicycle pump to compress the gas and release its heat. If the gas flowed round and round the loop, expanding when it was inside the refrigerator and compressing when it was outside, it would constantly pick up heat from the inside and carry it to the outside like a heat conveyor belt.

And, surprise surprise, this is almost exactly how a refrigerator works. There are some extra details worth noting. Inside the refrigerator, the pipe expands through a nozzle known as an expansion valve. As the gas passes through it, it cools dramatically. This bit of science is sometimes known as the Joule-Thomson (or Joule-Kelvin) effect for the physicists who discovered it, James Prescott Joule (1818–1889) and William Thomson (Lord Kelvin, 1824–1907). You won't be surprised to discover that the compressor outside the refrigerator is not really a bicycle pump! It's actually an electrically powered pump. It's the thing that makes a refrigerator hum every so often. The compressor is attached to a grill-like device called a condenser (a kind of thin radiator behind the refrigerator) that expels the unwanted heat. Finally, the gas that circulates round the pipe is actually a specially designed chemical that alternates between being a cool liquid and a hot gas. This chemical is known as the coolant or refrigerant.

Here's what's happening inside your refrigerator as we speak! The left-hand picture shows what's happening on the inside back wall of the chiller cabinet. The right-hand picture shows what's going around the back of the fridge on the outside.
1. The coolant is a liquid as it enters the expansion valve. As it passes through, the sudden drop in pressure makes it expand, cool, and turn into a gas (just like a liquid aerosol turns into a cool gas when you spray it out of a can).
2. As the coolant flows around the chiller cabinet (usually around a pipe buried in the back wall), it absorbs and removes heat from the food inside.
3. The compressor squeezes the coolant, raising its temperature and pressure. It's now a hot, high-pressure gas.
4. The coolant flows through thin pipes on the back of the fridge, giving out its heat and cooling back into a liquid as it does so.
5. The coolant flows back into the expansion valve and the cycle repeats itself. So heat is constantly picked up from inside the refrigerator and put down again outside it.





Old Refrigerators
If you look at the back or bottom of an older refrigerator, you'll see a long thin tube that loops back and forth. This tube is connected to a pump, which is powered by an electric motor.
Inside the tube is Freon, a type of gas. Freon is the brand name of the gas. This gas, chemically is called Chloro-Flouro-Carbon or CFC. This gas was found to hurt the environment if it leaks from refrigerators. So now, other chemicals are used in a slightly different process (see next section below).
CFC starts out as a liquid. The pump pushes the CFC through a lot of coils in the freezer area. There the chemical turns to a vapor. When it does, it soaks up some of the heat that may be in the freezer compartment. As it does this, the coils get colder and the freezer begins to get colder.
In the regular part of your refrigerator, there are fewer coils and a larger space. So, less heat is soaked up by the coils and the CFC vapor.
The pump then sucks the CFC as a vapor and forces it through thinner pipes which are on the outside of the refrigerator. By compressing it, the CFC turns back into a liquid and heat is given off and is absorbed by the air around it. That's why it might be a little warmer behind or under your refrigerator.
Once the CFC passes through the outside coils, the liquid is ready to go back through the freezer and refrigerator over and over.


Today's Refrigerators
Modern refrigerators don't use CFC because CFCs are harmful to the atmosphere if released. Instead they use another type of gas called HFC-134a, also called tetrafluoroethane. HFC turns into a liquid when it is cooled to -15.9 degrees Fahrenheit (-26.6 degrees Celsius).
A motor and compressor squeezes the HFC. When it is compressed, a gas heats up as it is pressurized. When you pass the compressed gas through the coils on the back or bottom of a modern refrigerator, the warmer gas can lose its heat to the air in the room.
Remember the law of thermodynamics.
As it cools, the HFC can change into a liquid because it is under a high pressure.
The liquid flows through what's called an expansion valve, a tiny small hole that the liquid has to squeeze through. Between the valve and the compressor, there is a low-pressure area because the compressor is pulling the ammonia gas out of that side.
When the liquid HFC hits a low pressure area it boils and changes into a gas. This is called vaporizing.
The coils then go through the freezer and regular part of the refrigerator where the colder liquid in the coil pulls the heat out of the compartments. This makes the inside of the freezer and entire refrigerator cold.
The compressor sucks up the cold gas, and the gas goes back through the same process over and over.


How Does the Temperature Stay the Same Inside?
A device called a thermocouple (it's basically a thermometer) can sense when the temperature in the refrigerator is as cold as you want it to be. When it reaches that temperature, the device shuts off the electricity to the compressor.
But the refrigerator is not completely sealed. There are places, like around the doors and where the pipes go through, that can leak a little bit.
So when the cold from inside the refrigerator starts to leak out and the heat leaks in, the thermocouple turns the compressor back on to cool the refrigerator off again.
That's why you'll hear your refrigerator compressor motor coming on, running for a little while and then turning itself off.
Today's refrigerators, however, are very energy efficient. Ones sold today use about one-tenth the amount of electricity of ones that were built 20 years ago. So, if you have an old, old refrigerator, it's better to buy a new one because you'll save money (and energy) over a long period of time.


For more information go to:
• Argone National Laboratory - Ask A Scientist (http://newton.dep.anl.gov/newton/askasci/1993/eng/ENG30.HTM)

• Mr. Hand's 8th Grade Science Site (www.mansfieldct.org/schools/mms/staff/hand/heatrefrig.htm)

• How Stuff Works - Refrigerator (www.howstuffworks.com/refrigerator.htm)

• Science Treasure Trove - refrigerator page (www.education.eth.net/acads/treasure_trove/refrigerator.htm)

Reference

1.www.explainthatstuff.com
2.www.energyquest.ca.gov

What are micro controllers? They are what their name suggests. Today they can be found in almost any complex electronic device - from portable music devices to washing machines to your car. They are programmable, cheap, small, can handle abuse, require almost zero power, and there are so many variaties to suit every need. This is what makes them so useful for robotics - they are like tiny affordable computers that you can put right onto your robot.

In a pure sense, a micro controller is just an IC (integrated circuit, or a black chip thing with pins coming out of it). However it is very common to add additional external components, such as a voltage regulator, capacitors, LEDs, motor driver, timing crystals, rs232, etc to the basic IC. Formally, this is called an augmented microcontroller. But in reality, most people just say 'microcontroller' even if it has augmentation. Other abbreviations would be ucontroller and MicroController Unit (MCU). Usually when I say 'microcontroller' what I really mean to say is 'augmented microcontroller.'

As a beginner it is probably best to buy an augmented micro controller. Why? Well because they have tons of goodies built onto them that are all assembled and debugged for you. They also often come with tech support, sample code, and a community of people to help you with them. My micro controller parts list shows the more popular types that you can buy. They tend to cost from $30 to $150 depending on the features. This will give you a good introductory to micro controller programming without having to be concerned with all the technical stuff.

In the long term however you should build your own augmented microcontroller so that you may understand them better. The advantage to making your own is that it will probably cost you from $10-$30.

Between getting a full augmented board and doing it yourself is something called a development board. These boards come pre-augmented with just the bare basics to get you started. They are designed for prototyping and testing of new ideas very quickly. They typically cost between $15 and $40.

There is a huge variety of microcontrollers out on the market, but I will go over a few common features that you will find useful for your robotics project.

For robots, ore important than any other feature on a microcontroller, is the I/O ports. Input ports are used for taking in sensor data, while output is used for sending commands to external hardware such as servos. There are two types of I/O ports, analog and digital.

Analog Ports are necessary to connect sensors to your robot. Also known as an analog to digital converter (ADC), they recieve analog signals and convert them to a digital number within a certain numerical range.

So what is analog? Analog is a continuous voltage range and is typically found with sensors. However computers can only operate in the digital realm with 0's and 1's. So how does a microcontroller convert an analog signal to a digital signal?

First, the analog is measured after a predefined period of time passes. At each time period, the voltage is recorded as a number. This number then defines a signal of 0's and 1's as shown:

The advantage of digital over analog is that digital is much better at eliminating background noise. Cell phones are all digital today, and although the digital signal is less representative than an analog signal, it is much less likely to degrade since computers can restore damaged digital signals. This allows for a clearer output signal to talk to your mom or whoever. MP3's are all digital too, usually encoded at 128 kbps. Higher bit rates obviously mean higher quality because they better represent the analog signal. But higher bit rates also require more memory and processing power.

Most microcontrollers today are 8 bit, meaning they have a range of 256 (2^8=256). There are a few that are 10 bit, 12 bit, and even 32 bit, but as you increase precision you also need a much faster processor.

What does this bit stuff mean for ADC? For example, suppose a sensor reads 0V to an 8 bit ADC. This would give you a digital ouput of 0. 5V would be 255. Now suppose a sensor gave an output of 2.9V, what would the ADC output be?

Doing the math:

2.9V/5V = X/255

X = 2.9*255/5 = 148

So how do you use an analog port? First make sure your sensor output does not exceed your digital logic voltage (usually 0V -> 5V). Then plug that output directly to the analog port.

This bit range could also be seen as a resolution. Higher resolutions mean higher accuracy, but occasionally can mean slower processing and more succeptability to noise. For example, suppose you had a 3 bit controller which has a range of 2^3=8. Then you have a distance sensor that outputed a number 0->7 (a total of 8) that represents the distance between your robot and the wall. If your sensor can see only 8 feet, then you get a resolution of 1 bit per foot (8 resolution / 8 feet = 1). But then suppose you have an 8 bit controller, you would get 256/8=32 ~ 1 bit per centimeter - way more accurate and useful! With the 3 bit controller, you could not tell the difference between 1 inch and 11 inches.

Digital ports are like analog ports, but with only 1 bit (2^1=2) hence a resolution of 2 - on and off. Digital ports obviously for that reason are rarely used for sensors, except for maybe on/off switches . . . What they are mostly used for is signal output. You can use them to control motors or LED's or just about anything. Send a high 5V signal to turn something on, or a low 0V to turn something off. Or if you want to have an LED at only half brightness, or a motor at half speed, send a square wave. Square waves are like turning something on and off so fast that its almost like sending out an analog voltage of your choice. Neat, huh?

This is an example of a square wave for PWM:

These squarewaves are called PWM, short for pulse width modulation. They are most often used for controlling servos or DC motor H-Bridges.

Also a quick side note, analog ports can be used as digital ports.

A serial connection on your microcontroller is very useful for communication. You can use it to program your controller from a computer, use it to output data from your controller to your computer (great for debugging), or even use it to operate other electronics such as digital video cameras. Usually the microcontroller would require an external IC to handle everything, such as an RS232. To learn more, read my microcontroller UART tutorial.

A timer is the method by which the microcontroller measures the passing of time - such as for a clock, sonar, a pause/wait command, timer interrupts, etc. To learn more, read my microcontroller timer tutorial.

I^2C

I^2C (pronounced 'I-squared-C') is also useful for communicating, but I have never used it. Just make sure your controller has some method of communicating data to you for easy and effective debugging/testing of your robot programs. Its actually somewhat complicated, but usually the manufacturer has simplified it so all you have to do is plug-n-play and do a few print statements. To learn more, read the I^2C tutorial.

To run a DC motor you need to either have an H-Bridge or a Motor Driver IC. The IC is great for small robots that do not exceed 1 or 2 amps per motor and the rated motor voltage is not higher than about 12V. The homemade H-Bridge would need to be used if you wanted to exceed those specs. There are a few H-Bridge controllers commercially available to buy, but usually they are way too expensive and are designed for battlebot type robots. The IC is small, very cheap, and can usually handle two motors. I highly recommend opting for the IC. Also, do not forget to put a heatsink onto the motordriver. Motordrivers give off pretty fireworks when they explode from overheating =)

Another interesting note, you can stack IC's in parallel to double the allowable current and heat dissipation. Theoretically you can stack as many as you want, as long as the current is high enough to still operate the logic of the IC. This works for voltage regulators too.

Im referring to anything that can be used for debugging by communicating information to you. LED's, buzzers, LCD screens, anything that gives output. The better the indicator, the easier the debugging. The best indicator is to have your robot tethered and print or data log sensor and action data to your computer, but it isn't always possible to have your robot tethered.

The lowest form of programming languages is the machine language. Microcontrollers need to be programmed with this.

An example of machine language:

3A 10 51

E6 DF

32 38 00

Obviously neither of us could ever memorize what all those seemingly random numbers and letters do, so we would program in a higher language that makes much more sense:

If (language = = easy)

print "yay!";

These higher languages would then be compiled automatically into a machine language, which then you can upload into your robot. Probably the easiest language to learn would be BASIC, with a name true to itself. The BASIC Stamp microcontroller uses that language. But BASIC has its limitations, so if you have any programming experience at all, I recommend you program in C. This language was the precurser to C++, so if you can already program in C++, it should be really simple for you to learn. What complicates this is that there is no standard to programming microcontrollers. Each has its own features, its own language, its own compiler, and its own uploading to the controller method.

This is why I do not go into too much detail because there are too many options out there to talk about. The support documents that come with the controllers should answer your specific questions. Also, if you decide to use a PIC, understand that the compiler program (at least the good ones) can cost hundred of dollars. Most microcontrollers also require a special interface device between your computer and the chip for programming which could also cost from $10-$40.

With possibly the exception of DC motors, the microcontroller is the most expensive part of your robot. There is just no escaping the costs, especially for the beginner. But remember, after buying all this for your first robot, you do not need to buy any of it again as you can reuse everything. So here is the breakdown of costs. The chip itself, without augmentation, would only cost dollars. But understand the chip is useless without the augmentation, so you would need to do it yourself if you do not buy it already augmented. This could potentially cost just as much with the augmentation, and could cause you many frustrations.

If however you are more experienced (and for some odd reason still reading this), you can customize your own circuit to do exactly what you want. Why have a motordriver when you are only using servos anyway? If you decide to buy an augmented MCU, the cost will range from about $50-$150. To compile your program, you would need to get special compiling software. Atmel and BASIC Stamps have free compilers. PIC's however have fairly expensive compilers. There are some free ones available online, but they are of poor quality in my opinion. CCSC PIC C compiler is about $125, but I think it is worth getting if you are going to use PIC's.

You will also need an uploader to transfer the program from your computer to the chip. This generally requires more special software and a special interface device. The Cerebellum PIC based controller has this built in which is really nice and convienent, but for any others expect to spend from $10-$40. People often opt to just make their own as the circuit isnt too complicated.

As a prototyper, what you probably want most is a MCU development board. These augmented microcontrollers are designed for the prototyper in mind. To find these augmented MCU's, do a search for 'pic development board,' 'atmel development board,' 'stamp development board,' etc.

If you have more specific questions about microcontrollers, or would like me to go into more detail about something, just write me and I will.

I've created a microcontroller product, called the Axon, that's both easy to learn and powerful in features. I use it for all my robot creations now, and will continuously release source code updates and tutorials using it. Feel free to check it out!

The engineering technology faculty at Wake Technical College undertook a study in the fall of 1978 to determine if our basic science and mathematics offerings were relevant to graduates' needs on the job. Since 1964, when Wake admitted its first engineering technology students, the engineering technology division had expanded to six fully accredited two year associate degree curricula with over 200 students enrolled. Feedback from employers and graduates indicates that the curricula are equipping graduates with the necessary entry ¬level skills. The explosion in technological information, however, has placed demands on two year ET curricula to include more state of the art subjects at the expense of fundamental science and mathematics subjects. Since only a limited number of topics can be covered in two years, ET curriculum planners must scrutinize subject matter to ensure that it does help to prepare students for jobs as science and engineering technicians, and to avoid technical obsolescence as their field changes.

We surveyed graduates of Wake's six ET programs and their employers to learn what they considered the basic science and mathematics topics most needed by engineering technicians on the job. We also sought to obtain comments about topics not listed on the survey which may be needed.

Of the 697 participants selected to receive our questionnaire, 470 had graduated from one of the six ET programs at Wake from 1969 through 1977, and 227 were employers of graduates of these programs. The questionnaire was drafted by a group of department heads and a second group of people involved with two and four year ET programs nationwide.

Results

Table 1 summarizes the basic science and mathematics topics needed by engineering technicians, as determined by the 29 percent of the enployers and 23 percent of the graduates who responded to the questionnaire. The findings are based on response patterns for a given item in which at least the group of employers or the group of graduates agreed with the combined group of respondents by a majority response in eitherthe essential (E), desirable (X) or not needed ( ) categories.

1) The strongest support for the items under mechanics came from respondents in the architectural, chemical, civil engineering, and industrial engineering technologies.

2) The items under the fundamentals of electricity/electronics were unani¬mously supported by respondents in the computer, electronic engineering, and industrial engineering technologies.

3) All groups of respondents supported the study of the general theory of

light, but only the electronic engineering technology respondents indicated support for all the items under light.

4) The study of the items under sound was supported by three groups of respondents: architectural, computer, and electronic engineering technologies.

5) All groups of respondents supported the study of heat.

6) Modern physics was important only to responding chemical technicians and electronic engineering technicians.

7) Only the chemical technology respondents supported the study of the chemistry subjects.

8) Items listed under biology were needed only by chemical technicians.

9) Civil engineering technicians were the only group who needed a knowledge of all the items under geology.

10) The two items under data processing were important to all but architectural technicians.

11) The study of algebra, trigonometry, logarithms, geometry, analytic geometry, and calculus was supported by all respondents.

12) The chemical, civil, electronic, and industrial engineering technology respondents indicated support for the items under statistics.

At the end of the questionnaire, the study participants were given the opportunity to make further comments, such as to be more specific with regard to certain topics or to list further topics they thought should be included.

In general, their comments addressed specific skills and knowledge required by technicians to do well in their jobs. The comments did reflect an awareness of the rapidly changing requirements in engineering technology and an appreciation of the value of basic science and mathematics in keeping abreast of these changes.

In addition to determining the basic science and mathematics topics most needed by engineering technicians, the study revealed several other trends:

Graduates and employers in all six engineering technology fields indicated that a knowledge of mathematics ranging from algebra to calculus was important for engineering technicians. The extent to which a certain mathematical topic was important depended upon its direct usefulness in solving day to day problems on the job. Support for the study of other mathematical topics resulted from a need for a foundation in mathematics which would afford the technician an opportunity to keep abreast of technological changes, as well as to develop analytical skills.

The respondents believed that an engineering technician needs a knowledge of basic science topics, which provide a foundation for applying skills and knowledge in their particular field. For example, chemical technicians indicated support for a study of the basic science of chemistry. Electronic technicians, on the other hand, indicated an interest in the fundamentals of electricity and electronics that explain the electrical phenomena associated with the application of electronics and electricity.

In the case of data processing, all participants except those in architectural technology believed that a knowledge of at least one scientific programming language was important. In addition, respondents indicated an interest in the study of COBOL.

Analysis of the response patterns of employers and graduates showed that graduates were more supportive of a knowledge of basic science and mathematics topics. Employers, on the other hand, tended to support only those topics that

were immediately useful in solving day to day problems. This difference in response patterns can be attributed to the desire of engineering technicians to stay abreast of technological change, while their employers appear interested primarily in the knowledge and skills that contribute to immediate productivity.