The Outside World The computer in an embedded system interacts with the other parts of the system, receiving and sending various types of electrical signals. The input signals typically represent some aspect of the outside world, such as position, speed, temperature, force, communications input, user input, or whatever. The output signals may control some mechanical device, varying the speed or direction of motion of a motor, turn on or off power to a heater, communicate with other devices, and so on, and will often display status/results to the user. Setting up these signals and connecting them to the embedded computer is the job of the hardware folk (poor devils), but it is very desirable that software developers have an understanding of the basic issues. So here is a quick summary of some of the main ideas involved : Review of basic electricity/electronics Safety Inputs Outputs Real Time & Embedded Systems 1.1 CA480 Dublin City University 2007 The Electric Review (1) Electric charge comes in two types, ‘plus’ and ‘minus’. The basic unit is the charge on a fundamental particle, the electron, which happens to have a minus charge. This is too small for everyday use, so the ‘Coulomb’, which is the positive equivalent of about 6,000,000,000,000,000,000 electron charges, is used instead. When charges move there is an electric current. This could simply be measured in Coulombs per second, but for some reason a new unit is used, the Ampere, shortened to Amp, or just A (he was a notoriously Absent-minded French scientist). One A is just shorthand for ‘one Coulomb per second’ – very helpful. It’s just like water in a river – Coulombs instead of litres, Amps instead of litres per second. Note current is not the speed of flow, but rather the quantity that goes past per second. The Amazon is not flowing very quickly as it enters the sea, but the number of litres per second is huge. The number of movable electrons in a metre of e.g. household wire is enormous, like a vast river, and a current of 1 A corresponds to the electrons moving along at a speed of only a few metres per month. Real Time & Embedded Systems 1.2 CA480 Dublin City University 2007 The Electric Review (2) If the electrons in a wire drift along very slowly, how can electrical devices react so quickly? The electrons in the wire repel each other so savagely, and are so light, that they behave like a very rigid rod – start one end moving with a push, and the other end moves almost immediately. Push one end of a solid rod so as to start it moving slowly. The push travels down to the other end at roughly the speed of sound in the rod, maybe about 1 mile per second, and the other end starts also to move slowly almost at once. The push travels fast, even though the rod travels slow. Push the electrons at one end of a wire, and the push travels to the other end at about two thirds the speed of light, about 120,000 miles per sec (the actual speed depends on many things, but can’t exceed that of light). A billion (i.e. 1,000,000,000) seconds is roughly 31 years. So a nanosecond, I.e. a billionth of a second, is to a second as a second is to 31 years. A computer with a 1 GHz clock does something new every nanosecond. In that time an electric push can only travel about 20 cms. So if a signal is to be sent from A to B and an answer received back in one nanosecond, A and B need to be closer than 10cms. This is one reason why it can be desirable for computers to be small. Real Time & Embedded Systems 1.3 CA480 Dublin City University 2007 The Electric Review (3) If a solid rod is moving, it can be difficult to stop it, or to make it go faster, because of its inertia, which we measure by its mass. Electrons are so light compared to electric forces that they can be stopped moving very quickly – very high frequencies are possible. There is a catch though Moving electrons cause a magnetic field, and when the current changes the magnetic field changes. And a changing magnetic field causes an electric force, which will try to resist the change in the current – a bit like inertia. By arranging a wire as much as possible in the magnetic field of its own current, e.g. by shaping it into a coil, this effect can be made quite large. Don’t try to stop the current in a coil quickly – give it a chance to die off. This property is known as inductance, and if signals are to change quickly inductance needs to be kept low. Interacting electric and magnetic effects are also involved in transformers, and are the basis for electromagnetic waves, I.e. radio, light, X-rays and so on. Real Time & Embedded Systems 1.4 CA480 Dublin City University 2007 The Electric Review (4) So what gets electric charges to move – what causes a current to flow? In the case of the river, which only flows downhill, gravity is the explanation. Electric charges feel gravity, but also an enormously stronger force, the electric force (we can forget about gravity by comparison). This force is the source of the atomic bomb’s energy, for example, and should be treated with respect at all times. All electrically charged objects exert an electric force on each other – the force gets stronger as the charges get nearer, and weaker as they get further apart, but never dies out entirely. The force can either push or pull - like charges repel each other, unlike charges attract. Luckily most objects, including ourselves, have equal amounts of positive charge (at the centres of atoms) and negative charges (the electrons on the outsides of atoms) and the repel and attract forces exactly cancel out. This is just as well. Electric charges also feel a force if they move in an area where there is a magnetic force, or if the magnetic force changes. Note that for magnetism to cause an electric force, there must be a change, either in position or in the size of the magnetic force. It turns out that electric and magnetic forces are really just different aspects of the same thing, the electromagnetic field, but that’s another story. Real Time & Embedded Systems 1.5 CA480 Dublin City University 2007 The Electric Review (5) So how to measure the size of the electric ‘push’ that is trying to get the current to flow? Standard way is to look at the energy involved in the charge going from A to B, the so called ‘potential difference’ – the ‘potential’ refers to ‘potential energy’. This will depend on the amount of charge involved, so let’s talk in terms of unit charge being moved. If one Coulomb gets one Joule of energy in getting from A to B, we say the potential difference is one Volt, or one V (after Volta, an Italian scientist), and we talk about voltage. Looking at energy difference is fine for the overall push, but what about the actual force, the electric force strength - how big is it. Energy change is got by multiplying force x distance, so we can measure the electric force in units of Volts per metre. For a given voltage difference, if the distance between A and B is large, the electric force will be small, and if the distance is small, the electric force will be large. In objects that are very small, voltage differences must be kept small, to keep down the maximum electric force. Circuits on chips are now so tiny that voltages are often kept below 2V. Notice also that voltage refers to energy per coulomb – if the voltage is very high but the charge very small, so that the energy involved is tiny, objects like humans may be little affected, although objects with small features, such as chips, may be badly damaged. Real Time & Embedded Systems 1.6 CA480 Dublin City University 2007 The Electric Review (6) So how hard is it, for a given ‘push’, to get charge to go onto B? As we get charge onto B, it repels the next lot of charge coming on, and does so very strongly. The bigger B is however, the more the existing charge can spread out, and the easier it will be to get more charge on. So the ‘capacitance’ of B will depend on its size, and it turns out it also depends on its shape. We can measure the capacitance of B in terms of the amount of charge we can get onto it per Volt of push, in units of Coulombs per Volt. To help ensure the writers of glossaries of technical terms have enough to do, this is named the Farad, or F (after Michael Faraday, a great scientist and admirable man). Most objects have capacitances so small they can be regarded as zero. Some devices, known as capacitors, allow both positive and negative charges to be stored close to each other, and are shaped in such a way that the repulsive forces are mostly cancelled by attractive ones. These can store useful amounts of charge even at low voltages. They are widely used in electronics, often as a kind of reservoir to damp out fluctuations in the supply voltage (decoupling capacitors). Because the capacitance of most things is so small, if charge does get transferred onto them, say by rubbing them against something else, a very high voltage results, sometimes big enough to cause a spark, as the charge jumps through the air to somewhere it can spread out more. Real Time & Embedded Systems 1.7 CA480 Dublin City University 2007 The Electric Review (7) So how hard is it, how much ‘push’ is needed, to get charge to travel through the connection between A and B? We might expect that some energy will be used up just in getting charge to travel through a connection, and that there will be a voltage difference between the two ends of the connection as long as a current is flowing. This is generally true, although special ‘superconductors’ are an exception. We might expect that this voltage difference will be greater for bigger currents. This also is generally true, though again there are exceptions. What few expected was that for many materials, the voltage difference is just proportional to the current. Double one, you double the other, and so on. Poor Ohm, a German academic who suggested this in the early 1800s, was sacked for suggesting something so ridiculously simple, and only re-instated more than ten years later. Ohm’s famous law, Voltage = Current x Resistance, or V = IR, where the voltage V is in Volts, the Current I is in Amps, and the ‘Resistance’ R is in Ohms, is probably the most widely used formula in electronics today. The resistance of a connection depends on the ‘resistivity’ of the material from which it is made, the thickness of the connection (wide pipes have less resistance to flow), and its length (long pipes have more resistance). Real Time & Embedded Systems 1.8 CA480 Dublin City University 2007 The Electric Review (8) Few (if any) physical properties have as wide a range of values as the resistivity of materials. Metals generally have low resistivity (silver and copper in particular). They are good conductors. Some substances have middle resistivity – ordinary water with the usual amount of impurities, the fluids of which biological organisms are mostly made, semiconductors like silicon with small amounts of special impurities added, and so on. Some substances have very high resistivity – they are good insulators. Rubber, most plastics, glass, dry paper, dry air, the vacuum are all examples. Even a high voltage will only manage to force a tiny current through them. It is worth noting that there are no perfect insulators. A big enough electric force (Volts per metre) will ‘break down’ any insulation, and let a catastrophically large current flow if enough charge is available. Very good insulators can cause problems, any charges rubbed onto them cannot leak away easily, and high voltages can result – static electricity. In humid environments condensation can cause problems due to leakage currents. On the other hand very dry environments encourage static due to lack of leakage currents. So humidity control is a good idea. Real Time & Embedded Systems 1.9 CA480 Dublin City University 2007 The Electric Review (9) Getting a charge to go from A to B is all very well, but the capacitance of B is typically so small that only a very small charge will be involved. It’s more interesting to get the charge to move around in a circuit. This current may be driven by a chemical reaction in part of the circuit (a battery), where one compound in the battery wants to give out electrons, and another wants to take them in. Or it may be driven by a changing magnetic field, as in a generator or a transformer. Or it may be generated by light falling on special materials, and so on. The total voltage ‘pushing’ the current around will be equal to the sum of the voltages across the different parts of the circuit, including the voltage needed to push current through the voltage source itself (the internal resistance of the source). The most basic circuit in electronics, and in many ways the most common, is arguably the ‘potential divider’, where the same current goes through two components connected ‘in series’. The voltage across each component is then the same fraction of the total voltage as its resistance is of the total resistance. By varying the resistance of one component the voltage across it can be made to change. All the electronic devices we deal with involve circuits. You should know the basic rules for parallel and series connections and power – consult ‘The Art of Electronics’ or the WWW if you’ve forgotten Real Time & Embedded Systems 1.10 CA480 Dublin City University 2007 Safety, Safety, Safety (1) Electric current heats the substance it passes through. Too big a current can cause fire. Wires must be thick enough (low resistance) to carry the maximum current without overheating. To ensure this, circuits include fuses, thin wires which melt first and breaks the current if it exceeds a safe limit. Never bypass a fuse. Human nerves use electrical impulses to tell the muscles to contract. An electrical current passing through the body can have the same effect, and cause muscles to contract very strongly and lock up out of control. This ‘electric shock’ can be very painful. Worse, since the heart is a muscle, and breathing depends on muscles also, it can be fatal. Even a small torchlight battery of 1.5 V will pass a current through the human body. The body’s resistance is high enough to ensure this current is too small to be noticed. Voltages less than 20V or so are unlikely to cause a big enough current through human tissues for any electric shock to be felt. But the mains voltage of 230V can push enough current through the human body to stop the heart and stop breathing, and cause some burning, and is lethal. Many low voltage supplies in computers (also e.g. car batteries) can put out big currents, in excess of 100A. They won’t give an electric shock, but If a finger ring or metal watch strap shorts out such a supply, the resulting high current can make the ring or strap red hot, or even melt it, resulting in the loss of a finger, or badly damaged wrist and paralysed hand. Real Time & Embedded Systems 1.11 CA480 Dublin City University 2007 Safety, Safety, Safety (2) The main danger of electric shock is from the mains 230V. One side of the this supply is connected to earth by the power company, so unless you are floating in mid-air, you personally are connected to the mains supply. The connection is really good (low resistance) if you are standing on wet ground. If you now touch the ‘live’ side of the mains, the one trying to send a current to the side connected to earth, it will see you as an attractive alternative path, and send a current through you. This is best avoided (understatement). This current can be cut off quickly by an ‘earth leakage trip’ or ‘residual current detect RCD’ device if this is installed. So : Ensure that the mains supply to any device you are working on has an RCD – if not, or if uncertain, buy one of the plug-in types. Ensure a fuse or circuit breaker is protecting wiring from excessive currents, which can cause fires. Ensure the metal case is earthed, so that it cannot become live if accidentally put in contact with the mains. Wear nothing metal on hands or wrists – a molten ring will remove a finger in a second. Wrists are worse. Most important, take care - RCDs fail, earth links work loose, … And if you see someone getting a shock, get the supply turned off before helping the victim, or you may become a victim yourself. Real Time & Embedded Systems 1.12 CA480 Dublin City University 2007 Safety - Handling Components If removing or installing circuit boards or components, care needs to be taken to avoid causing damage. The two main dangers are mechanical abuse and static electricity. It is surprisingly easy to misalign pins and sockets, causing bent or broken pins or perhaps even electrical damage due to incorrect connections. In very dry environments, charges can build up on the human body as it rubs against other materials and against the air, and be unable to leak away because everything around is such a good insulator. These charges will spread out to any uncharged object brought near, perhaps jumping as a spark since the voltage can be very high. This may give the human a slight jolt, but is not serious, as the total energy involved is tiny. On the other hand, it can totally destroy electronic components. So done Take care to check alignment of pins and sockets. Do not force things – check and see what’s going wrong. Usually you can feel that things have located properly – there’s a kind of ‘clunk’. If this is missing, be suspicious. Earth yourself before touching any components. This can be by touching any earthed metal, or by wearing an earth strap which is connected to earth. If the components comes in a protective wrapper, earth yourself while holding this, before touching the component. Real Time & Embedded Systems 1.13 CA480 Dublin City University 2007