Diode Tutorial

Diode? Oh yes, that’s a rectifier. It turns AC into DC.

That’s how it started out in the beginning, with galena and cat’s whiskers and with tubes, then with selenium and a bit of germanium, and finally with silicon. But semiconductor technology moved quickly. Next it was transistors, then small scale integrated circuits, and now the multimillion-transistor ICs that are our microprocessors and other amazing chips. The concepts of digital electronics opened up new avenues of application for this simple device that lets electricity flow in one direction but not back the other way. Terms arrived on the scene like ‘blocking,’ ‘isolation,’ ‘steering,’ and even ‘diode OR circuits.’ Things that just wouldn’t have been practical to do with tubes.

Here’s the symbol used in schematic diagrams for the diode. It’s easy to visualize what happens there. Current can flow in the direction of the arrowhead, but if it tries to go the other way it hits a roadblock. The diode is the one-way street of electronics! The arrowhead side of the diode is called the anode (meaning positive), and the roadblock side is the cathode (negative). Remember these words, as they show up in crossword puzzles.


Though it’s easy to think of the diode being a perfect short circuit in the forward direction, just as it’s easy to think of wires being perfect conductors, reality doesn’t work quite that way. Wires have resistance and incur a voltage drop that’s proportional to current, following Ohm’s Law. But diodes exhibit a constant voltage drop that depends on the material they’re made from, not from the current passing through them. (There will be a negligible amount of Ohm’s Law going on in there of course, but it doesn’t matter in the big picture.) Typical voltage drop values for the common silicon diode are in the range of 0.6 to 0.7 volts. If you don’t have that much to start with, nothing is going to flow. Not so surprisingly, there’s also a limit on how much voltage they can block in the reverse direction, though that’s more a matter of construction, and you have a number of different choices there. More on all that later.

Let’s move quickly to a Forester example of what you can do with diodes that doesn’t involve turning AC into DC. Then we’ll get back to more general discussions. Here’s the diagram of my map lights with dome light mod, as shown in this post in the long interior lighting thread. The object of the mod is to turn both map lights on when the dome light comes on, but to keep the map lights themselves working the same way they used to work, with each map light’s switch controlling that light and only that light. And not to have any destructive side effects that reach out elsewhere in the Forester, which is a classic risk associated with all kinds of mods—not just electrical ones!


Well, you could get the map lights to come on with the dome light simply by wiring all three in parallel. That would satisfy the first objective of the mod, but it would certainly fail at the second. Turning on either map light would turn on all three, and that’s not what we wanted to happen. As for potential harm to the health of the Forester, the map light switch would ground the output of the dome light controller. We just don’t know what would happen if a door were opened and the controller tried to put the dome light through its dimming process while its output was shorted to ground. And we don’t want to find out the hard way. Those sealed modules are expensive.

So we inject a couple of diodes into the circuit to isolate the dome light and its controller from the effects of the map lights, and the map lights from each other. Here’s how it works.

According to the official wiring diagram on which my drawing is based, the map lights and the dome light are controlled by ‘low side switching.’ In other words, they’re always connected to the battery, but they don’t complete the circuit to ground till their switches are closed. This is not uncommon in automobile wiring (though it’s against Code in home and other AC wiring).

[Actually the map lights were miswired in most if not all SG Foresters, and perhaps others as well. The switches were located between the battery and the lights, and the other side of the lights was always grounded. This ‘high side switching’ worked fine in the simple map light circuitry, but that wiring wouldn’t be right for the mod due to the way the dome light controller works. They all have to go the same way. So I first wired mine back the way they were supposed to be. Never trust the diagram. Always test first!]

Now, back to my diagram. When the dome light is grounded, either by its switch or by the ‘integrated module’ (dome light controller I call it) that connects to the door switches, this also creates a path to ground for each of the map lights through their respective diodes. Electricity flows in the direction of the arrowheads, and all three lights come on. But when either of the map lights is grounded through its own switch, electricity from the dome light, its controller, or the other map light can’t get to that ground because it’s blocked by the turned-on map light’s diode.

That’s blocking and isolation. Depends on how you look at it.

Isolation is especially important in mods, as the complexity of automobile wiring, together with some of the peculiar shortcuts that designers often take and the likelihood that the mod could cause multiple circuits to become intermixed, can lead to the surprise discovery of ‘sneak paths’—unintended routes that electricity will follow that produce unexpected results. See my discussion of what happens when a headlight fuse blows for an example. As an exercise for yourself, think about how diodes could prevent the situation that is illustrated there. In reality, however, you probably wouldn’t want to wire it that way as the diodes would eat up power, produce heat, and cost money, and the sneak path situation isn’t really dangerous or harmful. It’s just very peculiar and confusing.

Here’s another simple example of blocking. You have a piece of electronics that would be damaged if it were plugged in backwards. One simple solution is to include a diode at the place where it gets its electricity from the outside world, aiming in the proper direction. This way, current can only flow the right way, and it will be blocked if the device were plugged in backwards. But do remember that the diode will consume about 2/3 of a volt. This probably won’t matter in most applications.

Now for steering. At the expense of three more diodes (and one more voltage drop), you can make it so that the device will work when plugged in either way, at least when viewed from the outside. Here’s a picture of how to do this.


The device is over on the right side, and it requires that its top terminal B be connected to positive and the bottom one C to ground. Plug it in wrong and it’s dead. So the job of the four diodes is to steer electricity to B and then back out from C, regardless of which way the entire circuit is connected to the outside power supply at A and D.

Consider first the situation where A is connected to positive and D to ground. Follow the red arrows. Current comes in at A and is faced with two possible routes. But due to the orientation of the diodes there, it can only pass through D1 at the top. The diagonal diode D3 is facing the wrong way, so current can’t go there. And once it gets through D1 it encounters a similar situation that forbids it from heading straight down the D2 path to ground. So it continues to B, which is where it’s supposed to end up.

From B it goes through the device, coming out at C. Again, two choices: through the diode D4 at the bottom, or back up to A through the diagonal diode D3. D3 is now facing the right way, but there’s this little problem that A is more positive than C, and that’s just not the way electricity flows. It doesn’t want to go back where it came from. A similar situation prevents flow back up through D2. So now we have a nice complete circuit: A – D1 – B – Device – C – D4 – D.

Now switch the power supply around so that D is positive and A is ground, and follow the blue arrows. We’ll take it a little faster this time, since the principles are all the same as before. Current comes in at D and can only go up the diagonal D2. That takes it to B, where it’s supposed to be. It goes through the device to C, from which it goes up through the other diagonal D3 to A, which is ground. All other routes are blocked, either by diodes facing the wrong way or by destinations being too positive.

That’s steering. Remember this circuit, because you’re going to see it again soon.

Now let’s generalize a bit. In the map light mod, the light came on if its switch was on, or if the dome light was on, or if both were on. In the terminology of digital electronics, that’s the OR function. Turn something on when any one or more different other things are on, but don’t let any of those things interfere with the others. So let’s look at a simplified version of what we did with the map lights and think about what it could do for us in general. Here’s a three-banger. You want electricity to show up at point A when one or more of three other things X, Y, or Z get turned on. But you don’t want any of those things to turn each other on or to get inside each other’s circuitry and mess things up there. Isolation like this is especially important in a mod, because car wiring is complicated and things are related to other things in sometimes very strange ways that don’t always show up on the same page of the wiring diagram. So, three diodes and you’ve got it. Isolation, blocking, or in a logic statement:


This circuit presumes that X, Y, and Z all employ high side switching, and that you want A to work the same way. You could make it work if all four used low side switching by turning the diodes around in the other direction. That’s closer to what’s happening in the dome light illustration.

If you’d like to pursue an advanced curriculum in diode-based logic, check out this description of ‘Mickey Mouse Logic!’

So far we’ve pretty much confined our discussion to DC, but now for completeness let’s move on to the other thing that diodes do—turning AC into DC, which was their original job in life back in the beginning. We edged our way up to it in our discussion of blocking and steering. There we spoke about how to design a circuit which would ensure that backwards current wouldn’t flow into a circuit (blocking), and another one which would ensure that current would always end up flowing the right way regardless of how the circuit was plugged in (steering). But we were thinking about DC at the time. How about if we drive these circuits with AC? What’s the difference! AC is just DC that changes back and forth really fast and changes its voltage up and down as well.

Yes, each of those circuits is a standard way of creating a rectifier, turning AC into DC. A single diode constitutes a half wave rectifier, chopping off the negative half of the AC sine wave and keeping the positive half. The four-diode steering circuit is a full wave rectifier. It not only keeps the positive half of the AC, but it flips the negative half upside down and fills in the spaces that would have been left empty by the half wave approach. The result of full wave rectification contains twice the energy of half wave (less that associated with one extra voltage drop), and it’s easier to filter out the ups and downs to create a constant DC voltage than it is with half wave. It’s an extremely common application.

Diodes are generally sorted into three categories: small signal diodes, general purpose diodes, and power diodes, and these categories overlap by quite a bit. They aren’t formal definitions. The major differences involve current handling capacity, forward voltage drop, construction, and heat.

Small signal diodes are of light duty construction. They are able to handle the small amount of current necessary to drive other elements of circuitry which require very little power, and some are made from germanium, which offers a much lower forward voltage drop than silicon in exchange for a lower current rating and/or a higher cost. Typical small signal diodes include the germanium 1N34 and the silicon 1N914. (Although the 1N914 is called a small signal diode, its specs are actually pretty well into the general purpose category. It can handle 75 milliamps and block 75 volts, so there are indeed useful things that can be done with it outside of small signal applications.) These diodes are typically constructed with a wire coming out each end. The cathode end (negative end, blocking end) is marked with a stripe. Think of it as the blocking line in the schematic diagram.

Power diodes are intended for use as rectifiers in high current applications. They are typically constructed so that they can bolt directly onto a heat sink, because they do get hot. Remember, Watts equals Volts times Amps, and there’s a forward voltage drop of about 2/3 of a volt across every silicon diode. So for every Amp you push through one of them, you are losing about 2/3 of a Watt of power inside it. Where does it go? Heat! And what does heat do? It destroys semiconductors. So you need a method of taking heat away: a heat sink.

Note also that power diodes often use their body or attachment point as the cathode terminal, so you can’t just go bolting these things down to a convenient piece of already grounded metal and expect them to work unless that’s how the circuit is supposed to be. (See the picture below. Two of the power diodes there are indeed connected straight to ground because that’s how the circuits go. The other two are insulated.) Consult the manufacturer’s data sheet to determine the terminal configuration for your diode. Insulators and thermally conductive grease are available for use in these applications. If it’s a full wave rectifier that you have in mind, you can buy these already fabricated and ready to wire rather than dealing with four individual diodes. They’re often called bridges.

In between are the general purpose diodes. The standard family here is the 1N400x, where x runs from 1 through 7 and indicates the maximum reverse voltage that the diode can tolerate. The 1N4001 starts the family out at 50 Volts, running up to 1000 Volts for the 1N4007, so any of them would work fine in a car. All can handle a current of up to an Amp. For currents up to 3 Amps there’s the 1N540x family, just a bit more expensive. It also runs from 50 to 1000 volts reverse voltage. All of these are constructed and marked similarly to the small signal diodes though a bit larger.

These three categories are very loosely defined. There’s a lot of overlap. If you’re just doing things with cars, you can pretty much stick with the general purpose ones. They typically cost under a nickel apiece in quantity on eBay. Five dollars will buy a lifetime supply. If you need to handle higher currents, you’ll have to do some research and perhaps think about heat sinking. Or ask yourself if there’s another approach to the problem.

There is one caution as you move into higher current applications. Unlike many types of components where you can create greater values by putting things in series or in parallel, don’t try that with diodes. While a series connection would be pretty much meaningless, one could be tempted to put two in parallel for greater current handling ability. That’s not as easy as it seems. Due to manufacturing tolerances, there could be a minor difference in the forward voltage drops of the two, which could lead to an unpleasant conflict between them.

Here’s a picture of all three kinds.


One further note, in response to a question that was raised in another thread. LEDs are diodes, so can we rely on them to provide their own appropriate blocking or isolation? In theory, yes you could. They really are diodes, and they exhibit diode behavior. But there are a few drawbacks.

• LEDs don’t tend to feature an extremely high reverse voltage specification, and often you won’t even find that specification published anywhere—especially with a LED array rather than a single bare LED. So you could be taking a chance if the reverse voltage in your application got too high.
• Some LEDs, especially white ones, have a rather high forward voltage drop up in the order of 3 volts, which could be inappropriate in some applications. And LED arrays usually have, by design, a very high forward voltage drop—about 12 volts!
• Some LED arrays have internal steering diodes so that they will work in either direction! They don’t behave like diodes.

Summary:

• Diodes allow current to flow in the forward direction, but they exact a price of about two thirds of a volt for doing that.
• The forward voltage drop times the number of Amps flowing through a diode equals the number of Watts of heat that will be produced in a diode. Large amounts require heat sinks.
• Diodes can block current from flowing the wrong way.
• Diodes can isolate circuits from each other, preventing unintended flow in the backward direction.
• Through a combination of these capabilities, diodes can steer electricity of either polarity, or even AC, to places where only one polarity would be appropriate.
• Diodes can rectify AC, turning it into DC.
• The cathode end of small signal and general purpose diodes is marked with a stripe. Consult the manufacturer’s data sheet for terminal configuration of power diodes.

I’ve also posted a tutorial about multimeters, one about different applications of relays, and one about how to work with LEDs. Additional suggestions?
©