# How to Read Data Sheets: Rectifier Diodes

Diodes allow current to flow in only one direction. This makes them useful in many applications, including converting AC to DC, regulating voltages, and processing high frequency signals.

In previous installments of this miniseries, we started by considering the problems associated with reading datasheets in general, and then we looked at linear regulators (Part 1 and Part 2). Now it's time to turn our attention to diodes.

What is a diode?

Diodes come in many varieties, and it's not always obvious how to choose the right one. Let's start by answering two questions: "What is a diode?" and "why are there so many types?" Actually, let's ask one more question: "How many diodes are there on the board shown below?" (The answer is given at the end of this column.)

A diode is a device that allows current to flow in one direction but not in the other. This ability makes diodes useful in many applications, including converting AC to DC, regulating voltages, and processing high frequency signals. Early diodes were implemented using a variety of techniques and technologies, from "cat's whiskers" to vacuum tubes (you can find more about the origin of diodes on the Engineering and Technology History Wiki).

These days, of course, we predominantly use semiconductor diodes in our electronic systems. If you're not familiar with diodes, then SparkFun has a very nice tutorial that explains the basics, and a more in-depth discussion is found in these lecture notes.

See also the AC circuits, DC circuits, and diode tutorials on Electronics-Tutorials.ws.

What to look for when selecting a diode

So, how do we know what to look for when selecting a diode? The answer (as in many things electronic) is that this depends on the target application. One typical application where diodes are used is to convert an AC (alternating/bidirectional) voltage into a DC (direct/unidirectional) voltage. This is called "rectification," and diodes intended for this purpose are referred to as "rectifier diodes" or "rectifiers."

Let's look at a simple circuit that you might find in a power supply as illustrated below:

This is the simplest possible rectifier circuit. As you can see in the waveforms below, the diode converts the AC waveform on the input to a pulsing DC waveform. If we were to add a capacitor, we would end up with something close to a steady DC voltage, but that's a discussion for another time.

In this circuit, it's easy to see that all the current in the circuit we want to power (represented in the model by RL) has to pass through the diode. If you look closely at the result of the simulation, you will see that the output voltage is less than the input voltage. This means the diode will dissipate power.

Dissipating power

If you read my recent columns on linear regulators (see Part 1 and Part 2), you should remember that dissipating power caused the regulator to get hot, Fortunately, the voltage drop with a diode is much less, so this shouldn't be a problem. Or should it?

To answer this question, let's look at the data sheet for the 1N4001 diode that appears in the SPICE model (I found the data sheet that we will be using for our discussions on the ON Semiconductor site).

You may want to refer to this data sheet as we proceed. By the way, according to Wikipedia, this component has been around since being introduced by Motorola in 1965, and it's still being made by several manufacturers in the original leaded package with variants available in SMT packages.

When looking at this data sheet, the first thing we see is that seven different part numbers are called out: 1N4001 to 1N4007. The first page doesn't tell us much about the differences between these parts, but when we get to page two we see the following maximum ratings table:

From this table, we can see that each part in the family has a different maximum reverse voltage. Glancing through the rest of the data sheet, no other differences are called out, so this must be the distinguishing feature for the different parts in the family. But why are there three different types of reverse voltage ratings and what do they all mean?

Unfortunately, the meaning of the different ratings is not explained in this data sheet but with a little thought we can figure them out as follows:

• The Peak Repetitive Reverse Voltage or DC Blocking Voltage is the maximum reverse voltage that the part can withstand continuously without damage.
• The Non-Repetitive Peak Reverse Voltage given in this data sheet is what the part can withstand on a "one time" basis (note the conditions given).
• The RMS Reverse Voltage is the maximum AC voltage that the diode can withstand. This only applies to an AC waveform (for the 1N4001 note that 35V RMS is slightly less than 50V peak, so the two are essentially the same).

It's important to pay attention to this when selecting which diode to use. If you attempt to use a 1N4001 to rectify a 115V AC mains voltage, you will be in for a very unpleasant surprise. When a diode is faced with too much reverse voltage, it breaks down and starts conducting current in the opposite direction from what is wished for. While there are diodes that are intended and designed to operate safely in this mode (most notably Zener diodes), many diodes will fail catastrophically and -- by allowing the reverse current to flow -- will often damage other parts as well. Ideally the maximum reverse voltage would be infinite, but these are not ideal diodes, so we need to pay attention.

Just for fun (I know how to have a good time), I tried running my SPICE model with the AC voltage increased to 100V peak instead of 10V. The results are shown below. The -50V peak that the simulation shows would not be good if you were expecting this to be a DC supply. Also note that this simulation doesn't include catastrophic failure.

Moving along, we see that the max average forward current for this series of parts is 1A, with an astounding 30A for a single peak. The instantaneous forward voltage drop is 1.1V, while the maximum average forward voltage drop is 0.8V. We also note that the reverse current is specified in the micro-amps. Ideally, this would be zero, but microamps aren't too bad for a 1A part that was designed in 1965.

On to the graphs. On the graph showing typical forward voltage, we see that the voltage increases with increasing current and decreases with temperature. We also see that the reverse current increases with temperature and that the capacitance decreases with reverse voltage. Wait a minute, capacitance? I don't remember seeing anything about capacitance. That's right, it's only by looking at the chart that we see that these parts have capacitance. For a typical application where these diodes are used in a power supply, this may not be important, but we may need to use caution in other applications.

Now, what did we miss so far? We know the voltage drop and the max current, so we can calculate the power dissipation. Using the average forward drop we get the following:

PD = 0.8 * 1 = 0.8W

This doesn't sound like a lot, but this isn't a very big part. The thermal resistance specification refers us to Note 1. This note, which can be found on Page 4, contains several numbers along with diagrams of possible mounting methods.

It seems that the ability of this part to dissipate heat depends greatly on how it's mounted. The best case scenario is to mount the part with one end in contact with a ground or power plane on the PCB. The worst case option is to mount it away from the board with long leads.

For the best case, this part will have a temperature rise of T = 50 *0.8 = 40ºC. This means that if our board is at 25ºC, then our diode will be at about 65ºC (149 ºF). That's a bit hotter than I'd like it.

In our example application, however, our diode is only conducting for half the cycle, so we're really only dissipating half that power on average. Thus, we have half the temperature rise, which means our diode will be at 45ºC (113 ºF), which is a much more comfortable value.

But what about the worst case scenario when we have a much higher thermal resistance? In that case (using the same assumptions), our diode will be 69.6ºC (157.3 ºF). Although the diode will be well within its thermal specifications, it's still going to be warm enough that we should pay attention so that we don't get burned. So, once again, we find that things are not always as easy as they seem when dealing with power.

How many diodes?

Returning to the circuit board that I showed at the beginning of this article, there are 12 diodes on the board. In addition to eight discrete components, there's also a bridge rectifier containing four diodes (remember I asked how many diodes there were, not how many packages).

And once again, we've run out of time, so we'll have to look at other types of diodes in a future column. In the meantime, as always, I welcome your questions and comments.

## Join the Conversation!

User must log-in to comment.
• Very nice treatment Elizabeth,  you should have been (maybe are?) a teacher...
• Elizabeth

there are 12 diodes on the board. In addition to eight discrete components, there's also a bridge rectifier containing four diodes (remember I asked how many diodes there were, not how many packages).

What about the LED in the photocoupler? LOL

• @Aubrey: "...What about the LED in the photocoupler? LOL..."

Hi Aubrey -- Max here -- good catch -- FYI this isn't Elizabeth's fault -- she sent me the original image to include in the column -- it was while I was doing my copy-edit that I thought it would be fun to add this "how many LEDs" question -- I just knew I'd miss one

I know -- let's pretend it was a trick question LOL

• @Aubrey: "...What about the LED in the photocoupler? LOL..."

Hi Aubrey -- Max here -- good catch -- FYI this isn't Elizabeth's fault -- she sent me the original image to include in the column -- it was while I was doing my copy-edit that I thought it would be fun to add this "how many LEDs" question -- I just knew I'd miss one

I know -- let's pretend it was a trick question LOL

• @David: Thanks for the compliment!

When I was a child I wanted to be a teacher when I grew up. That was before I discovered engineering. I had an opportunity to teach a few years back and discovered that teaching engineering is harder than engineering.

The idea for this series was from a class I taught. After going over the theory, I went over the data sheets for the parts we'd be using in that week's lab.

• @Elizabeth... "teaching engineering is harder than engineering."

Teaching anything is harder than just doing it.  Whoever said "those that can, do, those that cannot, teach" obviously never tried teaching.

I was on a course on PLCs some time ago and was by far the oldest member of the class (including the instructor).  One of the other students was a real larrikin (that's Australian for juvenile delinquent) and was constantly disrupting the class.  There was not much the instructor could do about it (in Australia everyone has rights but no one has responsibilities...). I decided then that I'd never teach.  Fortunately the young man has mellowed a bit as he grew up and now runs his own successful business.

• @Elizabeth... "teaching engineering is harder than engineering."

Teaching anything is harder than just doing it.  Whoever said "those that can, do, those that cannot, teach" obviously never tried teaching.

I was on a course on PLCs some time ago and was by far the oldest member of the class (including the instructor).  One of the other students was a real larrikin (that's Australian for juvenile delinquent) and was constantly disrupting the class.  There was not much the instructor could do about it (in Australia everyone has rights but no one has responsibilities...). I decided then that I'd never teach.  Fortunately the young man has mellowed a bit as he grew up and now runs his own successful business.