Chapter 2 – Diodes
NOTICE: As a courtesy, you are viewing a preview version of an upcoming book, Transistors and Semiconductors: Clearly, Simply, Intuitively, and Concisely Explained by Michael LoneWolf. Since this is just a preview, the text below has been stripped of formatting and images. I need YOUR help to have this book published. Please HELP by visiting my Funding Request page at the main website. DISCLAIMER: This is an obsolete (draft) version of book manuscript. It is not intended to be used for study or reference, but for book preview purposes only. Support the project so that you can purchase a print copy of the most up-to-date book version with all formatting and illustrations!
We are going to begin our study of semiconductors with a look at diodes.
Fig 2.1
A diode is a two-terminal (there are two connections available to the reader) semiconductor device. The standard symbol for a diode is as illustrated, where the direction of the arrow indicates the fact that polarity matters. The cathode end of a diode is marked on the device. On a thru-hole part (as opposed to SMD, “surface mount device”), it is marked with a band.
If we connect the diode in a circuit with positive polarity connected to the anode, we will see that current will flow in this direction.
Fig 2.2
This will be our first circuit we will build. The “battery” can be anything you have that will supply 12V, or any other voltage. Everyone has plenty of wall-wart power supplies that no one knows what they belong to. If you use a different voltage, then select another light bulb.
The diode used is not critical. It can be anything that your local RadioShack stocks. It should have a “forward current” rating of at least 1A.
The bulb should have a voltage rating no less than that of the power supply. It would be easier to have it mounted in RS 272-357.
If we connect the diode with a positive polarity connected to the cathode, we will see that no current flows in this direction – bulb does not come on.
Real life diodes have several parameters which must be satisfied:
Reverse Voltage, or PIV (peak inverse voltage) – maximum voltage which can be connected across the diode
Forward Current – maximum current which can flow thru the diode
Let us look into diode behavior in more detail. We want to see their precise in-circuit behavior, so we will connect an adjustable (or variable) power supply across the diode, and measure the current which flows thru the device. We then plot the current vs. voltage relationship. This technique is always used in electronics to understand device behavior.
Fig 2.3
Fig 2.4
The device in a circle with an A is an ammeter. We can omit a load resistor if we keep currents below device limits, and keep supply voltage to very specific limits as well.
If we graph the voltage-current relationship, there are several interesting points:
1. In forward direction, the device does not pass current at a very low voltage across the device. Little current flows at tenths of a volt. Current starts to increase very fast (exponentially) at about 0.6V.
This is diode behavior in “forward conductance” or “forward bias”. After about 0.6V, it will pass a large amount of current thru itself. Coincidentally, as we will see later, this same 0.6V is the amount which gets dropped across the diode, or “stolen” from the voltage applied to it. If there was a load attached it would see a voltage of (Vsupply–0.6V).
2. In the reverse direction, very little current passes thru the device (50μA for 1N4001) as more and more voltage is applied. A diode will block current flow in reverse bias. At a high reverse voltage, 50V for a 1N4001, current thru the device abruptly increases. This is called “breakdown” region, and a diode must not be operated at this reverse voltage level. Operation at this region will permanently damage the device. This region is sometimes called by the very descriptive term “avalanche breakdown”, for the sudden and great surge of current as the device gives up and destroys itself.
The diode, therefore, conducts current in only one direction, if voltage across it is >0.6V, and less than breakdown.
I have split the two plots for the reader's convenience. Most books put both of the plots on one combined plot. When they do so, you have to watch out for axes scale change: +x axis is in volts, +y axis is usually in amps or milliamps, -x axis is in tens of volts, and -y axis is in μA. You will get to see such piling of lots of data on a single plot very often in electronics:
Fig 2.5
Example
Calculate all voltages and currents:
Fig 2.6
Answer: With the diode connected in the forward direction, it conducts current. Voltage VD across the diode will be 0.6V. Voltage VR across the resistor load will be (3–0.6V). Current IR thru the load resistor will be about 2.4mA.
In the reverse direction, diode will not conduct any current. IR will be close to zero. VD and VR will both be 3V, but no current will flow.
2.1 Diode bias
Diode with positive voltage applied to the anode is said to be in “forward bias” mode of operation. The fancy term “bias” will always refer to the manner in which a diode is connected. It is also applied to other electronics devices. For example, if someone would say to you “oh, that device is forward-biased”, you would interpret that as the device being connected in circuit in a manner which would permit current flow thru the device, and that a necessary minimum 0.6V has been applied to the device. Or, someone else might say “to place this device in forward bias, such and such voltage is required”, and you would translate that as the device being already placed in circuit properly, but that the task remains to see what conditions would ensure forward current flow thru the device. This mumbo-jumbo is used so that people “in the know” appear to speak a special language which no one else could possibly understand.
If, on the other hand, someone says “oh, it's reverse-biased”, then you would translate that as being a condition of voltages applied which place that particular device into the reverse operating condition, where no current flows. Voltages in electronic circuits change; an adjustment of a part like a potentiometer, or circuit response from an external signal may easily change that “reverse bias” condition to a “forward bias”. It is your job to know which conditions may cause this when you analyze diode circuits.
Fig 2.7
The “forward bias” voltage (see, we are using mumbo-jumbo ourselves already!) is something that must be kept in mind. A diode in forward conduction will drop about 0.6V across itself, as in a circuit below:
Due to Ohm's Law, this voltage drop (0.6V) multiplied by the current IL thru the diode will result in heat generated by the diode. Therefore, this is another parameter of a real life diode which has to be factored in device selection.
2.2 Physical diode limitations and imperfections
If you are choosing a diode, a few diode parameters are very important to know:
Maximum reverse voltage across the diode. This is also sometimes stated as PIV (peak inverse voltage), or breakdown voltage. Usually, it is written as VRmax
Maximum forward current thru the diode, IFmax.
Power dissipation ability of the diode and maximum junction temperature. All semiconductor devices are sensitive to temperature, and will have their parameters change with temperature increase. All semiconductors also have maximum device (or junction) operating temperature specifications.
Equivalent internal resistance. When looking at the I-V plot, we see that a limited amount of current flows at a limited forward voltage across the device. Current does not rise instantly as high as it can go after the 0.7V are exceeded, as would be the case for an “ideal” diode. This non-vertical slope signifies an “internal resistance” (limited current flows at any particular voltage; more current flows with higher voltage potential across device – sounds like Ohm's Law!). This is not an actual resistor inside the device; it is just what the device acts like. Furthermore, we see that the equivalent resistance (slope) changes with changing operating conditions. This non-ideal behavior will result in internal I*V losses and heat dissipation.
2.3 Reverse, leakage current
There is a very serious and undesirable characteristic of semiconductors – reverse leakage current. This current is not much of a problem at room temperature. However, it goes up rapidly with increasing temperature. For every 10°C increase, the current goes up 2.5 times! Device operation at 80°C is very different from one at room temperature.
This might not seem to be a big issue during our experimentation at 1A current levels. However, small-current or precision operation will be affected.
Furthermore, every device covered in this book will be related to diodes. We will see plenty of cases later when this increasing leakage current will be a deal-breaker.
2.4 Datasheets
In electronics, a device you purchase has an accompanying datasheet, which describes its expected operation. You should obtain and save datasheets of all electronic devices you use. They can be obtained either from the seller such as Digi-Key, or from a website such as www.AllDatasheet.com. Datasheets for all devices described in this book can be obtained as a single compressed file from the book's accompanying website, www.MKRD.info/SemiconductorsBook. Let us look at a real diode datasheet, for the 1N4001 diode manufactured by Vishay. You will note that AllDatasheet lists many manufacturers for a 1N4001, and that the datasheet I have chosen for you lists General Semiconductor. This is very common in the electronics industry. 1N400x series is a “commodity” item, and just about anyone makes one. In addition, its datasheet has been written a long time ago, by its original manufacturer, General Semiconductor. That company has since been swallowed by Vishay. You will see that most datasheets essentially copy each other, while one or another has more information than others. This is quite common in electronics.
Another important point is the use of different subscripts by different manufacturers. Don't expect all of the subscripts which you encounter in the book to be adapted by every datasheet you encounter. You will see slightly different subscripts, nomenclature, and test conditions in other books or datasheets.
In the datasheet, several more interesting specifications can be seen:
VRRM, a term they use instead of VFmax. Using non-standard nomenclature is very common in electronics. "Repetitive" here refers to marketing department's love for stating unrealistically high conditions which the device can withstand, for a very brief amount of time. We are more interested in prolonged operation, so we use "repetitive" ratings.
VRMS, maximum AC (alternating current) voltage.
IRmax of about 40μA at 70°C.
Fig 1 of the datasheet, which shows that device maximum forward current rating will drop after a junction temperature of 75°C.
Notice that most specifications are given at a specific temperature, or at several temperatures, or in a plot which plots the characteristic against junction temperature. All semiconductor devices are sensitive to temperature.
2.5 Checking a diode with a DMM
A good DMM has a diode check function. The switch position is labeled with the diode symbol. The OL on the display stands for “overload”. A more intelligent idea for the manufacturer would have been to display OPEN, which is what it means. If the read lead is connected to the anode and the black to the cathode, the diode should start conducting, and the DMM will display a reading of the voltage drop across the diode. Reversing the polarity will again show OL.
If what you have is an analog multimeter, then the X1000 or X10,000 resistance check range can be used to check a diode, in both directions.
Special Insert – Thermal considerations
As the reader can already guess, temperature is a factor which must always be kept in mind when using semiconductors. In general, with increasing temperature, semiconductors increase the amount of current they conduct (with other circuit parameters not having changed). At higher temperatures, lifespan of semiconductor devices decreases as well.
However, the more important point to keep in mind are the two different temperatures, one being ambient (surrounding air) temperature, and the other being internal to the device. All semiconductors generate heat due to P=I*V. Adapting the equation to a forward-biased diode, PD=IF*VF, where PD is the amount of power dissipated, IF is the forward current thru the device, and VF is the forward voltage drop thru the device.
Why “dissipated” instead of “generated”? Because unless the semiconductor device is to become a little oven with an ever-increasing internal temperature, it must dissipate the internal heat to the surrounding area, be it to the PCB, a heatsink, or the surrounding (ambient) air.
The measure of how well a device allows heat transfer from its internals to the ambient surroundings is Θ, thermal resistance, where Θ is the Greek letter theta.
Any time when not all of the heat generated within the device is transferred to the surroundings, device temperature rises. The temperature rise can be calculated from
Δ Temp rise = ΘJA * Power Generated
where Δ temperature rise is relative to the initial ambient condition (for an unpowered device at room temperature, 20-25°C, for a warm-to-hot semiconductor, about 60°C). JA stands for “from junction to ambient”. Junction means the internal semiconductor material of an electronic device, which is housed in a plastic case that you see.
Any semiconductor device lists in its datasheet the maximum temperature it can handle. But for all Silicon-based semiconductors (the majority), the maximum temperature must be maintained below 110°C. This is above the water boiling point, and a device this hot will burn anyone who touches it, among other things. A better idea is to keep device maximum temperature to below 80°C.
The equation above is typically used in reverse. A maximum temperature is chosen for the entire circuit, 60 or 80°C, ΘJA is looked up from the datasheet, and a maximum power dissipation (generation) value is calculated for the device. Then, device operation in the circuit is examined to see if operating conditions exceed that power generation limit. Or, one of circuit parameters is held constant (diode forward voltage drop of 0.65V for example), and a calculation is made of how big the other parameter can rise (forward current passing thru the device).
Fig 2.8 TO-220 tab
Low-power semiconductor device are encased in plastic. These devices can dissipate a low amount of heat. How can more powerful devices be made? At higher power, we give up on trying to dissipate all of the heat thru a plastic body, and we connect the semiconductor internals inside to a small metal tab. That metal tab is designed to be bolted to a heatsink.
A heatsink has a large surface area, usually achieved by using fins. It can therefore transfer a large amount of heat to the surrounding air.
There are a few factors to be aware of when using heatsinks.
1. Most semiconductors have an internal connection from one of the terminals to the tab. If two or more devices are mounted on a heatsink, then those terminals will be electrically connected to each other thru the conductive metal heatsink. If those terminals must not be connected together in the circuit, either a device with an insulated (or isolated) tab is used, or electrical isolation from the heatsink to the tab is created with a thin sheet of mica or a “thermal pad”.
2. When an isolator is not used, thermal grease is applied between the device and the heatsink. The grease fills tiny surface imperfections on both sides, and leads to a good thermal contact between the two surfaces.
Fig 2.9
3. Each of the “layers” in the Figure above represents a different and separate thermal resistance. The total Θ must now be calculated, by adding up thermal resistances of separate entities which are connected together.
ΘJA = ΘJC + ΘCS + ΘSA
C stands for case, and S is for heatsink. Typical ΘCS is 0.5 for thermal grease, mica, or a thermal pad (a grease-less silicone insulator).
The resulting junction temperature can be then calculated:
TJ = PD*ΘJA + TA
A realistic ambient temperature TA is at least 40°C, not room temperature, because heat generating devices on a board are often together in a cramped space with poor airflow.
The junction temperature TJ result must not exceed a chosen temperature limit, perhaps 80°C. Alternatively, the calculation can be run the other way, given a known dissipation amount. The minimum ΘSA is then calculated, and a heatsink is chosen with an equal or lesser specification. Note that the specification is given for a vertical position with unrestricted cool air flow. If heatsink position is less than optimal, or a very low ΘSA is required, then a forced air flow may be needed.
Very little thermal grease is applied, just enough to fill surface imperfections, but enough to cover the whole area. It should be apparent that the thicker the grease layer, the bigger the separation from the tab to the heatsink, and the worse the heat transfer. Note also that since the device is tightly bolted to the heatsink, any excessive heatsink is squeezed out under pressure, and there better not be too much grease to be squeezed out!
Fig 2.10
Notice the thin layer of thermal grease. Notice also that the device was kept in its ESD foam, and is on an ESD tabletop. We will study ESD a little bit later.
Heatsinks are bulky, heavy, and expensive. Also, it frequently happens that when the calculation is done with a given amount of power generated within the device, the ΘSA required is a very low value. A low ΘSA means a large heatsink size.
Fans can be used to force cool ambient air to quickly replace a warmed-up air right next to the heatsink, and increase heatsink power dissipation capability. In fact, most heatsink datasheets list the ΘSA parameter at several values of ambient air flow points.
Note, however, that fans, being mechanical devices, have many critical limitations:
Noise
Additional power draw
Requirement for filtering
End users never replace the filters, severely degrading air flow. System designer must never rely on an airflow figure given with a clean filter!
Poor reliability. All fans die eventually, much sooner than the solid-state electronic device they are cooling.
Dust deposit on the PCB and the heatsink, which leads to poorer heatsink performance and possible circuit degradation.
Fig 2.11
A good small-size heatsink for TO-220 packages available from Digi-Key for hobby and experimentation uses is the FA-T220-64E from Ohmite, at $2.01. Its datasheet is included with this book, and the reader is encouraged to look at it.
Device parameter vs. ambient temperature plot
Fig 2.12 All datasheets contain an interesting plot somewhere near the end of the datasheet. The plot shows an important device parameter, such as forward current flow, vs. an ambient temperature.
Several things are going on here.
1. The horizontal area represents maximum current thru the device limitation, independent of ambient conditions (due to size of internal wires and other factors).
2. The diagonal area decreasing the total power dissipation due to the rising product of IF*VF.
3. A vertical area specifying a maximum device temperature, no matter how low the electrical operating parameters values.
Note that the x-axis is erroneously and confusingly labeled with Ta (ambient). It is more intuitive to label the axis with TDEVICE, with TDEV = Tamb + ΔTrise. Then we can easily analyze this plot. At a given initial ambient temperature, a limited amount of temperature rise due to internal dissipation is allowed before the TDEV limit of 100°C is reached. Equation shown before can be used to determine maximum device electrical operating parameters at any combination of ambient and internal device temperatures.
This plot is most often called a power derating plot, obviously signifying that high ambient or internal temperatures derate or decrease the amount of power (I*V) which the device can handle.
Example: Calculate maximum forward current thru 1N4001 which will keep its junction temperature below 75°C.
Answer: The datasheet lists ΘJA=50°C/W.
TJ = PD*ΘJA + TA
W
PD=IF*VF A
Since 1.2A exceeds device maximum of 1A, we must go with 1A. However, our results show that 1A which the device states in fact will not significantly raise device temperature, and a forward current of 1A can be passed indefinitely thru the device.
2.7 Variability of parameters
Forward voltage drop, of 0.6-0.7V for a common diode, is a ballpark figure, and not an exact number. Diodes will show manufacturing spreads of several parameters, this being one of them. This voltage for different diodes will change, and can go as low as 0.4V or as high as 0.8V. In addition, it can be different for completely different diode types, which we will cover later. But its typical value is of 0.6 to 0.7V. In this book, I have purposefully randomly mixed specifications of 0.6, 0.65, or 0.7V, to keep a reader alert and on his feet, and for him to remember about variability of parameters.
Semiconductors are temperature-sensitive devices. Device temperature will vary due to both the internal power dissipation of I*R, as well as the ambient temperature. The diode will show variation in several parameters as the temperature at which it is operating is changed from typical (room) operating temperature, and up to the maximum temperature. In particular, the following parameters will change:
Reverse leakage current increases with increasing temperature
Maximum forward current decreases with increasing temperature
Breakdown voltage may decrease with increasing temperature
You need to consult the datasheet regarding all device parameters and operating conditions.
In addition, an effect as bad as the avalanche breakdown is “thermal runaway”. Common to different semiconductor devices, it is due to the fact that electronic devices usually start to allow more current thru themselves as they warm up. In the Fig.1 of 1N4001 datasheet, a nice diagonal line is drawn from 1A, 75°C point to 175°C point of higher temperature. This is a plot not of device limitation, but of limitations which you, the circuit designer, must place on your circuit. The nice diagonal line is a thermally safe region of operation, where the combination of forward current and device temperature will not damage the device. Area under the line is safe; but points of operation above the line are not allowed. The device can be improperly operated at 1A and temperatures higher than 75°C, ignoring this plot. At these operating conditions, heat generated by the device will not be completely dissipated by the device package, and junction temperature will rise. Increasing junction temperature and prolonged operation will lead to further temperature increases. Increasing temperature will also lead to higher current thru the device, leading to increasing temperature, and so on, in a spiral to certain destruction.
Therefore, a diode must be selected for a circuit with a safe margin of operation, or circuit parameters must be limited to safe values.
2.8 Special diodes
Diodes are chosen based on their intended application. We have so far discussed a diode commonly referred to as a “rectifier diode”. How exactly they do this task will be illustrated in the next section, but for now we will just say that a rectifier diode is made to have a relatively high forward current rating, high reverse breakdown voltage, and large physical size to handle 0.6VF*IF thermal dissipation.
A separate class of diodes is referred to as “signal”, “switching”, “high speed” or “fast recovery” diodes. These are made to respond to circuit changes quickly (thru lower device capacitance), and large forward current is of second concern. One such diode type is a Schottky diode. Its first notable feature is a typical forward voltage drop of only 0.2-0.3V. The second feature is its extremely fast response speed.
An older type of diode, of Germanium construction, no longer being made, had a forward voltage drop of 0.1-0.2V and fast response to signals. It suffered from high leakage current.
2.9 Diode applications
Rectifiers
So, why is a common diode called a “rectifier” diode? Because they can be made to rectify, or convert, AC to DC current.
Fig 2.13
What we see is a circuit which is found in every wall-wart power supply which the reader has lying around. High AC voltage from the outlet is applied to a transformer, which converts it to a lower voltage AC current, as well as isolates the output from the AC line (read the AC Voltage Safety section for more details on electrical isolation). I have plotted the AC output of the transformer secondary winding. That is connected to a diode bridge, commonly represented in the circuit as a square sitting on a vertex, with a diode symbol inside. This symbol stands for a special arrangement of four diodes, which the reader can see on the right. The output of the diode bridge, VD, is also shown. It consists of pulses of current which are of positive polarity only, as opposed to AC current, which switches from positive voltage to negative, in the form of a sine wave. See if you can understand how the arrangement of four diodes accomplishes this task.
Notice also from the bridge output plot, as compared to its AC input plot, that the peak of every pulse is of a lower voltage than the AC input. Question for the reader – Why? Answer: keep in mind the two diode drops, because at any point in time two diodes in series are conducting current.
The output, while only containing positive voltage, is a pulsating output, and is not DC. It would be OK to power a light bulb, but if it was used as a power supply for an audio device, the user would hear a very loud and annoying “AC buzz” of 120Hz. That is why a filtering capacitor CF is used. Each voltage peak charges it up, and the capacitor gives back its stored energy when DC pulses drop in voltage, 120 times a second. It can only store a limited amount of energy – the output after the capacitor is plotted as VC. It is an output more closely resembling a DC current.
The four diodes are actually sold as a pre-built “bridge” module, which is often coincidentally square, and has four terminals, so that the reader does not have to interconnect four diodes himself.
Flyback diode for driving inductive loads
Inductive loads are nasty loads because an inductor by definition resists any voltage level changes. That means it will oppose you trying to increase voltage across it, as well as trying to decrease it. If power supply to an inductor is suddenly disconnected, the inductor will attempt to maintain that current flow, by converting its stored magnetic field into current. Resulting voltage level, since that inductor is not loaded, can reach very high levels. A 12V relay briefly producing 100V-1kV is typical. More voltage is produced by bigger inductors.
That voltage will be applied to the circuit according to the circuit below. Polarities specified are correct; by “opposing”, an inductor produces a voltage opposite to what you are trying to apply to it.
Fig 2.14
After we have the current flowing steady-state in a counter-clockwise direction, but cease it quickly, the inductor will oppose the change. It will collapse its magnetic field, converting it to electrical current, in the direction the current was flowing previously.
For moderate sized inductors, a voltage in the range of kilo-volts can be produced, because there is nothing loading that voltage. When the switch is thrown open, in circuit on the right , voltage between the contacts can jump to several kilovolts, break down dielectric property of the air between closely spaced switch contacts, and produce a visible spark (or arc). This spark, besides being dangerous to people which may operate the switch, also wears out and destroys switch contacts.
We can use a flyback diode to clamp those negative voltage spikes:
Fig 2.15
When that switch is closed, inductor will store electrical energy in its magnetic field, while the diode will be in reverse bias and will not affect the circuit. When the switch is later opened, that inductor will produce a voltage spike across itself. That voltage spike will be applied across the diode, which will open and conduct, and dissipate the energy inside itself. The switch will not see a voltage spike and no arcing and sparks will be generated.
As for part numbers, that diode must: a) respond quickly b) withstand high voltage levels c) be able to dissipate all of that energy thru itself, and d) have a current rating equal to that of the normal current flow thru the inductor.
For our trivial loads and power supply voltages, a 1N4003 is sufficient. For driving larger loads or at higher power supply voltage, a Schottky diode (a fast diode) with appropriate breakdown voltages and surge current rating must be used.
From now on, any load which can behave as an inductor (a winding of electrical wire, and / or having inductance), such as relays, motors, etc must always have a flyback diode attached.
Diodes as clippers and clamps
Power OR / backfeed prevention diode
Reverse polarity protection
Electronic circuits do not tolerate a reversal of power supply polarity well. This is easy to do in real life conditions; for example, wall-warts are lying around in everyone's house which put + on the outside (shield) connector. There are also AC wall-warts. Also, a 9V battery can be easily pressed against the connecting clip in the wrong orientation, and so on. The easiest way to protect against polarity reversal is by passing the + power supply thru a diode. The problem with this approach is that a high-current diode is needed (large in PCB estate and expensive), that the diode with drop the power supply by 0.6V (0.2V if a Schottky is used), and that it will dissipate I*R losses.
A cheap trick is to use a diode in reverse, in combination with a fuse. The circuit will probably need a fuse without the reverse polarity protection anyways, and the diode in reverse has little impact on the circuit with correct applied polarity. If the polarity is reversed, that diode will conduct and blow out the fuse.
Fig 2.16
Other diode applications
2.10 Diodes for voltage control
Suppose that you needed a regulated source of 1.2V. By “regulated”, we mean a device which functionally does the following:
Fig 2.17
Therefore, it accepts any voltages from 1.2V (below which we cannot expect it to be able to output 1.2V) to high voltages, while only outputting 1.2V. The load may request zero or higher current, and output voltage cannot change with the load current changes. It is easy to see why a resistor divider or a potentiometer cannot satisfy these requirements.
We can use a trick like the following:
Fig 2.18
Two diodes in series will have a forward voltage drop of 1.2V. But one of the first circuit limitations is that these diodes would have to be hand-picked for their 0.6V forward voltage drops, as this is a parameter which somewhat varies from diode to diode. Those diodes will draw current thru the current limiting resistor, which will drop (Vin–1.2V) across itself. By “regulating” this circuit actually “drops it down to” a low voltage. Limitations of this circuit are very apparent:
Voltage higher than 1.2V will require more diodes wired in series
Current limiting resistor will dissipate a lot of power (Task for the reader: calculate power rating of a resistor and circuit energy efficiency at Vin=12V, Iout=1A).
Maximum current to the load is limited by the resistor. Using a higher resistance to decrease circuit power dissipation will limit maximum current which it can supply.
The circuit must ensure that a minimal value of diode current flows thru the diodes to “turn them on”. If the resistor does not supply enough current, a very high resistance load will see an unregulated voltage of Vin at the “regulated” output.
Diodes will dissipate a low of power.
We can do something about only one of these limitations. (In a later chapter, we will deal with power dissipation as well). Forward voltage drop is a very low voltage phenomenon for all diode types available, but there is a region of the plot which roughly resembles the characteristic forward voltage region. Task for the reader: find it in the diode I-V plot:
Fig 2.19
The “breakdown” region of the plot resembles the forward conduction plot, but occurs at a much higher relative voltage. However, the reader has been warned not to operate the device there. Well, there is a special diode type device, called a Zener diode, which has been made specifically to be safe to operate at “reverse breakdown”. Furthermore, it has been manufactured with very specific, and relatively low, “reverse breakdown” points (couple of volts to tens of volts). Operating in reverse breakdown, a Zener works much like our diode “regulator” circuit, but at higher voltages. In forward conduction, a Zener is like a normal diode.
Fig 2.20
We use the same circuit, but the diode is connected in reverse.
The Zener diode conducts current thru itself. The value of the resistor must be low enough so that some minimal current is established, which is required for the Zener to start conducting in the reverse breakdown region. The lower the resistance, the more current will flow both thru the Zener and out to the load. Large amounts of current will result in large power dissipation in the resistor and the Zener diode.
Let's look at a Zener datasheet of the two RadioShack devices. NXP Semiconductors (formerly Philips Semiconductors) has a well-written 10-page datasheet which covers both devices. We encounter some new ratings;
A tolerance specification of ±5% of the regulated voltage specification
An IZ (working current in Zener configuration) – maximum amount of current allowed thru the device
A VZ working voltage specification for the Zener configuration (subject to the stated 5% variation)
Test current Itest: current at which the Zener will break down and start regulating. This is a maximum value (due to parameter spread or variability) for any device in a batch from the manufacturer.
Ptot maximum power dissipation due to the VZ*IR product
Imperfection resistance rdif
Note that just like with all diodes, Zener leakage current will increase with temperature.
Other textbooks can be used if the reader wishes to build a Zener regulator circuit, and calculate necessary values such as the resistance. We will not discuss a Zener voltage regulator circuit because of its limitations.
Circuit protection
Suppose we had a circuit or an electronic device which had to be protected from the voltage (either at the power supply terminals, or on any circuit external connections) exceeding a certain maximum value, perhaps as a “spike”, or “transient”, which are very common in real-world operation.
A case in point is the car's electrical supply system. You may want to build a circuit which converts “cigarette lighter” 12V output to 5V to charge a USB device. That “12V” is not an actual value. It is more like:
With ignition OFF, it is 11.6V or lower. A lead-acid battery does not supply an exact 12V, and the voltage drops with any load attached.
With ignition ON and the engine running, it is 13.8V or higher. This is the “charging” voltage for the battery, but the whole car electrical system has to be supplied by this “charging” potential.
Since the car battery is being charged by the very electrically noisy alternator (which means that its output is not a clean and steady 13.8V), and since it has more on-board electrical noise generators (ignition system, etc), that 13.8V sees spikes up to positive and negative hundred volts. On an oscilloscope, that “12V” cigarette lighter output would look like:
Fig 2.21 Typical automotive transients
Width of the spikes has been exaggerated. They are briefer (“narrower” at less than a second). I highly recommend the Harris Corporation Application Note 9312 for reading.
We can use a Zener diode as a “clamp” to not allow those spikes to go over a certain voltage.
Fig 2.22
In the circuit on the left, Vz is chosen to be somewhat higher than a normal Vin level, but less than Vout maximum acceptable value. If a spike with a voltage higher than Vz appears, the Zener diode will clamp it and dissipate it thru itself. We have omitted a current-limiting resistor by assuming that the spike will be brief enough so that it does not exceed Zener power dissipation rating. That is a bad assumption. The circuit on the right uses a fuse, which is selected so that it will blow if that Zener absorbs too much current for too long of a time period. This circuit adds the size and expense of a fuse, and when that fuse blows, a puzzled circuit or electrical device user will wonder why it no longer works (if they are not aware of the fuse). Amazingly, electrical engineers prefer the circuit on the left: should a high voltage appear at Vin, the Zener will eventually overheat and destroy itself, which will apply that high Vin to the circuit, and which will be destroyed later as well.
Back to the car's electrical voltage system. We have voltage spikes of both polarities. How can we use two Zener diodes to protect a circuit from spikes of either polarity? We connect them like so:
Fig 2.23
At either polarity of a spike, one of the Zeners is in forward conduction, while the other regulates the spike down to the Vz. Not surprisingly, there is a special separate electrical device which contains just that – two Zeners. It is called a TVS (Transient Voltage Suppressor). It is made specifically to withstand very high voltage spikes, and to be able to dissipate as heat those high-voltage, high-energy spikes. For a very brief amount of time, these devices can withstand a high amount of current and energy. They are most often used to protect a circuit from inductive kickback, ESD, and lightning
Let us look at a TVS datasheet for Littelfuse SA15CALFCT.
We see a PPPM rating of a maximum power which the device can dissipate inside itself, for an extremely short amount of time, and a much smaller steady-state PD dissipation rating. Note that the device is only designed to operate with bursts of current which it must clamp, and not for “steady-state” operation.
We see an IFSM peak current surge rating.
We also see a response time specification. TVS diodes are much faster compared to other over-voltage protection devices we will study later, but are not as powerful, and are more expensive than other devices.
There is a “reverse current” IR specification. Since we have a bi-directional device, this is actually “leakage current at voltages less than breakdown”. A low current means that the device will not affect (“load”) the circuit it is connected to, until the time when that spikes comes.
There are two voltage ratings listed. One is stand-off, and another one is breakdown. The stand-off voltage is the maximum working voltage of the circuit the TVS is connected to, at which the TVS will not affect the circuit in any significant way besides the leakage current. The breakdown voltage is the voltage at which we want the TVS to act and turn on.
The test current IT specification is that for the minimum amount of current which must flow thru the device, at a voltage higher than breakdown, for the device to turn on and clamp the voltage. This current value can be used for testing purposes to verify device functionality.
The device above clamps voltages exceeding a specific value, of either polarity. Such a device is called a bi-directional TVS. If the device only needs to clip one polarity of voltage spikes, then a uni-directional TVS can be used instead.
Fig 2.23 Uni-directional
In the reverse direction, the device is equivalent to a diode. Notice that this device is extremely similar in its operation to a Zener diode. However, the two are designed for two different tasks, and are not constructed the same.
TVS Application. A recommended circuit to attach to the input of any device or circuit which is fed from the car's electrical system is:
Fig 2.24
This circuit will also eliminate that annoying buzz or noise if a device connected to the cigarette adapter is listened to thru the on-board audio system. Car's ignition system, especially in older cars, is a generator of noises which extend into the AM radio band (1MHz!). Fuse is a slow-blow type, and should be of 1.5x normal or “highest expected” circuit current draw. If it is not a special self-resetting type, it must be user—accessible and replaceable. The output of this circuit is usually attached to a voltage regulator, which reduces the noise further, and produces a necessary regulated voltage level.
2.11 Light Emitting Diode (LED)
An LED is a fascinating device, primarily because most of the electrical energy input is converted to light and not heat as with common light bulbs. However, many people operate an LED without regard for proper parameters.
A common light bulb has the following characteristics which make it so simple to use:
When connected to a voltage source equal to the indicated value, a light bulb draws a limited amount of current.
If connected to a voltage source somewhat higher than rated voltage, it will operate brighter, hotter, and will draw more current, but a small over-voltage condition will be tolerated.
If connected to a voltage which is somewhat lower than the rated voltage, or a voltage source which can be varied from operating voltage down to zero, the lightbulb will glow dimmer proportional to the voltage, all the way down to zero volts.
Ambient temperature around the bulb (within limits) has small effect on light bulb operation.
An LED is very different from a light bulb because it is a semiconductor and a diode. In particular:
It will not start to operate until a relatively large (2.1V and up) forward voltage drop is reached (compare to a diode's 0.6V).
At voltages higher than the forward voltage drop, the LED will operate like a diode, and pass an exponentially increasing current thru itself.
If connected to a variable voltage source, and operated at a voltage other than the forward voltage drop, the brightness of the LED and the current passing thru it are not linearly proportional to the voltage level.
Above a the amount of safe operating current of the LED, it will overheat and permanently damage itself.
Temperature variations, especially at higher temperatures, have a large effect on an LED.
An LED has the following characteristics:
Forward voltage drop VF – minimum voltage to turn the LED ON. Alternately, in a current-limited driving circuit, this is the voltage which the LED will drop across itself.
Forward typical current IF at which a specified brightness is reached, and at which power dissipation in the device (due to the product of VF*IF) will not exceed device absolute rating.
Because the device is encased into transparent plastic, and LED has a low PD power dissipation limit.
Above the voltage level VF, current IF thru the device increases exponentially, and so does the dissipated power thru the device. A small increase above VF leads to device meltdown and destruction. Neither VF nor IF can be exceeded without shortening LED life or completely destroying it.
The life of an LED is irreversibly shortened by driving it with current which exceeds the maximum specified value, or by subjecting the LED to high temperatures, either from internal dissipation or from ambient temperature levels.
Fig 2.25
It is often the case that the following trivial circuit is used to drive an LED:
Resistor value is calculated to be
This circuit has several problems:
Voltage source is assumed to be constant. If it changes, a voltage which will cause a large current flow thru the LED will destroy it, and a low voltage can drop below the forward voltage drop.
LED voltage drop is assumed a specific value. An LED with a different color, semiconductor material, or due to variability, can have a significantly higher or lower voltage drop.
A proper, but somewhat more complicated, LED operating method is to use a constant-current supply. We will discover circuits and devices which can supply this proper current later in the book.
I need to make sure that the reader understands that an LED must be powered by a current-limited voltage supply, the voltage level of which must be limited to the LED forward voltage. Do not ever do the amateur's error in powering an LED off a constant-voltage supply. This picture shows what will happen:
A constant-voltage power supply was attached to this LED, which barely exceeded the VF specification, but which resulted in a huge current flow thru the device, and melting of the plastic encasing.
Fig 2.26
Low LED reverse voltage
An LED has a low voltage rating when connected in reverse. A possible scenario in which an LED is destroyed is when attempting to test it, but connecting it in reverse. It is easy to keep increasing the power supply voltage, to more than a typical VRmax=5V, and to destroy the LED in the process, without ever seeing it come ON (what we were trying to do by increasing the voltage, but failed at).
LED brightness adjustment
It is possible to adjust LED brightness by varying the current thru the device, perhaps by varying the supply voltage to the series resistor-LED circuit. However, the resulting brightness will not be linearly dependent on the supplied voltage. Small voltage increases at the low end of the range will lead to small brightness increases, but small voltage increases near device typical values will lead to large brightness (and current) increases. An added problem is the device turning off suddenly and completely below the voltage of VF. There are two methods to vary LED brightness:
1) Complicated – vary the supplied LED current, while keeping the voltage steady. This involves circuits which contain half a dozen of parts or more.
2) Simpler – pulse the LED quickly with an unvarying supplied current and voltage. If the LED spends most of the time with the current applied, and is off only for very brief amounts of time, then it will appear to be near its maximum brightness. If the LED spends half the time being ON and half the time being OFF, while being pulsed quickly (hundred times to ten thousand times a second), it will appear to the eye to be at half brightness, and so on. This is commonly called PWM – pulse width modulation. We will see this mentioned in the book several more times.
Simplest constant current LED drive circuits
We will have to wait until a later chapter before we can design LED drive circuits which do not use a resistor.
Other LED parameters
LED brightness is measured in candela, although there are other measurement units available.
View angle is the range over which the LED output drops by 50%. 20° to 60° are common. As an example a 30° device plot is shown below, where the y-axis is the multiple of total brightness:
Fig 2.27 ©Cree, C503B device
How to select an LED. With the plethora of devices available, the problem is often in how to limit the choices of selecting among thousands of different devices. Knowing what exactly you need helps you greatly in the electronics parts selection process.
Let's walk thru a typical selection process on Digi-Key:
We will click on Product Index tab on the main page, and then on "LEDs - <75mA, Discrete"
We will select "In Stock" and hit Apply Filters. The number of choices drops significantly.
We will choose Red color.
We will choose Through Hole devices only. This always significantly reduces the number of choices.
Hit Apply Filters
We will order our results by "Millicandela Rating", descending.
We will then pick a device which can be sold in quantities of 1, is cheap, has high brightness, and has a wide view angle. Notice that super-high brightness devices may achieve that status by being extremely focused, to perhaps <20°.
40-60° being a typical value for LEDs which are to be easily viewable off-center, we select a 30° Cree C503B-RCN-CW0Z0AA1 or C503B-RCS-CW0Z0AA1. Their ratings are listed as VF 2.1V, IF 20mA, 7500mcd, at 14 cents each.
Let us look at the datasheet of the chosen device.
We see several specifications which are new to us. First we look at the Absolute Maximum Ratings section.
First one is IFmax. LEDs are designed to be operating with a typical current value. The maximum value listed here is the most the device will tolerate, but only at specific conditions such as low ambient temperature. If the current is exceeded beyond this, the device will be permanently damaged.
Notice the very low Vr reverse voltage rating. Like we discussed before, it is easy to damage the LED in the reverse direction.
PD power dissipation is, of course, the VF*IF product.
We are now in the Typical Operating Condition section.
We see specifications for forward current and voltage, dominant wavelength, intensity, and viewing angle. Question: how can most LEDs listed with the same forward voltage and current ratings have such varying intensities?
Answer: Intensity variations are due to the focusing (viewing angle), efficiency, LED size (the larger the better), better thermal management inside the device, and other factors. However, it pays to comparison shop for the LEDs which are brighter, if they are at the same parameter and cost level as other similar devices.
Connecting several LEDs together
If more than one LED is needed, it would be ideal to use a dedicated power supply for each LED, but that is simply not economically and practically possible.
Fig 2.28 Series connection
A simple way to connect several LEDs together is the series arrangement.
To minimize power waste in the resistor, whenever possible, the number of LEDs should be chosen so that Vsupply ≈ VF * (# of LEDs)
A resistor must still be used even if the equation product on the right only slightly exceeds the power supply voltage rating.
Note the significant flaw of this connection – if one LED goes out, so does the whole chain. You have seen this happening in LED traffic lights where a whole portion of the light is not lit. Note that an LED itself does not have to fail for this to happen – a solder joint or a PCB trace could fail just as well.
Fig 2.29 Parallel connection
At lower power supply voltages, a parallel connection can be used instead.
Why is a resistor connected to each LED? Because of device variability, also known as parameter spread. One of the LEDs will happen to have a lower VF than other devices. Because of this, that device will conduct more forward current than other devices in the chain. Consequently, its temperature will rise higher than other devices. Increasing temperature will lead to an even lower VF, higher current flow, and greater luminosity, which will again increase its forward current, and so on in a spiral towards destruction.
A resistor will eliminate this effect. Whenever a single device will conduct more current than other devices, voltage drop thru that resistor will increase, and this will effectively increase LED VF, canceling the effect. The value of the resistor, usually relatively small, is chosen so that this temperature stabilization mechanism can function over the whole range of temperatures and currents. An understanding of by how much VF will change is needed to calculate the required resistance value.
LED life claims
The ignorant media often claims unsupported and unverified claims of LED lifetimes of 10 years or so. Somewhat similar claims were, and are, being made for CFL (compact fluorescent) light bulbs. Any homeowner who has purchased CFLs knows that they are thrown out about every other year. The bulb itself may last 10 years in a controlled environment of a museum, but it is other factors, which people are ignorant of, that shorten the lifespan.
Here is an example. LEDs were just recently introduced (relatively speaking) into the traffic light application. The problems are easy for the public to see:
1. The waste heat generated by the incandescent traffic lights of the past was “useful” heat in the winter, when it melted snow off the traffic light. LEDs do not produce heat, and traffic lights get obstructed by snow. The first lesson is to use a lighting solution for a proper application.
2. Burned-out or blinking segments of a traffic light, composed of many individual devices. Either a cheap series connection was used, and a single voltage spike, or a single bad overheated device took out the whole series-connected chain, or a single connection point broke.
3. An incandescent light bulb was of simple and reliable construction, as far as the base of the bulb was concerned. In LED lights, this is replaced with a PCB and many solder connections. Solder connections fail due to bad manufacturer’s processes and final inspection, due to thermal cycling, due to corrosives, and due to the use of lead-free solder. Many solder connections are many opportunities for an eventual failure.
From my own experience in electronics manufacturing, I actually expect other components (solder, wire, and PCB connections) to fail long before a properly driven LED would have failed.
It is possible to manufacture an LED lighting solution which will last 10 years and which will be tolerant of several individual LED devices failing, but it will cost more for the manufacturer and for the consumer.
LED polarity identification
Cathode is identified on an LED, with both a flat spot (or a chamfer) on the body, and a longer cathode lead (for new, unclipped devices). If neither of the two is available, it is possible to look inside the plastic body part, and observe which of the two internal metal connections is larger in size – that is the cathode. A DMM can also be used to identify polarity. The “diode check” of modern DMMs is usually sufficient to dimly illuminate an LED in forward bias. Since LEDs are not identified with a part number on them (to save a little money), care should be taken to not mix LEDs together. Additionally, other non-LED electronic parts exist which look like LEDs, but cannot be operated like an LED. If everything is mixed together, it's asking for trouble!
How to find VF and Imax of an unknown LED
Limit power supply current to 20mA
Set power supply voltage to 0
Find anode and cathode leads and connect the LED to the power supply with a proper polarity
Slowly increase the voltage, up to a few volts until LED starts to come on
Note the voltage at which LED is ON and current increases to 20mA, or its maximum value. A table for typical forward voltages based on LED color is given below. The power supply voltage set for the LED you have cannot exceed the maximum given in the table.
To find out LED maximum current and / or power (only if necessary): current limit to 20mA (small LEDs), 50mA (large or high-brightness LED), or higher for very powerful LEDs. Then drive them at the EXACT forward voltage given for that LED type (or voltage found from the previous steps), and record the current.
Color
Voltage drop V
Infrared
V < 1.9
Red
1.63 < V < 2.03
Orange
2.03 < V < 2.10
Yellow
2.10 < V < 2.18
Green
1.9 < V < 4.0
Blue
2.48 < V < 3.7
Ultraviolet
3.1 < V < 4.4
White
V > 3.5
Fig 2.30 ©Wikimedia Creative Commons
2.12 Diode as a Sensor
A special type of diodes is made which are sensitive to light. They are called photodiodes, and are connected to the circuit in reverse bias condition – the cathode is “more positive” than anode. With the diode in a dark environment, it behaves just like a regular reverse-biased diode – little current flows from cathode to anode. However, if the diode is exposed to light, larger reverse leakage current starts to flow, proportional to the amount of light applied.
Fig 2.29
The resulting current is small compared to diode forward conduction current. A typical device has a “reverse dark current” of 20nA flowing at darkness, increasing to a “reverse light current” of 20μA at bright sunlight level. An amplifier is usually needed to control a load from the low photodiode current. We will cover such amplifier circuits later in the book. In general, photodiodes pass too little current to be easy to work with. We will study other devices shorty which are easier to use (photoresistors and phototransistors).
Photodiodes have a varying amount of sensitivity across the visible light spectrum of 740nm to 380nm and in datasheets, a “spectral response” plot is given for this purpose.
As a typical application, there is a NIR (near-infra-red) photodiode inside every piece of electronics which is controlled by a remote control.
Let us look at the datasheet of OSRAM Opto Semiconductors SFH229 photodiode.
We see that its spectral response is shifted towards the infrared spectrum. This is typical for Silicon-based photodiodes. At visible colors (380-740nm), its response is below the maximum.
And, like the reader should know by now, its response is highly temperature dependent, particularly with dark current (noise) increasing greatly with high temperature increase. In particular, a high temperature may be interpreted by a circuit as a presence of illumination on the device!
Fig 2.30
Its light current vs. irradiance plot is linear (hard to see from this log-log plot), which means that device behavior is highly predictable, and an amplifier circuit can be of simple construction. For irradiance reference, full summer sunlight at noon radiance is about 30mW/cm2. That is something like 10,000 lux. Room lighting levels are a lot lower.
Fig 2.31
Notice that the plot below shows two completely different setups. The IP is the reverse current when 5V is connected across the photodiode:
Fig 2.32
Vo, on the other hand, specifies an “open circuit” voltage across the device, like this:
Is the photodiode generating voltage? It sure is! We will use this property shortly.
The plot above was given at a specific set of conditions. The full I-V plot of a photodiode looks like:
Fig 2.33
x-axis shows reverse voltage across the device. We see that the current increases linearly (spacing between the several horizontal lines is the same) with raising illumination levels. Plots like this show three variables at the same time. We will see plots of this type very often.
For intuitive analysis, only two variables should be viewed at a time. For example, by looking only at the illumination level fc vs. the y-axis relationship, we see that reverse currents (y-axis) are given at several values of illumination levels (fc).
Once we are comfortable with this knowledge, we can introduce one more variable, which will expand, but not contradict our previous findings. For example, we see that the previous statement is true, but after a specific value of reverse voltage. We see that at a reverse voltage value lower than a certain value, the plots are no longer horizontal, and are becoming dependent on the reverse voltage value. This is not a desirable behavior. Therefore, we can, for example, constrain our circuit to illumination between dark to 3,000 fc, which will only require a reverse voltage of ≥10V.
Note in particular the relatively high reverse voltage needed across the device to ensure linear response – about 10V in this case.
Photodiode leakage current
Just like with any other diode, a photodiode will show an increasing reverse leakage current with increasing temperature. This increase will be read by an amplifying circuit as an increasing illumination level (it does not know any better). It would not take much of a temperature rise for the rising leakage current to be interpreted as a “full illumination” level.
Opto-isolator
If an LED and a photodiode are placed in a single package, several interesting electronic devices can be made. One possible device is an opto-isolator. Its task is to electrically isolate one part of the circuit from another.
Remember that taking measurements or passing signals from one circuit to another requires a common reference point (COM) between the two circuits. At high voltages such as household AC current, there must not be a connection or a direct current path between AC hot and neutral wires and the power supply output. As an example, there is a special type of a power supply, different from wall-wart transformer-based design. It consists of special electronics which drives a small high-frequency transformer. In order to produce an extremely stable output voltage over the range of 100-240V AC and 50/60Hz frequency, that power supply controller must have feedback, to know what it is producing on the output, and to adjust accordingly.
Fig 2.34
This is a simplified design of an isolation power supply. One LED and one photodiode in a dashed-line rectangle is a single opto-isolator device. It provides feedback to control circuitry while completely isolating the left part of the circuit from the right side, along the dashed vertical line.
Such a device can also be called an opto-coupler. One name or the other is used depending on an application. The reader should pick a device based on the needed output type, and examine its datasheet.
Opto-interrupter
Fig 2.35
If a slot is placed between the LED and the photodiode, the resulting device is called an opto-interrupter. A mechanical part of the electrical device either obstructs or frees the slot, and the electrical part of the device can detect when that happens. For hobby applications, a much easier opto-interrupter device to work with has a logical-level output.
Fig 2.36 Logical-level output photo-interrupter
The triangle-looking part of the device translates the very low μA-level output of the photodiode into very certain and somewhat high current (10mA) logic output, which switches between LOW and HIGH based on whether the slot is blocked or not. This convenient tiny circuitry comes at a little added cost compared to devices without the logic circuitry.
Let us look at the datasheet of a Sharp Microelectronics GP1A52HRJ00F. What we see on the input is a basic LED with VF of 1.1Vtyp, and IF=5mA. On the output is a very easy to to work with (VCC 4.5-17V) logic output, which can output a maximum of 16mA into a digital logic input or a circuit. We also see a total response speed of 25μS, which means that the beam can be obstructed or freed as fast as 20 thousand times a second.
2.13 Diodes as Solar Cells
If a photodiode is placed in bright illumination, and the voltage across it is measured with a sensitive voltmeter, then it will be apparent that photodiodes also produce voltage from light, without having a power supply attached (as we did with photodiodes as sensors).
There is a special form of a photodiode, with a very large surface area, which is called a photo-voltaic solar cell (PV cell).
A PV cell is the easiest, but by far the most expensive method of converting sunlight radiant energy into electrical energy.
A solar panel is a collection of solar cells, each producing 0.6V, wired in series or series-parallel, for higher voltage and current.
Among solar cell specifications are:
No-load voltage. Voltage which the cell produces at illumination, without heavy loading.
Short or maximum current. Maximum current which would flow if the cell was loaded with 0Ω load. It is not a good idea to short-circuit solar cells!
Spectrum response. Like all semiconductor products, solar cells do not respond to light spectrum uniformly, and they show higher response towards the red and infrared part of the spectrum.
Maximum power output, which is the I*V product, but not of the short circuit and open-circuit voltages simultaneously.
Efficiency. At moderate latitudes, at noon of a summer day, the amount of solar radiant power impacting the Earth surface can reach 300W/m2. Typical solar cell efficiencies are now about 20%. Thin-film solar cells, which will be all the rage in the future, will have efficiencies less than 10%. Reflectivity of glass covering the solar panel also factors into total solar panel efficiency.
Variation of output with temperature. Solar cells and panels are highly sensitive to heat. Higher temperature significantly reduces solar cell output and degrades its life.
Rated life. Solar cells degrade from the effects of high temperature and UV radiation, and do not last more than 10 years in outdoors conditions.
Another huge factor is cloud cover. Human eye adjusts instantly to an occasional cover obstructing sunlight, and light clouds do not visibly impact total amount of sunlight. Solar cell output, however, drops very appreciably with cloud cover. Heavy cloud cover just about eliminates solar panel output.
Another crucial solar panel requirement is that it remains pointed at the sun, usually with the aid of automated motorized trackers / positioners.
Due to their unpredictable variability, solar cells are most often used to recharge a battery when light is available, and have a circuit run off the batter when it is not. When a solar cell is not supplying power, its “diode nature” must be kept in mind;
Fig 2.39
The solar cell at illumination will supply current to the battery, recharging it (this circuit is simplistic; the battery charging process is not so simple). However, without illumination, battery voltage will exceed potential across the solar cell, and without the blocking diode DB, current will flow thru the “reverse biased solar cell diode”, draining the battery, and potentially damaging the solar cell.
For a datasheet, we can look at Sanyo Energy M-5412CAR. The reader should not buy this solar panel, or any new solar panel, for that manner. There are plenty of discount resellers (All Electronics or e-Bay for example) which sell solar panels which are more affordable for their power output.
Leakage current
We see that a voltage will be developed across a diode, in a forward direction. However, just like all diodes, the solar panel will have an increasing amount of reverse leakage current with increasing temperature. This leakage current will steal from the forward current generated by the panel. Furthermore, that sunlight energy converted into leakage current will be dissipated within the panel, further raising its internal temperature.
Solar panels don't do very well on a hot summer day. Sitting on top of a hot roof and being warmed by the sun, the panels become just about useless at 80°C due to the huge leakage point at that temperature.
Blocking diode
The blocking diode in the schematic above should, of course, steal as little voltage from the panel as possible. A Schottky diode should be used in this application. Note that no diode is needed if the solar panel is not directly charging a battery cell like is shown.
Cell I-V plot
Fig 2.40
We can easily see the need to list testing conditions for open circuit voltage and short circuit current. By the way, it is not a good idea to check a solar panel by shorting an ammeter across it. A specific value of an adequate-wattage load resistance should be used to prevent short-circuit conditions, at which output voltage will drop below optimal.
Neither does it make much sense to just measure the open circuit voltage with a voltage meter. Again, voltage and current should be both measured with a load resistance.
2.14 Controllable Diodes
Fig 2.41
Special types of diodes exist which can be controlled. The first device we will study is a DIAC. A DIAC starts to conduct current only after a certain voltage level has been reached for a brief amount of time (tens of volts, specific to a device part number).
DIAC symbol looks like:
It is easy to see that a DIAC is bidirectional, and can be used with AC voltages.
Its typical behavior can be described by its operation plot,
Fig 2.42
Two quadrants are mirrors of each other, so we will only examine the positive, positive quadrant (for just one of the devices inside).
What we see is voltage VBO across the device increasing, without causing much of initial increase in current IF. The device is, therefore, closed. At some relatively high voltage VBO, forward current IF thru the device starts to increase, up to IBO value, and then forward current thru the device starts to increase quickly, and voltage drop across the device decreases significantly. The device is is now conducting, and it will pass current without dropping much voltage across the device. To turn off the device, voltage across and current thru the device it have to be dropped to a low enough value. This "turn off" current value is called "holding current".
Since "resistance" of the device in a certain region of the plot decreases with increasing voltage, many engineers call this "negative resistance". It is negative only in the sense of behaving opposite to the Ohm's Law V=I*R. Don't get carried away with fancy terms and just substitute a mental "opposite to Ohm's Law" for "negative resistance". The device is not a resistor, and no physical device with "–2Ω" exists.
There is no control mechanism, per se. DIACs are usually called "triggering diodes". They are used to trigger an another controllable device, when a certain control voltage level is exceeded. When the voltage level is exceeded, the DIAC opens, and turns ON the other device in a very predictable and certain manner.
Let us look at a datasheet of a Fairchild Semiconductor DB3CT. We have:
VBO Breakover voltage 28-36V
IFRM Repetitive peak forward current 2A
IBO Breakover current 100μA
Minimum current thru the device at which breakover will occur.
IB Leakage current 10μA
Device will leak this much current when in the OFF condition.
Other datasheet parameters are by now self-explanatory. We will see how DIACs are used in a bit.
We now look at devices which can be controlled externally. The first device of interest is a Thyristor (it is also called a SCR).
Fig 2.43
Thyristor symbol is as shown on the left.
The "Gate" is a control terminal. Normally, no current flows from the Anode to the Cathode, even at high enough voltages (>0.6V). The diode is turned OFF unless a control voltage is applied to the Gate. After the control voltage is applied, the diode conducts like a regular diode, but it stays on even if the gate control voltage is turned off. The only way to turn the Thyristor OFF is to have the voltage across the device and current thru it drop to a specific low value (near 0), by either disconnecting the load or the power supply. This "turn off" current value is called "holding current".
A positive voltage, with respect to the Cathode, must be applied to the Gate terminal. The minimum current needed to do this is called gate threshold current, and is generally indicated by IGT.
A plot of its characteristics is:
Fig 2.44
Which is similar to a DIAC plot and should be interpreted the same. Looking at the right-side "OFF state" portion of the plot, we see that with a very low voltage across the device, little current would flow. Current increases a little at several volts across the device, but does not grow by much even when a large voltage value is reached. At this time, a control voltage is applied. Dotted line, being read from right to left, means that the device is not found in the dashed portion of the plot; instead, device jumps from the right side point of the dashed line, to its left side. In this "ON state" region of the plot, little voltage increases lead to large current increases thru the device (it is open, and drops little VTM voltage across itself). In the "ON state", device "moves up and down" the "ON state" plot, unless current from the device drops below IH. If it does, the device jumps immediately to the "OFF state" plot portion.
In reverse mode, the device plot looks like a diode reverse breakdown plot, and it behaves very similarly.
The reader may think that the only way to turn off the constant flow of current from the power supply thru the thyristor and into the load is by having a separate power switch at the power supply. However, there are actually many applications where either the power supply supplies pulsed current, or the load is not drawing constant current. One such load is – surprisingly – DC brush-commutated (most popular kind of a) motor. The motor does not draw constant current – at some periods of rotor position, the brushes are not connected to the winding, and the motor for a brief amount of time is not drawing any current. Perhaps the most popular application of thyristors is in cordless power tools like screwdrivers. A pressure/position dependent control in the trigger is connected to a thyristor, which controls the DC motor. This results in a very compact and efficient control unit. Braking can be implemented as well, by placing a thyristor across a motor, and turning it ON to quickly stop the motor by shorting it.
Let us look at a datasheet of a TO-220 STMicroelectronics TYN608RG. The “sensitive gate” specification of the device signifies that little gate current and voltage is needed to turn the device ON. We have:
VDRM repetitive peak off-state voltage 600V
Maximum voltage which can be applied to the device. "Repetitive" here refers to the marketing department's love for stating unrealistically high conditions which the device can withstand, for a very brief amount of time. We are more interested in prolonged operation, so we use "repetitive" ratings.
IT(RMS) On−State Current RMS 8A
180° refers to using the device with AC supply current (when the device will only turn ON during half the AC waveform). DC current operation will mean a lesser current rating capability, since the device will not cool down in AC waveform "OFF" periods. The reader needs to inspect the datasheet further to find a specification for DC operation (such as Figure 2 and 3).
IL latching current 6mA
With the device turned ON by gate terminal voltage, current thru the device (Anode to Cathode) must be of this minimal value for the device to "latch" in the ON condition. If the current is less, removing control voltage will result in the device turning OFF due to insufficient latching current thru the device.
VTM Peak Forward ON−State Voltage 1.6V
Voltage drop thru a device in the ON state. Multiply by RMS current thru the device for ohmic I*R losses and PD amount which would have to be dissipated by the device. This power calculation cannot exceed maximum device power dissipation rating, or a recommended device temperature rise.
IDRM off-state current 1mA @ 125°C
Current leakage thru the device in the OFF condition. It rises x200 over a 100°C increase from its room temperature value!
IGT gate trigger current 200μA
This is the current which must be supplied into the gate terminal.
VGT gate trigger voltage 0.8V
VGD gate turn-off voltage
Gate voltage must drop below this value for the device to turn off.
VRG reverse gate maximum voltage
Maximum voltage in the reverse polarity which can be applied to the gate (negative voltage means a cathode more positive than gate)
IH holding current 5mA
Current thru the device must drop below this value for the thyristor to turn OFF with gate voltage no longer applied.
Vt0 threshold voltage
Voltage across the device must fall below this value for the device to stop latching. Note that a certain small amount of time is required for the device to respond. dV/dt and dI/dt
Voltage or current cannot rise faster than this in the OFF condition, or the device will turn ON by itself (an undesirable characteristic).
Rd dynamic resistance
How close the device is to an ideal switch. Like with the diode non-vertical I-V plot, a large dynamic resistance means that little current will flow thru the device with little voltage across the device, and current will increase with increasing voltage. An undesirable parameter if it is not equal to 0Ω.
A thyristor is a DC device. It can be made to control AC current by connecting two Thyristors together in one package.
Fig 2.45
This device called a Triac. It can control AC current, since at both voltage polarities, one of the "Thyristors" inside will conduct. The turn-off control is even simpler, since when the AC waveform crosses zero during current reversal, the Triac turns OFF if no gate voltage is being applied at that point. This happens 120 times a second, therefore varying the control voltage to the Triac results in the device "adjusting" its setting 120 times a second, fast enough for most applications.
To trigger a Triac ON, either positive or negative voltage is applied to the gate terminal, with respect to the MT1 terminal. This voltage is specified in the datasheet as "VGT – gate trigger voltage".
In a typical TRIAC, the gate threshold current is generally low, but it is dependent on:
Temperature. The higher the temperature, the lower the gate current needed.
Quadrant of operation (combination of possible polarities across the device and at the control terminal).
The voltage applied on the two main terminals MT1 and MT2, with higher voltage between MT1 and MT2 requiring less gate current.
The value of latching current varies with: gate current pulse (amplitude, shape and width), temperature, and quadrant of operation.
Let us look at a datasheet of a TO-220 NXP Semiconductors BT137X-600E,127. This device was chosen to have a “logic level gate” (same as sensitive gate), as well as “four-quadrant operation”. The reader will notice that the tab of this device is coated in plastic. I have intentionally chosen this feature. Triacs are usually used as high-voltage control devices (residential 120V AC and higher). If the device does not have both MT1 and MT2 isolated from the tab, then a dangerous voltage will be present on the device tab. If the device is not insulated from the heatsink, then the dangerous voltage will be present on the heatsink as well. Even when isolation from the heatsink is used, and the isolation hasn't broken down, a troubleshooter can still get an electric shock by touching the tab. I am therefore willingly paying for safety by buying an inherently insulated device, which will cost a little more, and have a somewhat worse thermal conduction compared to a metal tab TO-220 device.
Four-quadrant operation means that the device will trigger with all possible polarities applied to the main and control terminals (switching AC polarity across main terminals and ±control voltage on gate terminal)
Quadrant II
Voltage across positive
Gate voltage negative
Quadrant I
Voltage across positive
Gate voltage positive
Quadrant III
Voltage across negative
Gate voltage negative
Quadrant IV
Voltage across negative
Gate voltage positive
Fig 2.46 Triac four quadrants of operation Looking at the datasheet, we have:
VDRM repetitive peak off-state voltage 600V
Maximum voltage which can be applied to the device. "Repetitive" here refers to their love for stating unrealistically high conditions which the device can withstand, for a very brief amount of time. We are more interested in prolonged operation, so we use "repetitive" ratings.
IT(RMS) RMS on-state current 8A
Since a TRIAC controls an AC waveform, an RMS value is given for maximum current which can be passed thru the device.
IL latching current 35mA
With the device turned ON by gate terminal voltage, current thru the device (MT1 to MT2) must be of this minimal current for the device to "latch" in the ON condition. If the current is less, removing control voltage will result in the device turning OFF due to insufficient latching current thru the device.
VT on-state voltage 1.65V
Voltage drop thru a device in the ON state. Multiply by RMS current thru the device for ohmic I*R losses and PD power which would have to be dissipated by the device. This power calculation cannot exceed maximum device power dissipation rating, or a recommended device temperature rise (50 °C for this device).
ID off-state current 0.5mA
Current leakage thru the device in the OFF condition.
IGT gate trigger current 25mA max
Four values are given. These four are the possible combinations of AC polarity at the terminals, and polarity of gate control voltage. To be sure the device will be switched on, use the largest value. This is the current which must be supplied to the gate terminal.
IH holding current 20mA max
Current thru the device must drop below this value for the Triac to turn OFF with gate voltage no longer applied.
VGT gate trigger voltage 1.3V
All other parameters not listed here have already been discussed in the previous Thyristor section. Thermal data listed in the datasheet should also be easy to understand from the previous section on diode datasheets.
Let us build practical circuits.
Thyristor Brushed DC motor on-off control
The advantages of such a circuit over a mechanical switch are:
Silent "switch" operation
No switch arcing and wearing out
Small size
A finger is not required! This circuit can be controlled from a micro-controller
TRIAC AC load control from AC wall-wart, a load, and phase chopping
TRIAC triggered from DIAC:
Perhaps the most popular applications for Triacs are household incandescent light bulb intensity control and motor speed control. The method of control of a load connected to AC power is like so:
And a typical circuit to accomplish the task is:
Since the device is either On or Off for different lengths of time during the AC waveform, I*R (ohmic) losses are small. A small device can control a kilowatt load, with the controlling circuit being of small size and simple construction, and with the device controlling a large load effortlessly (without any heat generation, etc).
The two control devices mentioned so far only allow an On/Off control of a load. We will now study devices which can be controlled to produce an intermediate and controllable amount of current to the load.
2.15 Diode-type semiconductors DMM test
Forward direction
Reverse direction
Special test
Diode
0.5-0.8V
Open
–
Schottky
0.2-0.5V
Open
–
Zener
0.5-0.9V
Open
Limited current reverse breakdown
Uni-directional TVS
0.5-0.8V
Open
Limited current reverse breakdown
Bi-directional TVS
Open
Open
Limited current bi-directional breakdown
LED
>2.5V if DMM uses high test voltage
Open
Limited current forward conduction / luminance
Fig 2.52
