Although quite simple in its overall nature, this was my first major project encompassing the complete design cycle of concept, design, simulation, purchasing, prototype, testing, cost evaluation, and production and taught me many valuable lessons along the way. This power supply still resides on my work station and is used quite regularly to provide a stable five volt source for use in my many bench top designs before a suitable regulator design has been encompassed into the project. And even though a simple regulator can be used in place of this design, it still serves as a template when a non-standard voltage supply or current is needed. Once the basics of circuit theory are understood, a power supply is an excellent place to start for almost any beginner in electronics as it will most definitely be useful in almost all of one's future projects.
Mousey Power Supply
Ω resistor. The entire project cost around $10 to build, including the estimated costs of materials freely obtained, and supplier quantities of 1000+. The project is split into two sections: rectification and regulation. Although the entirety of the regulation section could be replaced by a single regulator chip and pair of capacitors, the goal here was to design a circuit from the base level up.
To understand how a power supply works, you must first realize that the voltage coming from the power grid is not the same as what comes from a battery. Battery voltage is called "direct current," or DC. Outlet voltage is called "alternating current" or AC. The difference is that the voltage leaving a battery is of a constant level. The magnitude does not change (voltage level or current direction). This is the desire output from a power supply. Voltage from a power outlet is different. It is actually a continuously changing voltage which fluctuates between a positive and negative level meaning that the current flows in one direction for half of the time before switching direction for the other half of the time. Measuring this voltage on an oscilloscope would result in a sinusoidal waveform. The rectification process begins with the transforming of the high voltage alternating current to a low voltage alternating current. This is done by a transformer. The second step is the actual rectifier. A diode bridge is used to force all current, both positive and negative, to flow through the positive supply line. A filter capacitor is then used to create a constant voltage level with minimal ripple to be supplied to the regulator. These components can be seen in detail in Figure 2.2.
In this circuit design, a transformer with a 10:1 ratio is used. This means that the voltage entering primary side of the transformer will be ten times higher than the voltage leaving the secondary side. The resultant voltage is not necessarily what you might expect without a bit of electrical knowledge. Although a typical wall outlet in the United States is considered to be 120V. This is actually the "root mean square" value of the voltage. Using an RMS value is typical for alternating current and is calculated by dividing the peak voltage by the square root of two. reversing this equation reveals that the actual peak voltage from the standard wall outlet is 170V meaning the voltage leaving transformer secondary is around 17V. As mentioned, the next step is to pass this lower voltage sinusoidal wave through a full-wave rectifier bridge. This will cause all of the negative portions of the wave to become positive at the output. This works because current can only flow through a diode in one direction. This current flow is illustrated by focusing on the diode bridge of Figure 2.2 and comparing it to the resulting voltage waveform, shown in Figure 2.3.
As shown in the image, no matter what portion of the voltage waveform enters the bridge, the result will always be a positive. It is then the capacitor's job to level off this voltage to a semi-constant level. This entire process is known as AC to DC conversion. In theory, the rising edge of each voltage peak will work to charge the capacitor. Then, when the voltage level begins to drop, the capacitor will release its voltage to try and keep the line level the same as it was. The end result is DC voltage containing some amount of ripple. The ripple is like the original alternating current, but is very small and oscillates around the DC voltage level instead of circuit ground. It can be calculated by dividing the capacitor current by two times the product of the capacitance and oscillation frequency [VR = IC / (2 C f)] as shown in Figure 2.4.
The voltage leaving the rectifier flows through the control element. The output is then sampled and fed into an error amplifier which compares it to a reference voltage level. The error is then seen by the control element which adjusts its output to compensate. A Dalington transistor, part TIP120D was used for the control. This is actual two series transistors packaged together. The leads have been connected to offer a larger output gain. It also has a metal packaging which will be useful in heat dissipation. A resistive voltage divider is used as the sampling circuit. Increasing the top resistor or decreasing the bottom resistor will cause the output voltage level to rise. This can also be performed by using a potentiometer. A 2N2222 transistor is used as the error amp; however, an actual operation amplifier could be used here as well. The reference voltage is generating by using a Ebenezer diode, part LM385. This diode guarantees that 2.5V will be seen as a reference voltage. This circuit was simulated using PSPICE; the schematic is shown in Figure 3.2.
As shown, an LED is also used to indicate the power status of the supply. This circuit was run through multiple simulations to approximate the outcome when built. The two switches shown in the schematic are used to represent the power being turned on and off at specific times. This will allow the resulting simulation waveforms to show the rise and fall times of the circuit. A large rise time will mean the power supply does not reach its desired value very quickly. A large fall time is not goo either, as this circuit is not requiring any sort of a backup power time once the source has been removed. Ideally, the LED will go out within a second of the power switch being turned off. In the first simulation, the load resistance is given as infinity to represent the open load voltage. The result is seen in Figure 3.3.
As shown, the voltage level is expected to rise incredible quickly and drop within few hundred milliseconds when the source is removed. This same simulation was the performed under varying loads, each yielding similar results. Of course, as the load decreased larger current path was available, the fall time also decreased. Next, the ripple current was also simulated. It was desired to keep this value as low as possible to provide a very stable voltage level. A high ripple voltage indicates a very noisy voltage source, and then can have very ill effects on complicated or fragile circuit designs. A lot of noise is commonly seen in circuits using the more efficient switching power supplies. In this case, we are sacrificing efficiency for stability. Trade offs like this are quite common in circuit design. The ripple voltage simulation is shown in Figure 3.4.
As seen, the ripple is well under five millivolts, peak to peak. This will result in a very stable power supply. Also of note, in a no load situation, the output should be seen a little bit higher than you are wanting. When a load is applied, it will drop the output voltage as current is drawn. Drawing too much current will drop the output voltage to an unusable level which can create havoc on a circuit needing to maintain a constant voltage such as microcontrollers or processors. If the voltage drops too low, these devices will reset in mid operation or become damaged altogether. This is why the feedback loop is necessary on a power supply. If a simple voltage divider is used to create a voltage level, the desired voltage will not be seen when a load is applied because the resistance creating the voltage level has changed. The error in the output must be fed back into the controller so that it can adjust the output as needed. In doing so, a lot of power is dissipated as heat, hence the inefficiency. For example, if our regulator is converting 17V DC to 5V DC for a load drawing one amp, the regulator is actually burning twelve watts of power in heat just to create five! [ (17V - 5V) * 1A = 12W ] That would be considered a bad design when the power supply is limited such as in portable devices, and a much more efficient regulator should be used.
With the simulations giving desired results, the actual circuit can be built and tested. It is first constructed on a breadboard to ensure the correct operation of all parts before being soldered and housed for use. To create the enclosure, I was able to take the base of an old LED lamp I had long since mined for parts and hollow out the battery compartment. There was plenty of space for the transformer and circuit to be fixed inside of this piece, and it would look much nicer than a standard rectangular hobby enclosure. My original design for the circuit housing is shown in Figure 4.1 while the case and circuit are should in Figure 4.2.
Figure 4.2I fixed a few standoffs inside of the enclosure and screwed the circuit to them. This would prevent the board from rattling around inside. The power chord used for the power supply was one from some other device I had long since destroyed. I also wanted easy access to the fuse we were required to include in the design, and because the purpose of this base was to hold the batteries for the lamp, it had a typical cover on the bottom which was held in by spring clips. This provided a quick means to open the box and switch out the fuse. The underside of the enclosure is shown in Figure 4.3.
On the upper side of the enclosure, a standard 12V automotive rocker switch was installed in a hole drilled through the top of the plastic. a green power indication LED was also fixed to the top. To screw post terminals were installed in the front of the supply to provide a means to access the created voltage. These two terminal posts can be seen as ears perked upright, which when combined by the power chord being viewed as a tail and the ovular body of the enclosure gives the power supply it mouse like appearance and name. The supply was measured using both an oscilloscope to confirm the correct voltage level and minimal ripple and a standard multimeter under no load and 50Ω load conditions. Although the multimeter results were slightly different then those obtained by the more expensive lab equipment (which were used in the abstract) the measuring by use of multimeter is shown in Figure 4.4 as proof of the correct operation of the circuit.