For quite a long time, the author of the YouTube channel “Open Frime TV” had a desire to repeat the design of the laptop power supply, as they have proven themselves from the best side. And now the day has come. In this article, we will analyze in detail the full process of creating such a device from development to testing and testing.
A considerable number of notebook power supplies have been in the hands of the author. I came across both excellent circuit solutions and, frankly, not very good ones.
In this project, the unit will be built according to this (see image below), a fairly common scheme that is used in many industrial power supplies.
The circuitry of this unit does not differ from microcircuits, which have both a controller and a transistor in the same case. However, this solution has a wider range of parameters that can also be adjusted, namely current, output power and frequency.
Now let's look at the circuit in more detail. Looking at the diagram, we see an ordinary flyback power supply.
The principle of operation of such a block was analyzed in detail in one of the author's previous projects. Now we will touch on only the main points, and start with the microcircuits that can be installed here. The UC3840 series includes 4 different microcircuits.
In turn, they are divided into 2 parts: with 50 percent filling and with 100 percent filling.
For those with 50% filling, the output frequency is 2 times less than the master one. To assemble a flyback power supply, it is better to use microcircuits with a 50 percent filling in order to protect yourself from core saturation. Therefore, the UC3844 and UC3845 chips are suitable for us. They differ from each other only in starting voltage, 9 or 18 volts.
At first glance, it may seem that this is not the most important point. But it is worth thinking about what voltage we will apply to the gate of the transistor. With an 18-volt microcircuit, we will have to apply just such a voltage to the gate. But with a 9-volt one, we can safely limit ourselves to 12-13V.
It is also worth adding that according to the datasheet, these microcircuits work stably from 0 degrees, therefore, for work at negative temperatures, it is advisable to use the UC2840 series.
Now we turn to the rest of the circuit. Let's look at the input filter first.
Manufacturers often save on it, but you shouldn't. In this example, the author used X 0.15 μF capacitors and a common mode choke on a special core.
The main tank is next.
It should be chosen at the rate of 1 μF per 1 W of output power. We cannot count the damper yet, we will return to it after winding the transformer.
In many laptop (and not only) power supplies, you can see that manufacturers install 2 resistors in series for self-feeding.
This is done for reasons of reliability, since in theory the mains voltage can break through one resistor.
Below we have a power transistor.
Here you can use transistors such as 4N60, 6N60 or even 10N60. It should be understood here that the more powerful the transistor, the greater the switching losses, so always choose the transistor wisely.
The current limiting resistor will be kindly calculated by the Starichka program when calculating the transformer.
At the output of the unit, everything is pretty standard: a diode, a capacitor, and an additional filter (choke and one more capacitor) is installed. This helps to reduce ripple at the power supply output.
TL431 acts as an error amplifier that drives the optocoupler.
The optocoupler must be powered directly from the output diode, and the feedback can be taken after an additional filter. Thus, we will avoid parasitic vibrations from the throttle.
We choose the current of the optocoupler LED approximately equal to 10 mA.
For a small adjustment of the output voltage, a trimmer resistor has been added to the board.
The output was not without a load resistor, which keeps the unit in working mode, preventing it from falling into relay mode.
The power dissipated on it should be approximately equal to 1W.
We have finished with the circuit, we are starting to manufacture the printed circuit board. You can download the schematic and the seal in the archive HERE.
The author decided not to bother and ordered the manufacture of a printed circuit board in a fairly well-known Chinese company for the production of printed circuit boards.
At the output we have printed circuit boards of excellent factory quality, which facilitates the process of assembling and setting up the device.
After soldering all the elements, proceed to winding the transformer.
In more detail, the whole process of manufacturing a flyback power supply is shown in the author's video:
We remember this parameter and go to the Old Man program and open the tab “Calculations of the RCD clamper”.
First of all, we are interested in these parameters:
With inductance everything is clear, but oscillations need to be found by yourself. To do this, connect the block to the network. We do not have a clamper, so we will not heavily load the unit, a maximum of 20% of the nominal value. We become the ground probe for the plus (+) power supply, and the measuring probe for the drain of the transistor.
We are now interested in free vibrations. To do this, use the cursor and estimate the length of one period.
As you can see, the period turned out to be in the region of 1.7 μs. We remember this value and look at the second period from the leakage inductance. It is at the beginning of the reverse stroke, immediately after the ejection. Also, using the cursor, we find the value of one period of interest to us.
As you can see, it is approximately 152 ns. We transfer the obtained values to the program, after which we make the calculation.
Now let's define the amplitude of the ejection. The transistor is 600V. The maximum power supply can reach 370V. Therefore, only 230V is left for overshoot and reflected voltage. The reflected voltage is approximately 130V. This means that the emission should not exceed 100V.
The program calculated the capacitance of the 4.3nF capacitor. The emission is 74V.
The next step is to determine the correct choice of the resistor.
The capacitor is discharged to a voltage of 120V, which is lower than the reflected voltage, and this is an extra cost for charging the capacitor. Based on this, you need to slightly increase the resistance of the resistor. This field will help us with this:
It is necessary to enter the value of the voltage remaining on the capacitor after the surge. We look, like the first time, with an oscilloscope. We need the difference between the maximum discharge of the capacitor and the value immediately after closing the diode.
That's basically all, the assembly of the flyback power supply is complete. You can start testing. First test under load:
As you can see, the voltage is in place, which means the feedback is working correctly.
The second test is a short circuit test. We shorten the exit and see how the current protection will work.
Well, the last test is a test for pulsations, where can we go without them.
As you can see, here everything is OK as well. That's all. Thank you for attention. Until next time!