The low-power switching power supply can be used in a wide variety of amateur radio designs. The circuit of such a UPS is particularly simple, so it can be repeated even by novice radio amateurs.
Main parameters of the power supply:
Input voltage - 110-260V 50Hz
Power - 15 Watt
Output voltage - 12V
Output current - no more than 0.7A
Operating frequency 15-20kHz
The initial components of the circuit can be obtained from available trash. The multivibrator used transistors of the MJE13003 series, but if desired, they can be replaced with 13007/13009 or similar. Such transistors are easy to find in pulse blocks power supply (in my case they were removed from the computer power supply).
The power supply capacitor is selected with a voltage of 400 Volts (in extreme cases, 250, which I strongly do not recommend)
The zener diode used was a domestic type D816G or an imported one with a power of about 1 watt.
Diode bridge - KTs402B, you can use any diodes with a current of 1 Ampere. Diodes must be selected with a reverse voltage of at least 400 volts. From the imported interior you can install 1N4007 (a complete domestic analogue of KD258D) and others.
The pulse transformer is a 2000NM ferrite ring, the dimensions in my case are K20x10x8, but large rings were also used, but I didn’t change the winding data, it worked fine. The primary winding (network) consists of 220 turns with a tap from the middle, the wire is 0.25-0.45 mm (there is no point anymore).
The secondary winding in my case contains 35 turns, which provides an output of about 12 Volts. The wire for the secondary winding is selected with a diameter of 0.5-1 mm. Maximum power The converter in my case is no more than 10-15 watts, but the power can be changed by selecting the capacitance of capacitor C3 (at the same time, the winding data of the pulse transformer is already changing). The output current of such a converter is about 0.7A.
Select a smoothing capacitance (C1) with a voltage of 63-100 Volts.
At the output of the transformer, you should use only pulse diodes, since the frequency is quite high, conventional rectifiers may not cope. FR107/207 are perhaps the most affordable of the switching diodes, often found in network UPSs.
The power supply does not have any short circuit protection, so you should not short-circuit the secondary winding of the transformer.
I didn’t notice any overheating of the transistors; with an output load of 3 Watts (LED assembly), they are icy, but just in case, they can be installed on small heat sinks.
List of radioelements
| Designation | Type | Denomination | Quantity | Note | Shop | My notepad |
|---|---|---|---|---|---|---|
| VT1, VT2 | Bipolar transistor | MJE13003 | 2 | 13007/13009 | To notepad | |
| VDS1 | Diode bridge | KTs402A | 1 | Or another low-power one | To notepad | |
| VDS2 | Diode bridge | 1 | Any up to 2A | To notepad | ||
| VD1 | Zener diode | D816G | 1 | To notepad | ||
| C1 | 220 µF 440V | 1 | To notepad | |||
| C2 | Electrolytic capacitor | 1000 uF x 16V | 1 | To notepad | ||
| C3 | Capacitor | 2.2 uF x 630V | 1 | Film |
As a rule, repairing such an inexpensive device is not economically profitable.
Especially in non-poor countries. average price 5 dollars.
But it happens that there is no extra money, but there is time and spare parts.
There is no store nearby. Circumstances do not allow. Then it's not about price.
In my case, everything was simple - one of my two chargers broke Nokia AC-3E, friends brought a bag of broken chargers. Among them were a dozen branded Nokia chargers. It was a sin not to take it.
The search for the circuit did not lead to anything, so I took a similar one and converted it to the AC-3E. Many chargers for mobile phones. As a rule, the difference is insignificant. Sometimes the denominations are changed, a little more or a little less elements, sometimes a charge indication is added. But basically the same thing.
That's why this description and the diagram will be useful for repairing not only the AC-3E.
The repair instructions are simple and written for non-specialists.
The scheme is clickable and of good quality. 
THEORY.
The device is a blocking oscillator operating in a self-oscillating mode. It is powered by a half-wave rectifier (D1, C1) with a voltage of approximately +300 V. Resistor R1, R2 limits the starting current of the device and acts as a fuse. The blocking oscillator is based on a transistor MJE13005 and a pulse transformer. A necessary element of the blocking generator is a positive feedback circuit formed by winding 2 of the transformer, elements R5, R4 C2.
The 5v6 zener diode limits the voltage at the base of the MJE13005 transistor to within five volts.
Damper circuit D3, C4, R6 limits voltage surges on winding 1 of the transformer. At the moment the transistor is turned off, these surges can exceed the supply voltage several times, so the minimum permissible voltage of capacitor C4 and diode D3 must be no lower than 1 kV.
PRACTICE.
1. Disassembly. Self-tapping screws holding the charger cover in this device They look like a triangular star. As a rule, there is no special screwdriver at hand, so you have to get out as best you can. I unscrewed it with a screwdriver, which, over the course of its use, had become sharpened into all sorts of crosses.
Sometimes chargers are assembled without bolts. In this case, the body halves are glued together. This indicates the low cost and quality of the device. Disassembling such a memory is a little more difficult. You need to split the body with a non-sharp screwdriver, gently pressing on the joint of the halves.
2. External inspection of the board. More than 50% of defects can be detected through external inspection. Burnt resistors and a darkened board will show you the location of the defect. A burst case or cracks on the board will indicate that the device was dropped. Chargers are used in extreme conditions, so falls from anywhere are a common cause of failure.
In five out of ten memory systems that I had the opportunity to do, they were trivial contacts bent through which 220 volts are supplied to the board.
To fix it, just slightly bend the contacts towards the board.
You can check whether the contacts are at fault or not by soldering the power cord to the board and measuring the voltage at the output - the red and black wires.
3.
Broken cord at the output of the charger. It usually breaks at the plug itself or at the base of the charger. Especially for those who like to talk while charging the phone.
Called the device. Insert the lead of a thin part into the center of the connector and measure the resistance of the wires.
4.
Transistor + resistors. If there is no visible damage, first of all you need to unsolder the transistor and ring it. It must be borne in mind that the transistor
MJE13005 the base is on the right, but it also happens the other way around. The transistor may be of a different type, in a different housing. Let's say MJE13001 looks like a Soviet KT209 with the base on the left.
Instead I installed MJE13003. You can install a transistor from any burnt-out lamp - a housekeeper. In them, as a rule, the filament of the bulb itself burns out, and the two high-voltage transistors remain intact.
5. Consequences of overvoltage. In the simplest case, they are expressed in a short-circuited diode D1 and a broken resistor R1. In more complex cases, the MJE13005 transistor burns out and inflates capacitor C1. All this simply changes to the same or similar details.
In the last two cases, in addition to replacing the burnt conductors, you will need to check the resistors around the transistor. With the diagram this will be easy to do.
Silicon transistors n-p-n structures, high voltage amplifiers. Production of transistors 13001 is localized in Southeast Asia and India. Used in low-power switching power supplies, chargers ah for various mobile phones, tablets, etc.
Attention! With loved ones (almost identical) general parameters at different manufacturers transistors 13001 can differ in pin locations.
Available in plastic housings TO-92, with flexible leads, and TO-126 with rigid leads. The type of device is indicated on the housing.
The figure below shows the MJE13001 and 13001 pinouts from different manufacturers, with different housings.
The most important parameters.
Current transfer coefficient 13001 may have from 10
before 70
, depending on the letter.
For MJE13001A - from 10
before 15
.
For MJE13001B - from 15
before 20
.
For MJE13001C - from 20
before 25
.
For MJE13001D - from 25
before 30
.
For MJE13001E - from 30
before 35
.
For MJE13001F - from 35
before 40
.
For MJE13001G - from 40
before 45
.
For MJE13001H - from 45
before 50
.
For MJE13001I - from 50
before 55
.
For MJE13001J - from 55
before 60
.
For MJE13001K - from 60
before 65
.
For MJE13001L - from 65
before 70
.
Current transmission limit frequency
- 8
MHz.
Maximum collector - emitter voltage - 400 V.
Maximum collector current (constant) -
200
mA.
Collector-emitter saturation voltage at collector current 50mA, base 10mA - 0,5 V.
Base-emitter saturation voltage with a collector current of 50mA, base current of 10mA - no higher 1,2
V.
Collector power dissipation- in TO-92 housing - 0.75 W, in TO-126 housing - 1.2 W without radiator.
Use of any materials from this page is permitted provided there is a link to the site
I present another device from the “Don’t Take!” series.
The kit comes with a simple microUSB cable, which I will test separately with a bunch of other laces.
I ordered this charger out of curiosity, knowing that in such a compact case it is extremely difficult to make a reliable and safe device mains power 5V 1A. The reality turned out to be harsh...
It came in a standard bag with bubble wrap.
The case is glossy, wrapped in protective film.
Overall dimensions with plug 65x34x14mm 
The charger immediately turned out to be inoperative - a good start...
At first, the device had to be disassembled and repaired in order to be able to test it.
It is very easy to disassemble - on the latches of the plug itself.
The defect was discovered immediately - one of the wires to the plug fell off, the soldering turned out to be of poor quality. 
The second soldering is no better 
The installation of the board itself was done normally (for the Chinese), the soldering was good, the board was washed. 
Real device diagram 
What problems were found:
- Quite weak attachment of the fork to the body. The possibility of her remaining disconnected from the socket is not excluded.
- Lack of input fuse. Apparently those same wires to the plug are the protection.
- Half-wave input rectifier - unjustified savings on diodes.
- Small capacitance of the input capacitor (2.2 µF/400V). The capacity is clearly insufficient for the operation of a half-wave rectifier, which will lead to increased voltage ripple across it at a frequency of 50 Hz and to a decrease in its service life.
- Lack of input and output filters. Not a big loss for such a small and low-power device.
- The simplest converter circuit using one weak transistor MJE13001.
- A simple ceramic capacitor 1nF/1kV in the noise suppression circuit (shown separately in the photo). This is a gross violation of device security. The capacitor must be of at least Y2 class.
- There is no damper circuit for suppressing reverse emissions of the primary winding of the transformer. This impulse often breaks through the power key element when it heats up.
- Lack of protection against overheating, overload, short circuit, and increased output voltage.
- The overall power of the transformer clearly does not reach 5W, and its very miniature size casts doubt on the presence of normal insulation between the windings.
Now testing.
Because The device is not inherently safe; the connection was made through an additional mains fuse. If something happens, at least it won’t burn you and won’t leave you without light.
I checked it without the housing so that I could control the temperature of the elements.
Output voltage without load 5.25V
Power consumption without load less than 0.1 W
Under a load of 0.3A or less, charging works quite adequately, the voltage maintains a normal 5.25V, the output ripple is insignificant, the key transistor heats up within normal limits.
Under a load of 0.4A, the voltage begins to fluctuate slightly in the range of 5.18V - 5.29V, the ripple at the output is 50Hz 75mV, the key transistor heats up within normal limits.
Under a load of 0.45A, the voltage begins to noticeably fluctuate in the range of 5.08V - 5.29V, the ripple at the output is 50Hz 85mV, the key transistor begins to slowly overheat (burns your finger), the transformer is lukewarm.
Under a load of 0.50A, the voltage begins to fluctuate greatly in the range of 4.65V - 5.25V, the ripple at the output is 50Hz 200mV, the key transistor is overheated, the transformer is also quite hot.
Under a load of 0.55A, the voltage jumps wildly in the range of 4.20V - 5.20V, the ripple at the output is 50Hz 420mV, the key transistor is overheated, the transformer is also quite hot.
With an even greater increase in load, the voltage drops sharply to indecent values.
It turns out that this charger can actually produce a maximum of 0.45A instead of the declared 1A.
Next, the charger was collected in the case (along with the fuse) and left in operation for a couple of hours.
Oddly enough, the charger did not fail. But this does not mean at all that it is reliable - having such circuitry it will not last long...
In short circuit mode, charging quietly died 20 seconds after switching on - the key transistor Q1, resistor R2 and optocoupler U1 broke. Even the additionally installed fuse did not burn out.
For comparison, I’ll show you what a simple Chinese 5V 2A tablet charger looks like inside, manufactured in compliance with the minimum permissible safety standards. 
Taking this opportunity, I inform you that the lamp driver from the previous review has been successfully modified and the article has been updated.
Most modern network chargers are assembled using a simple pulse circuit, using one high-voltage transistor (Fig. 1) according to a blocking generator circuit.
Unlike more simple circuits on a step-down 50 Hz transformer, the transformer of pulse converters of the same power is much smaller in size, which means the size, weight and price of the entire converter are smaller. In addition, pulse converters are safer - if in a conventional converter, when the power elements fail, the load receives a high unstabilized (and sometimes even alternating) voltage from the secondary winding of the transformer, then in case of any malfunction of the “pulse generator” (except for the failure of the optocoupler feedback- but it is usually very well protected) there will be no voltage at all at the output.
Rice. 1
Simple pulse circuit blocking generator
A detailed description of the principle of operation (with pictures) and calculation of the circuit elements of a high-voltage pulse converter (transformer, capacitors, etc.) can be read, for example, in “TEA152x Efficient Low Power Voltage supply” at the link http://www. nxp.com/acrobat/applicationnotes/AN00055.pdf (in English).
The alternating mains voltage is rectified by diode VD1 (although sometimes the generous Chinese install as many as four diodes in a bridge circuit), the current pulse when turned on is limited by resistor R1. Here it is advisable to install a resistor with a power of 0.25 W - then if overloaded, it will burn out, acting as a fuse.
The converter is assembled on transistor VT1 using a classic flyback circuit. Resistor R2 is needed to start generation when power is applied; in this circuit it is optional, but with it the converter works a little more stable. Generation is maintained thanks to capacitor C1, included in the PIC circuit on the winding, the generation frequency depends on its capacitance and the parameters of the transformer. When the transistor is unlocked, the voltage on the lower terminals of windings I and II in the diagram is negative, on the upper ones it is positive, the positive half-wave through capacitor C1 opens the transistor even more strongly, the voltage amplitude in the windings increases... That is, the transistor opens like an avalanche. After some time, as capacitor C1 charges, the base current begins to decrease, the transistor begins to close, the voltage at the upper terminal of winding II in the circuit begins to decrease, through capacitor C1 the base current decreases even more, and the transistor closes like an avalanche. Resistor R3 is necessary to limit the base current during circuit overloads and surges in the AC network.
At the same time, the amplitude of the self-induction EMF through the diode VD4 recharges the capacitor SZ - that is why the converter is called flyback. If you swap the terminals of winding III and recharge the capacitor SZ during forward motion, then the load on the transistor during forward motion will sharply increase (it may even burn out due to too much high current), and during the reverse stroke the self-induction EMF will be unspent and will be released at the collector junction of the transistor - that is, it can burn out from overvoltage. Therefore, when manufacturing the device, it is necessary to strictly observe the phasing of all windings (if you mix up the terminals of winding II, the generator simply will not start, since capacitor C1 will, on the contrary, disrupt generation and stabilize the circuit).
The output voltage of the device depends on the number of turns in windings II and III and on the stabilization voltage of the zener diode VD3. The output voltage is equal to the stabilization voltage only if the number of turns in windings II and III is the same, otherwise it will be different. During the reverse stroke, capacitor C2 is recharged through diode VD2, as soon as it is charged to approximately -5 V, the zener diode will begin to pass current, the negative voltage at the base of transistor VT1 will slightly reduce the amplitude of the pulses on the collector, and the output voltage will stabilize at a certain level. The stabilization accuracy of this circuit is not very high - the output voltage varies within 15...25% depending on the load current and the quality of the zener diode VD3.
A circuit of a better (and more complex) converter is shown in rice. 2
Rice. 2
Electrical circuit of a more complex
converter
To rectify the input voltage, a diode bridge VD1 and a capacitor are used; the resistor must have a power of at least 0.5 W, otherwise at the moment of switching on, when charging capacitor C1, it may burn out. The capacitance of capacitor C1 in microfarads must be equal to the power of the device in watts.
The converter itself is assembled according to the already familiar circuit using transistor VT1. A current sensor on resistor R4 is included in the emitter circuit - as soon as the current flowing through the transistor becomes so large that the voltage drop across the resistor exceeds 1.5 V (with the resistance indicated on the diagram being 75 mA), transistor VT2 opens slightly through diode VD3 and limits the base current of transistor VT1 so that its collector current does not exceed the above 75 mA. Despite its simplicity, this protection circuit is quite effective, and the converter turns out to be almost eternal even with short circuits in the load.
To protect transistor VT1 from emissions of self-induction EMF, a smoothing circuit VD4-C5-R6 was added to the circuit. The VD4 diode must be high-frequency - ideally BYV26C, a little worse - UF4004-UF4007 or 1 N4936, 1 N4937. If there are no such diodes, it is better not to install a chain at all!
Capacitor C5 can be anything, but it must withstand a voltage of 250...350 V. Such a chain can be installed in all similar circuits (if it is not there), including in the circuit according to rice. 1- it will noticeably reduce the heating of the switch transistor housing and significantly “extend the life” of the entire converter.
The output voltage is stabilized using the zener diode DA1 located at the output of the device, galvanic isolation is provided by the optocoupler V01. The TL431 microcircuit can be replaced with any low-power zener diode, the output voltage is equal to its stabilization voltage plus 1.5 V (voltage drop across the LED of the optocoupler V01)’; a small resistance resistor R8 is added to protect the LED from overloads. As soon as the output voltage becomes slightly higher than expected, current will flow through the zener diode, the optocoupler LED will begin to glow, its phototransistor will open slightly, the positive voltage from capacitor C4 will slightly open transistor VT2, which will reduce the amplitude of the collector current of transistor VT1. The instability of the output voltage of this circuit is less than that of the previous one and does not exceed 10...20%; also, thanks to capacitor C1, there is practically no 50 Hz background at the output of the converter.
It is better to use an industrial transformer in these circuits, from any similar device. But you can wind it yourself - for an output power of 5 W (1 A, 5 V), the primary winding should contain approximately 300 turns of wire with a diameter of 0.15 mm, winding II - 30 turns of the same wire, winding III - 20 turns of wire with a diameter of 0 .65 mm. Winding III must be very well insulated from the first two; it is advisable to wind it in a separate section (if any). The core is standard for such transformers, with a dielectric gap of 0.1 mm. As a last resort, you can use a ring with an outer diameter of approximately 20 mm.
Download: Basic pulse circuits network adapters for charging phones
If you find broken links, you can leave a comment, and the links will be restored as soon as possible.
Loading...
