Power supply: what can be made from an energy-saving lamp? Operation and repair of compact fluorescent lamps. Compact fluorescent lamp circuit.

Despite the small size of energy saving lamps, they contain many electronic components. By its design, it is an ordinary tubular fluorescent lamp with a miniature bulb, but only coiled into a spiral or other spatial compact line. It is therefore called a compact fluorescent lamp (CFL).

And it is characterized by all the same problems and malfunctions as for large tubular bulbs. But the electronic ballast of a light bulb, which has ceased to shine, most likely due to a burnt-out spiral, usually retains its operability. Therefore, it can be used for any purpose as a switching power supply unit (in short, UPS), but with preliminary refinement. This will be discussed further. Our readers will learn how to make a power supply from an energy-saving lamp.

What is the Difference Between UPS and Electronic Ballast

We will immediately warn those who expect to receive a powerful power source from the CFL - it is impossible to obtain high power as a result of a simple rework of the ballast. The fact is that in inductors that contain cores, the working magnetization zone is strictly limited by the design and properties of the magnetizing voltage. Therefore, the pulses of this voltage generated by the transistors are precisely matched and determined by the circuit elements. But such an electronic ballast power supply is quite sufficient to power the LED strip. Moreover, a switching power supply from an energy-saving lamp corresponds to its power. And it can be up to 100 W.

The most common CFL ballast circuit is a half-bridge (inverter) circuit. It is an autogenerator based on a TV transformer. The TV1-3 winding magnetizes the core and acts as a choke to limit the current through the EL3 lamp. Windings TV1-1 and TV1-2 provide positive feedback for the appearance of a voltage that controls transistors VT1 and VT2. The diagram in red shows the CFL flask with elements that ensure its launch.

An example of a common CFL ballast scheme

All inductors and capacitances in the circuit are selected so as to obtain precisely dosed power in the lamp. The performance of the transistors is associated with its value. And since they do not have radiators, it is not recommended to strive to get significant power from the converted ballast. There is no secondary winding in the ballast transformer from which the load is supplied. This is the main difference between it and the UPS.

What is the essence of ballast reconstruction

To be able to connect the load to a separate winding, it is necessary either to wind it on the L5 inductor, or to use an additional transformer. Conversion of ballast in a UPS provides:

For further conversion of the electronic ballast into a power supply from an energy-saving lamp, a decision must be made regarding the transformer:

• use the existing choke by modifying it;
• or use a new transformer.

Choke transformer

Let's consider both options below. In order to use an electronic ballast choke, it must be removed from the board and then disassembled. If an W-shaped core is used in it, it contains two identical parts that are connected to each other. In this example, orange adhesive tape is used for this purpose. It is carefully removed.
Removing the tape holding the halves of the core

The core halves are usually glued so that there is a gap between them. It serves to optimize the magnetization of the core, slowing down this process and limiting the rate of current rise. We take our pulse soldering iron and heat the core. We attach it to the soldering iron with the joints of the halves.

Having disassembled the core, we get access to the coil with the wound wire. It is not recommended to unwind the winding that is already on the coil. This will change the magnetization mode. If the free space between the core and the coil allows one layer of fiberglass to be wrapped to improve the insulation of the windings from each other, this should be done. And then wind ten turns of the secondary winding with a wire of suitable thickness. Since the power of our power supply will be small, a thick wire is not needed. The main thing is that it fits on the coil, and the halves of the core are put on it.

After winding the secondary winding, collect the core and fix the halves with adhesive tape. We assume that after testing the power supply unit it will become clear what voltage is created by one turn. After testing, we will disassemble the transformer and add the required number of turns. Usually, the alteration is aimed at making a voltage converter with an output of 12V. This allows you to get a battery charger when using stabilization. The same voltage can be made from an energy-saving lamp, as well as a flashlight powered by a battery.

Since the transformer of our UPS, most likely, will have to be completed, it is not worth soldering it into the board. It is better to solder the wires sticking out of the board, and solder the leads of our transformer to them during testing. The ends of the terminals of the secondary winding must be cleaned of insulation and covered with solder. Then, either on a separate panel, or directly at the terminals of the wound winding, it is necessary to assemble a rectifier on high-frequency diodes according to the bridge scheme. A 1 μF 50 V capacitor is sufficient for filtering during voltage measurement.

UPS testing

But before connecting to a 220 V network, a powerful resistor must be connected in series with our block, which was converted by hand from a lamp. This is a safety measure. If a short-circuit current flows through the switching transistors in the power supply, the resistor will limit it. In this case, a 220 V incandescent light bulb can become a very convenient resistor. In terms of power, it is enough to use a 40–100-watt lamp. If there is a short circuit in our device, the light will glow.

Next, we connect the multimeter probes to the rectifier in the DC voltage measurement mode and apply a voltage of 220 V to the electrical circuit with a light bulb and a power supply board. Strands and open live parts must be insulated beforehand. It is recommended to use a wire switch to supply voltage, and put the light bulb in a liter jar. Sometimes they burst when turned on, and the fragments scatter to the sides. The tests usually pass without problems.

More powerful UPS with separate transformer

They allow you to determine the voltage and the required number of turns. The transformer is being finalized, the unit is tested again, and after that it can be used as a compact power supply, which is much smaller than its counterpart based on a conventional 220 V steel core transformer.

To increase the power of the power supply, it is necessary to use a separate transformer made in the same way from a choke. It can be removed from a higher wattage light bulb burned out completely along with the semiconductor ballast products. It is based on the same circuit, which differs in the connection of an additional transformer and some other parts shown in red lines.

The rectifier shown in the image contains fewer diodes compared to the rectifier bridge. But for its work, more turns of the secondary winding are required. If they do not fit into the transformer, a rectifier bridge must be used. A more powerful transformer is made, for example, for halogens. Anyone who has used a conventional transformer for a lighting system with halogens knows that they are powered by a fairly large current. Therefore, the transformer is bulky.

If the transistors are placed on radiators, the power of one power supply unit can be significantly increased. And in terms of weight and dimensions, even several such UPSs for working with halogen lamps will turn out to be smaller and lighter than one transformer with a steel core of equal power. Another option for using workable housekeeping ballasts could be their reconstruction for an LED lamp. Converting an energy-saving lamp to an LED design is very simple. The lamp is disconnected, and a diode bridge is connected instead.

A certain number of LEDs are connected at the output of the bridge. They can be connected in series with each other. It is important that the LED current is equal to the current in the CFL. can be called a valuable mineral in the era of LED lighting. They can find use even after the end of their service life. And now the reader knows the details of this application.

Modern manufacturers offer energy efficient lamps of different sizes, powers, equipped with different bases. Also, lighting devices have a different structure, from which their schemes differ. Depending on the manufacturer, you can choose products with more complex mechanisms that will have high-quality elements of an electronic ballast (ECG).

Features of the schemes

There are inexpensive models on the market, but they often lack critical components that affect the life of the product. The most popular manufacturers in Russia are:

• Navigator (domestic manufacturer);
• MAXUS (international British-English corporation);
• DeLux (Chinese manufacturer);
• Camelion (umbrella brand, originated in Hong Kong and successfully integrated today in Europe, Asia and America).

The circuit of an energy-saving lamp is its so-called heart, with the help of which the entire lighting device functions. The composition of the electronic board may include parts of various qualities and sizes, depending on the good faith of the manufacturer. It is worth noting that high-power devices, equivalent to incandescent lamps of 105 watts and above, cannot have small elements, since the electrical circuit must be equipped with massive parts to ensure normal operation.

If we compare the “Maxus” and “Navigator” bulbs, one can be sure that their components will be different. This means that companies cooperate with different manufacturers of electrical parts or use different approaches to creating these elements on their own.

In general, all lamp circuits for 20, 30, 60 W and higher will be very similar to each other, which helps to repair them if some mechanisms fail.

The principle of the housekeeper

An energy-saving lamp works in much the same way as linear fluorescent lamps. Its glow is provided by the passage of voltage through the electrodes located at the edges of the glass bulb. The tube is filled with inert gas and mercury vapor or mercury compounds. When the medium inside the lamp heats up, ionized electrons are formed, which collide with gas atoms at high speed. All this leads to the formation of a low-temperature plasma that emits ultraviolet radiation.

However, a person cannot perceive either ultraviolet or infrared radiation. To convert it into light visible to our eyes, a special coating is used - a phosphor. Passing through it, ultraviolet rays turn into uniform, bright, saturated illumination.

Due to its low power, a 20 Watt economy has a higher efficiency than a 100W incandescent lamp. Consider why light bulbs help conserve energy and how they work.

Components of the circuit

An energy-saving lighting device consists of the lamp itself and an electronic ballast, also called an electrical circuit. All electronic elements are designed to ensure uninterrupted and correct operation of the lamp.The biggest distinguishing feature of these devices from conventional incandescent lamps is that they operate on direct voltage, and not on the alternating voltage supplied by the mains. It is for this reason that the electronic ballast is built into the lamp housing itself; it is used to preform, distribute and protect the mechanism. The connection diagram contains the following components:

• high-voltage low-power diodes;
• anti-jamming choke;
• medium power transistors;
• high-voltage electrolyte (most often 400 V);
• capacitors of different capacities, but of the same voltage (250 V);
• high frequency transformers (2 pieces);
• resistors.

How the lamp is ignited

When the voltage hits the dinistor, a pulse is generated that goes to the transistor and provokes its opening. After starting is complete, this part of the circuit is blocked by a diode. After opening the transistor, the capacitor is discharged, which is necessary to prevent the dinistor from re-opening. Transistors act on a transformer. It is made of a ferrite ring, processed with three windings arranged in several rows. The voltage across the filaments is supplied through a capacitor from a boost resonant circuit.

The glow in the tube starts at a resonant frequency, which is determined by a capacitor of higher capacity. At the moment of ignition, its voltage is up to 600 W. At startup, it exceeds the average by 5 times, therefore it is important that the flask is intact and sealed. Failure to do so may damage the transistors.

After complete ionization of the gas in the flask, the capacitor with the largest capacity, which determined the luminescence frequency, is shunted. This leads to a decrease in frequency and a transition of the generator control to the second capacitor. The generated voltage decreases, but remains within the range that is necessary to keep the light bulb burning.

The fundamental point is that the cathode and anode alternately change their places, this helps to ensure the smooth operation of the circuit and greatly simplifies the repair, if it needs to be done.

Lamp device

In addition to the electronic ballast mounted in the base, a lamp is an important element of an energy-saving lighting device. It is she who is responsible for the uniformity of light distribution, its saturation, color rendition and other properties of the device. It is possible to conditionally divide the sections of the flask into lower and upper. In the upper one, special holes are made for installing the tube. The lower part contains a board in which the parts are located, and from which the leads from the tube extend.

The top area of \u200b\u200bthe board is equipped with wires that lead to the base. The lamp elements can be fastened to each other using special latches. In cheaper models, the parts are glued together. If you need to make repairs, you need to draw a screwdriver along the joint line or disconnect the latches.

How is the repair done

In order to determine which elements of the circuit or the lamp itself are faulty, it must be disassembled. To do this, detach the top from the bottom and turn off the flask. With the help of an Ohmmeter, we check the bulb glow coils. If it is found that one coil has burned out, the flask is repaired. It can be closed with an 8-10 ohm resistor. The resistor must be of high power. You will also need to remove the diode that shunts the burnt-out coil, if there is one in the circuit.

If a resistor burns out in lamps of 30 W or more, there is a high probability that the transistors are also out of order. This is due to capacitor breakdown. The situation can be corrected by installing a new fuse (resistor) and transistors.

In addition to replacing damaged circuit elements, you can upgrade the lamp. This is done by drilling ventilation holes in the plinth. Some models already have them, and if the manufacturers have not taken care of proper cooling of the electronics, you can do it yourself.

Attention! If you have drilled ventilation holes in the base of a 30 W lamp or lighting fixture of a different wattage, it cannot be used in rooms with high humidity. This can lead to breakdown in the capacitor and lamp failure.

The expediency of intervention in the schemes

It is possible to repair lamps of 30 W or other power only if you are confident in your abilities and knowledge. When you do not understand how the lamp circuit works, and what can break down in it, it is best not to try to fix the breakdown yourself.

It is forbidden to perform any actions with housekeepers if the integrity of their flasks is violated. The tube contains mercury or its vapors, therefore, when it is depressurized, the device becomes dangerous to human health and life.

Let's summarize

The circuits are almost the same in all models. Differences may be in the presence of diodes, shunt coils and other elements. However, if you know the electronics of one device, then working with all the others will be quite simple.

Schemes are often interested in people who want to independently repair broken lighting devices. It is not difficult to do this if you have the necessary skills and are sure that the housekeeper can be brought into working order.

The designation "energy-saving lamp" (EL) is more related to fluorescent compact lamps with a threaded base of any power (7, 20 W and above). Thanks to their more compact dimensions, the standard Edison base in the design and the absence of the need to use a remote control gear, such bulbs are more popular than linear designs of the same type.

The nuances of work and devices

Consists of several main units: built-in, gas-filled flask, base. The principle of EL functioning is based on a phenomenon called luminescence. The inner surface of the flask is coated with a phosphor. This substance can have a different composition, which will determine the quality of lighting and, accordingly, the intended purpose of the light source.

The device of such a lamp assumes the presence of two electrodes, which are installed in the tube. An arc discharge occurs between them under voltage. The flask contains low concentration of mercury and an inert gas.

Thanks to this content, low-temperature plasma is formed, which is further converted into UV radiation, invisible to human eyes. At this stage, the main role is played by the phosphor, which is covered from the inside of the bulb. This substance absorbs ultraviolet radiation, resulting in visible light from the lamp.

The circuit of an 11 W energy saving lamp is as follows:

In the figure, you can see the supply circuits driving the choke L2, the fuse F1, the filter capacitor C4 and the diode bridge (4 diodes 1N4007). A dinistor and elements D1, C2, R6 are involved in the launch. Protective functions are implemented through the elements R1, R3, D2, D3.

To turn on the lamp, it is necessary to ensure the opening of the transistor Q2, which occurs with the help of R6, C2, as well as a dinistor: these elements form a pulse. Blocking of this section of the circuit is performed with the participation of diode D1. The transformer is excited by means of transistors. The voltage comes from the boost resonant circuit (L1, C3, C6, TR1).

Types of energy saving lamps

The choice of a light source is made based on differences in shape, type of holder, power. The brand of the product also plays a role. Most popular manufacturers: Navigator, Philips, General Electric, Osram.

The EL device can be different, which is determined by the type of base:

• E14, E27, E40 - Edison base, thanks to which a light source of this type can be installed instead of analogs with a filament;
• pin holders (G53, 2 D, G23, G24Q1-G24Q3).

The following EL versions are distinguished by color temperature:

• with a warm white glow (2,700 K);
• with cold light (6 400 K);
• daylight source (4,200 K).

There are also different flasks: U-shaped, spiral-shaped, spherical and pear-shaped. Energy-saving light bulbs also differ in tube diameter: 7, 9, 12, 17 mm.

Overview of technical specifications

When choosing, you should take into account all the main parameters of light sources:

1. Power (from 7 to 105 W). For the home, it is recommended to choose versions no more than 20 W. The fact is that the EL luminous flux directly depends on the power: the higher the value of this parameter, the brighter the light. For comparison, a 100 W incandescent lamp and a 20 W compact fluorescent analogue produce the same luminous flux.
2. Base type. It is selected based on the characteristics of the lighting device in which the lamp will be installed.
3. Flask shape. This parameter does not affect the quality of work.
4. Colorful temperature. If the light source was chosen incorrectly, such a light will cause discomfort, regardless of the power (7, 20 W and above) and other parameters.

In addition, when choosing EL, it is necessary to pay attention to the service life. On average, a lamp of this type works for 6,000-12,000 hours.

Pros and cons of operation

The popularity of such light sources is due to a considerable number of advantages:

• a decrease in the level of energy consumption (by 80%), respectively, a 20 W lamp works no less efficiently than an analogue with a 100 W filament;
• longer term of work;
• low heating intensity;
• uniform light;
• a wide range of designs with excellent color temperature.

The disadvantages include the relatively high cost, the presence of substances hazardous to health in the flask, a decrease in efficiency at low temperatures, and a negative impact on the mechanism of frequent switching operations.

In addition, the electrical circuit of such a light source does not provide for the use of a dimmer.

Thus, energy-saving light bulbs are in many ways superior to other analogues (halogen and incandescent lamps). This is primarily due to a reduction in electricity costs, as a 20 W light source can replace the 100 W incandescent version.

Even compact fluorescent bulbs emit less thermal energy, are reliable and compact in size. The shape of the bulb does not affect the efficiency of the work, except that the cost differs: spiral versions are offered at a higher price.

In the previous sections, referenced above, both general technical characteristics and features of specific compact fluorescent lamps were considered. But the subjects leave testing in a random way, and the properties of CFLs from different manufacturers are somewhat different, which inadvertently raises interest in the internal structure of the lamps and a detailed study of their technology. This material is designed for a trained reader, so I apologize for possible difficulties with the perception of the material.

Electronic circuit

Converters for powering CFLs can be built according to various circuitry, from vibration transducers to ... However, you should not bother yourself with tricky words, almost all CFLs of the target range are made according to the same concept of one thousand nine hundred double year - on a resonant half-bridge oscillator. Many controllers for fluorescent lamps have been developed, with different functions and an extremely careful attitude to the lamp, but all this did not take root.

Cause? I don’t think that the money, microcircuits quickly lose in price with large-scale production. Then what is holding back the development of progress? Most likely, conservatism of thinking ("the circuit works, and no one complains"), and a lack of interest in improving the quality and operating time of the device. I think you have already formed your own opinion on this issue, and therefore I modestly shut up and pass on to “our rams”.

Most often, the converter is performed according to the following topology:

An incomplete diagram is presented here - there is no input filter that suppresses high-frequency interference, diodes to protect transistors from reverse voltage, and other trifles. In normal CFLs, these components are present, but we are talking about lamps in the budget segment, and therefore - what is, that is. In addition, excessive gates complicate circuit analysis. Complete options for building converters are easy.

The circuit can be divided into the following parts:

• Input rectifier unit (brown block) - rectifies and smooths the AC voltage of 220 volts, forms a constant voltage of about 280 volts to power the converter.
• Startup circuit (blue block) - starts the oscillator when the device is turned on.
• Power section (green block) - converts the rectified mains voltage into high frequency alternating voltage.
• Control transformer TV1.
• Bulb assembly (purple block, together with choke L1) - matches the output of the power section with the fluorescent lamp bulb.

Now in a little more detail. The circuit is really quite interesting, although it looks simple.

The mains voltage is rectified by a diode bridge and smoothed by an electrolytic capacitor ("C1" in the presented electrical diagram), the voltage from it ensures the operation of the power unit. It is fed to two switches (Q1 and Q2) on bipolar npn transistors, which convert it into an alternating voltage and transmit it to the matching node with the bulb.

The whole design of the electronic ballast is an auto-generator. The device operates at a certain frequency, which depends on the individual characteristics of a number of components. I'm not going to dissemble, it really is - the work of auto-oscillator circuits depends on the mass of characteristics and is extremely unstable. In a normal circuit, a dedicated controller controls the power switches, and the resulting operating characteristics (frequency, duty cycle) are directly determined from the conditions for the correct operation of the fluorescent lamp. Here we have a "dumb" autogenerator that just works and that's it. However, I got a little ahead of myself.

Let's forget about the lamp and the starting circuit for now, this is a separate conversation. The power section consists of two switches on transistors Q1 and Q2, controlled by the transformer TV1, the voltage shape of which is formed from the current passing through the bulb, the latter in turn depends on the frequency and voltage value from the output of the key transistors Q1 / Q2.

He loved her.
She ate a piece of meat
He killed her.
I buried it in a hole
And wrote the inscription,
What:
The priest had a dog
etc.

This is exactly how the autogenerator works, “on its own”, and this vicious circle cannot be broken. To establish such devices is easier to shoot at once, they either work right away or ... well, if they don't explode. The only way to understand the issue is to split the device into parts and analyze them independently. During debugging, this is done, the positive feedback circuit is turned off, and a signal from a separate generator is fed to the control transformer. If you are too lazy and go the simple way with "just turn on", then everything will end with a clap and search for the next pair of transistors. To reduce the risk, it is recommended to turn on the lamp through the LN (incandescent lamp), which will perform the functions of a fuse in case of "excess" in the electronics. Reception is very good, only it does not save from burnt transistors.

So, power transistors Q1 and Q2 open alternately, which is ensured by the polarity of the control transformer windings. If we assume that at the beginning of the windings (marked with a dot) at some moment a pulse of positive polarity acts, then a positive voltage will be supplied to the input of the transistor Q1, and a negative voltage will be applied to Q2. This means that transistor Q1 will be on, Q2 off, and a voltage level close to the supply voltage will be formed at the output (slightly less by the amount of the collector-emitter saturation voltage of Q1). If the control voltage changes sign, then the state of the transistors will change in the same way - Q1 will close, and Q2 will open, thus a low level will be set at the output, almost 0 volts.

This means that an alternating voltage is obtained at the output with the levels "zero" - "all power supply" and a period depending on the control signal, which is formed by the transformer TV1. The load current acts as a setting input value for it. If we simplify the power section as much as possible, then it will look like this:

A load consisting of an inductor L1 and a lamp (with a pair of capacitors and a PTC thermistor) is connected through the right winding of the transformer to the output of the key stage on transistors Q1 / Q2. This means that the current through the lamp is the quantity that determines the shape of the signal, which in turn turns on the transistors. I just want to add: "And the opening transistors form a voltage that causes a current, which ...", the circle is closed.

In this "circle" there must necessarily be an element that determines the operating frequency of the entire device, otherwise stable operation will be impossible. For CFL autogenerating ballast, such a key element is a resonant circuit consisting of a choke L1, a capacitor C4 and an equivalent lamp resistance - a classic version of an RLC circuit.

The resonant frequency for this construction depends not only on the values \u200b\u200bof the reactive components (L1 and C4), but also on the reduced active resistance of the lamp. The formula looks like this:

You can read more about the resonant circuit with series and parallel loading in WikipediA. I would like to note an important point - with a decrease in the nominal load resistance, the resonant frequency of the system decreases.

Such a construction of the circuit will ensure the operability of the lamp, but there can be no question of any stabilization - the device will always try to work at the resonant frequency with maximum efficiency. This is too bad, automatic adjustment should be introduced, but how? Install a current sensor, generate a reference voltage and process errors with an amplifier? A little more and you can reach a full PWM converter. It will be great, just stupid - chips for converters of fluorescent lamps have long been developed, duplicating them on transistors is an idiotic task. How to get out of the situation?

The complication of the scheme will lead to its inexpediency, and this despite the fact that such a construction "almost suits". And the solution was found (and a long time ago), it is successfully used in devices with a similar principle of operation. The idea is that the control transformer is not made with a conventional core made of a soft magnetic material (ferrite), but a material with a rectangular magnetization reversal hysteresis loop is used.

In order not to cast a shadow on the wattle fence, we will immediately proceed to the consequence of replacing the usual ferromagnetic material with a "special" one. The switching criterion is energy (which causes the magnetic field strength in the magnetic circuit). As soon as the energy exceeds the threshold, a switchover immediately follows. For this circuit, the accumulation measure is the number of turns of the primary winding of the transformer and the current through it. These characteristics are the limiting factor for adjusting the pulse frequency to maintain constant lamp current.

Indirectly, the use of a special magnetic core material is indicated by the ratio of the number of turns - for the normal operation of the "current transformer", the control current of the transistors should be about ten times less than the output current; you cannot drive the transistors into deep saturation. In this case, the primary winding consists of eight turns, and the "secondary" of three, which means a transformation ratio of 2.7 and is clearly less than the previously announced figure. The tuning of the characteristics of the converter is carried out not only by the number of turns, but also by the values \u200b\u200bof the resistors in the bases and emitters of the transistors.

Fortunately, we do not have to calculate or optimize the converter block, so I happily skip all this "dense forest". Let's note the main thing - the scheme somehow works, and it is definitely not worth getting into it, this is a design "in itself" and does not accept simple modernization.

Okay, we figured it out a little with the converter, but this oscillator can only work if it "already" generates. If there are no pulses, then there is no current through the control transformer and, as a result, there are no signals to open the transistors, the system "sleeps". To wake it up, a trigger circuit is used that generates a single pulse to open the lower transistor (Q2), which triggers the oscillator.

Let's go back to the original scheme. The trigger block is highlighted in a blue rectangle, it consists of resistors R1 and R2, diodes D1 and D2, capacitor C2. A relaxation generator is assembled on these elements, it works as follows: the capacitor C2 is charged with a small current through the resistor R1 to the breakdown voltage of the dinistor D2, usually about 30 volts. When D2 opens, capacitor C2 is discharged through the base of transistor Q2, which creates a start pulse for the CFL converter. After a very short time, the voltage across the capacitor decreases to a value at which the dinistor turns off and then the cycle repeats - the voltage across the capacitor will slowly rise again until the dinistor is turned on.

There is a trigger pulse, why do we need diode D1? The fact is that the relaxation generator will generate its pulses constantly. Although they are rare, they can coincide with the moment of the open state of the upper transistor, which will lead to an additional opening of the lower transistor. As a result, a large current pulse will appear through both open keys; such an incident can result in only one thing - the combustion of the circuit. Thus, after the converter enters the switching mode, the starting circuit must be blocked from repeated attempts to generate, which is done using the diode D1 - it discharges the capacitor C2 at those moments when the transistor Q2 is open.

There is still a resistor R2, and the point of using it is that it sets a nonzero voltage on the collector of the transistor Q2 (or rather, on the capacitor C3). Well, judge for yourself, what's the point of sending a triggering pulse to the base of the lower transistor, if the collector voltage is zero and its inclusion does not affect the state of other elements in any way. Resistor R2 ensures that the collector voltage "will be" before starting, that is its meaning.

By the way, usually such "fixing" resistors are put not one, but two: the first - as shown in the diagram, the second - from the collector Q2 to the "-" circuit of the power supply. A very large initial impulse is harmful for a half-bridge circuit, and the use of a pair of resistors allows the amplitude to be halved. However, these are trifles.

The next element that I want to draw your attention to is the lamp interface. It consists of capacitors C3 and C4, resistor R7 and the lamp itself. Let's forget about the PTC, capacitor C3 for a while and consider a simplified diagram of the lamp unit.

"V1" here refers to the square-wave voltage (meander) that creates the transducer assembly.

To begin with, let's define a simple question - what is a lamp? It is a sealed container with a small amount of mercury and filled with an inert gas. Two directly heated cathodes are installed at both ends of the lamp. By the way, its heating is not an obligatory function, there are varieties of cold cathode fluorescent lamps (CCFL). After the discharge occurs between the cathodes, a current arises that flows along the filament spiral, regardless of whether a voltage is applied to the filament terminals. This means that even with short-circuited filament leads, the filament will be hot. However, the questions of the cathode operation can be omitted for now, only two points are important regarding the steady-state operating mode:

• The glow is always hot, even if its leads are shorted.
• The lamp current flows through the filament.

Let's finish with the glow itself and turn our attention to the lamp bulb. It is usually made in the form of a thin tube, curled in a fancy way ("U" or "spiral"). A discharge is formed in its depths, which causes a glow that is so valuable to us. To obtain a discharge between the cathodes, a high voltage must be applied, which will cause a breakdown followed by a transition to a glow discharge. This mode is characterized by lower voltage and higher current. It is logical to assume that the lamp has two stable states - breakdown (high voltage, low current) and normal mode (lower voltage, relatively high current).

For now, let's leave this common assumption under the question mark and continue the thought further - what happens if the converter increases the voltage across the lamp? More voltage - more current through it, what other options? Let's make a simple check - let's see the current through the lamp. I do not present the picture, because of its obviousness - the shape of the current completely repeats the shape of the voltage applied to the lamp. Well, while it all fits together. But “alas,” careful reading of the documentation brings some dissonance. In particular, app. note # (THE L6569: A NEW HIGH VOLTAGE IC DRIVER FOR ELECTRONIC LAMP BALLAST) See Figure 15 below so you don't waste time studying the entire document.

It follows from this graph that as the current through the lamp increases, the voltage across it decreases. H'm. The dissonance intensifies. In the steady-state mode at a high frequency of the converter, the shape of the current through the lamp is characterized by a purely active form, without reactive components, and according to the long-term change in modes, the average current value is very nonlinear. A decrease in voltage with an increase in current indicates a negative internal resistance of the lamp, which clearly implies its tendency to self-excitement. However, the plasma in the lamp is already in some mode of the volumetric oscillatory process - you must have noticed various floating sprites in its body. It is very annoying that the graph in the figure is limited to such a small range, 0.1-0.23 amperes.

I will try to assume that with a decrease in the current, the trend will continue, but the question is - will it be monotonous? Building your own converter with adjustable characteristics has a very long history, you can do with a conventional CFL with an autogenerating converter, but with one addition - add a supply voltage regulator. The electronic circuit works quite adequately from 70 volts of alternating voltage, which allows you to change the lamp power several times.

It is troublesome to change the value of the alternating voltage, thyristor regulators are generally inapplicable, so I used a smooth voltage supply device that has been used in my room for a long time. Initially, the smooth voltage control unit was conceived to reduce the stress of switching on CFLs in the absence of preheating in them and to reduce the unpleasant effects of sudden switching on of light at night. The phase of switching on the lamp was removed (16 seconds, 452 KB), you can see. The voltage rises quite quickly, so I had to discharge the frames a bit.

I don’t know how it will seem to you, but I’m seeing several "jerks". If you look at the brightness at several points in the frame and average it, then it will change approximately as follows:

At the initial moment of time, a discharge occurs and the beginning of the glow of mercury vapor, therefore, the interval up to 200 ms is not interesting, and there is nothing unusual there. But after 230 ms, there is a sharp increase in intensity with a slight stabilization, followed by a second sharp jump in brightness. The supply voltage rises monotonously and rather linearly, this was verified during the development of the unit, and therefore a sharp change in properties seems strange. There are two obvious "breakthroughs" on this chart.

It would be possible to blame everything on the heating of mercury and the formation of vapors, but turning on the same lamp at a nominal supply voltage does not show any unusual phenomena. Wait, something similar has already happened somewhere ... In the first part of the article, we considered the case of turning on a cold fluorescent lamp, and one strange thing was observed on the graph, which I could not then explain.

Pay attention to the middle of the graph in green. Do you see anything similar?

The explanation for this phenomenon is simple, and I have already encountered it - plasma has several stable states. In ancient Soviet times, we developed a small-sized pocket TV, I was entrusted with the issue of backlighting. Complete data on the characteristics of that lamp have not been preserved, but I remember the approximate figures - the breakdown voltage is 800 volts, the lamp is in this mode up to 0.8 mA. With an increase in current above this threshold, the voltage drops sharply to about 200 volts, this state remains up to a current of 25 mA. With a further increase in the current, the voltage drops to 45 V and then remains almost unchanged.

Thus, the backlight converter could be built at 45 volts, but with the obligatory provision of overshoot of the “200 V” state. Or you can stay in the "200 V" combustion mode, but with the risk of falling into a low-voltage mode. The TV was powered by NKGTs-045 batteries, and therefore there was no way to get excess power, I had to limit myself to a not very stable, but low-power option. By the way, they tried a full-fledged version, with a flyback converter and energy storage in capacitors, but the design turned out to be inconvenient, and Soviet capacitors could not withstand operation at a nominal, but impulse voltage. We installed a conventional resonant oscillator, now such a solution is often used in CFLs powered by 12 volts. However, I digress, sorry.

The moral of this fable is that the plasma in the flask has "stable" states that it can "occupy". I'll try to suggest that not only "borrow", but also switch between them, as long as she has negative internal resistance.

To summarize this section - the equivalent resistance of the lamp in combustion mode can be represented as a resistor, only the value of this "resistor" can take on different values, depending on the amount of current through it.

Let's go back to the electronic ballast circuit. Suppose the circuit works, but how is the brightness of the glow maintained? Previously, it was suggested that the stabilizing function is performed by a special design of the control transformer, which changes the duration of the open state of the transistors, that is, the operating frequency. But the converter generates a rectangular voltage (more precisely, trapezoidal), and a sinusoidal voltage comes to the lamp.

The fact is that between the lamp and the converter there is a resonant circuit formed by a series choke and a parallel capacitor. These elements "absorb" the energy of the converter and form a sinusoidal voltage in the load (that is, the lamp), giving energy to it. Therefore, the shape of the "exciting" voltage is not important, the output will always be "sine". However, slight distortions of the shape are still present, the quality factor of the circuit is not too high.

Let's take some "averaged" parameters of reactive elements for the tested 15-25 W lamps and make a simulation. In this case, the equivalent resistance of the lamp will be of the order of 1 KΩ, which will allow using a number of load resistors and 1-2-4-8 KΩ as a characteristic of the system's operation in different combustion modes.

The upper figure shows the voltage across the tube, the lower figure shows the current through the resonant capacitor.

The simulator shows results comparable to theoretical calculations - as the value of the load resistor decreases, the resonant frequency also decreases, the voltage decreases, and the "resonant" rise becomes less in magnitude (the circuit quality factor decreases). If we exaggerate very much, then the case with a small load (8 KΩ, red graph) can be equated to the initial phase of switching on the lamp, it is characterized by a high voltage. Note, however, the current through the resonant capacitor (bottom picture). If the load is normal (1-2 KOhm, light green-blue graphs), then the current through it is relatively small. I did not begin to mark the current through the load resistance, so as not to clutter up the diagram. For these two cases, the current through the capacitor is less than through the load resistance. If the resistance rating is increased, then a large current begins to flow through the capacitor. And if we take into account that the voltage on the same capacitor increases greatly, then the reactive power will turn out to be simply enormous.

According to the simulation, 0.92 amperes and 1.1 kV, or 1 kV * A. The term "W" in this case is not applicable, the power is reactive, and therefore it is marked as "B * A". It is clear that a real converter in a CFL is not capable of producing such power, even for a short time, but the stressful conditions of operation are provided. Such a case (a small load) occurs at the moment the lamp is turned on, so it is not surprising that electronics so "like" to explode exactly at the moment of switching on. In solutions using microcircuits, this stress state is mitigated by frequency control, not allowing the operating frequency to be set strictly to the resonance threshold ("warm-up" mode), which increases the life of the entire device.

And here we note an extremely important point - if there is a high voltage on the lamp (at the time of the discharge), then this means an extremely large reactive power flowing through the resonant capacitor. It is clear that the same power circulates in the resonant choke, but they do not "die like flies" in CFLs, which is so "characteristic" of resonant capacitors.

Earlier, a simplified, but rather functional version of electronic ballast was considered. However, there is also an even cheaper version of the same scheme. The main nodes remain the same, the launch node is "simplified". If in the first variant a special element (dinistor) was responsible for the launch, the cost of which ... I don't know exactly how much one match costs? But when the instruction “to save at any cost!” Follows, then we, the buyers, reap the fruits of the creativity of “these comrades”. The scheme of such an execution looks something like this:

At first glance, the circuit has become somewhat simpler, the components have been removed from the central part.

The whole circuit is an amplifier with positive output-input feedback, and therefore it simply must generate, the problem is only in starting. In the previously considered version of the circuit, the node on the dinistor was responsible for this moment, but here it is absent. To start, the transistors are transferred from the key to the low-current linear operating mode. Namely, it turns out "as if" an ordinary amplifier, which cannot but be excited. To transfer the transistors to the amplifying mode, it is necessary to provide at least a small collector current at rest, which is carried out by installing a resistor R1 between the collector and the base of the transistor Q2.

The figure shows a "simplified" version of the circuit with autostart, but there is also a more "complete" version with the transfer of both transistors to the amplifying mode. However, it has a drawback - you have to install more parts, and therefore it is less common. Since the upper transistor (Q1) does not conduct current at rest, then a resistor must be added to the circuit to create that current. In this implementation, this function is performed by the resistor R2.

If we compare the first and second versions of the ballast, it can be noted that:

• The power components are the same, the difference appears only at the moment of starting.
• The variant with a dinistor is characterized by a clear threshold of the converter turn-on voltage.
• The autoplay option has not received any clear boundaries and, potentially, may never turn on. There may be problems with starting at low or high temperatures, aging of electronic ballast components. This method is less reliable - electrolytic capacitors have a clear tendency to "dry out" at high temperatures.

In short, the second option is clearly worse. And, interestingly, it is not necessarily cheaper - the dinistor is replaced with an electrolytic capacitor, and which of them costs less?

Schemes with autostart are marked in the products of the GamBiT trademark, so I told about the existence of such a circuit solution, and so ... unpleasant. As a hardware developer, I have an extremely negative attitude towards autogenerator "things" - they either work or not work, "that's all." And the autogenerator with auto start is already the limit. By the way, a similar circuit solution has already been used in series, remember AT computer power supplies (do not confuse with ATX!). In them, to start, both transistors in the half-bridge were switched to a weak active mode, which facilitated the onset of generation. One "but", after starting, voltage was applied to the control microcircuit, and it intercepted the control over the switching of transistors. There is also a clean auto-generator. Well, the most budgetary solution, nowhere else. And, of course, at the expense of quality.

Due to low power consumption, theoretical durability and price reduction, incandescent and energy-saving lamps are rapidly replacing. But, despite the declared service life of up to 25 years, they often burn out without even having served the warranty period.

Unlike incandescent bulbs, 90% of burned out LED bulbs can be successfully repaired by hand, even without special training. The examples provided will help you repair a failed LED lamp.

Before undertaking the repair of an LED lamp, you need to present its device. Regardless of the appearance and type of LEDs used, all LED lamps, including filament bulbs, have the same structure. If you remove the walls of the lamp housing, then inside you can see the driver, which is a printed circuit board with radio elements installed on it.

Any LED lamp is arranged and works as follows. The supply voltage from the contacts of the electric cartridge is applied to the terminals of the base. Two wires are soldered to it, through which voltage is applied to the driver input. From the driver, the DC supply voltage is supplied to the board on which the LEDs are soldered.

The driver is an electronic unit - a current generator that converts the supply voltage into the current required for the LEDs to glow.

Sometimes, to diffuse light or protect against human contact with the unprotected conductors of the board with LEDs, it is covered with a diffusing protective glass.

About filament lamps

In appearance, a filament lamp is similar to an incandescent lamp. The device of filament lamps differs from LED lamps in that they use not a board with LEDs as light emitters, but a glass sealed gas-filled bulb in which one or more filament rods are placed. The driver is located in the base.

The filament rod is a glass or sapphire tube with a diameter of about 2 mm and a length of about 30 mm, on which 28 miniature LEDs coated with a phosphor are fixed and connected in series. One filament consumes about 1 watt. My operating experience shows that filament lamps are much more reliable than those made with SMD LEDs. I guess over time they will replace all other artificial light sources.

Examples of repairing LED lamps

Attention, the electrical circuits of the LED lamp drivers are galvanically connected to the mains phase and therefore extreme care should be taken. Touching an unprotected part of the human body to the exposed parts of the circuit connected to the electrical network can cause serious damage to health, including cardiac arrest.

LED lamp repair
ASD LED-A60, 11 W on the SM2082 chip

Currently, powerful LED bulbs have appeared, the drivers of which are assembled on microcircuits such as SM2082. One of them worked for less than a year and came to me for repair. The light went out haphazardly and turned on again. When tapped on it, it responded with light or extinguishing. It became apparent that the problem was poor contact.

To get to the electronic part of the lamp, you need to pick up the diffusing glass with a knife at the point of contact with the body. Sometimes it is difficult to separate the glass, since silicone is applied to the fixing ring when it is seated.

After removing the light-scattering glass, access to the LEDs and a microcircuit - the SM2082 current generator was opened. In this lamp, one part of the driver was mounted on an aluminum LED PCB and the other on a separate one.

External examination did not reveal any defective rations or broken tracks. I had to remove the board with LEDs. To do this, the silicone was first cut off and the board was pried over the edge with a screwdriver blade.

To get to the driver located in the lamp housing, it was necessary to unsolder it, warming up two contacts with a soldering iron at the same time and moving it to the right.

On one side of the driver PCB, only a 6.8 μF 400 V electrolytic capacitor was installed.

A diode bridge and two series-connected resistors with a nominal value of 510 kOhm were installed on the back of the driver board.

In order to figure out which of the boards the contact is missing, they had to be connected, observing the polarity, using two wires. After tapping on the boards with a screwdriver pen, it became obvious that the malfunction lies in the board with the capacitor or in the contacts of the wires coming from the base of the LED lamp.

Since the soldering was not suspicious, I first checked the reliability of the contact in the central terminal of the base. It can be easily removed if you pry it over the edge with a knife blade. But the contact was reliable. Just in case, I tinned the wire with solder.

It is difficult to remove the screw part of the base, so I decided to solder the soldering wires from the base with a soldering iron. When touching one of the rations, the wire was exposed. There was a "cold" soldering. Since it was not possible to get to strip the wire, it was necessary to lubricate it with active flux "FIM" and then re-solder it.

Once assembled, the LED lamp emitted light steadily despite being hit by a screwdriver handle. Checking the light flux for pulsations showed that they are significant at a frequency of 100 Hz. Such an LED lamp can only be installed in luminaires for general lighting.

Driver wiring diagram
lED lamp ASD LED-A60 on a chip SM2082

The electrical circuit of the ASD LED-A60 lamp, thanks to the use of a specialized microcircuit SM2082 in the driver for stabilizing the current, turned out to be quite simple.

The driver circuit works as follows. The AC supply voltage is fed through fuse F to a rectifier diode bridge assembled on the MB6S microassembly. The electrolytic capacitor C1 smooths out the ripple, and R1 serves to discharge it when the power is turned off.

From the positive terminal of the capacitor, the supply voltage is applied directly to the series-connected LEDs. From the output of the last LED, the voltage is applied to the input (pin 1) of the SM2082 microcircuit, the current in the microcircuit is stabilized and then from its output (pin 2) is fed to the negative terminal of the capacitor C1.

Resistor R2 sets the amount of current flowing through the HL LEDs. The magnitude of the current is inversely proportional to its rating. If the value of the resistor is reduced, then the current will increase, if the value is increased, then the current will decrease. The SM2082 microcircuit allows the resistor to adjust the current value from 5 to 60 mA.

LED lamp repair
ASD LED-A60, 11W, 220V, E27

Another ASD LED-A60 LED lamp, similar in appearance and with the same technical characteristics as the repaired one, got into the repair.

When turned on, the lamp lit up for a moment and did not shine further. This behavior of LED lamps is usually associated with a driver malfunction. Therefore, I immediately proceeded to disassemble the lamp.

The light-scattering glass was removed with great difficulty, since along the entire line of contact with the body, despite the presence of a retainer, it was abundantly greased with silicone. To separate the glass, I had to look for a pliable place along the entire line of contact with the body with a knife, but still, there was a crack in the body.

To gain access to the lamp driver, the next step was to remove the LED printed circuit board, which was pressed along the contour into an aluminum insert. Despite the fact that the board was aluminum, and it was possible to remove it without fear of cracks, all attempts were unsuccessful. The board was kept tight.

It also did not work to remove the board together with the aluminum insert, since it fit snugly to the body and was seated with the outer surface on silicone.

I decided to try to remove the driver board from the side of the base. To do this, first, a knife was pried out of the base and the central contact was removed. To remove the threaded part of the base, it was necessary to slightly bend its upper flange so that the punching points disengage from the base.

The driver became available and freely moved out to a certain position, but it was not possible to completely remove it, although the conductors from the LED board were soldered.

There was a hole in the center of the LED board. I decided to try to remove the driver board by striking its end through a metal rod threaded through this hole. The board has advanced a few centimeters and bumped into something. After further blows, the lamp body cracked in the ring and the board with the base of the base separated.

As it turned out, the board had an extension, which with its shoulders rested against the lamp body. It looks like the board was shaped to restrict movement, although it was enough to fix it with a drop of silicone. Then the driver would be removed from either side of the lamp.

The voltage of 220 V from the lamp base through a resistor - the FU fuse is fed to the MB6F rectifier bridge and after it is smoothed by an electrolytic capacitor. Then the voltage goes to the SIC9553 chip, which stabilizes the current. Parallel connected resistors R20 and R80 between pins 1 and 8 of the MS set the value of the LED supply current.

The photo shows a typical electrical schematic diagram given by the manufacturer of the SIC9553 chip in a Chinese datasheet.

This photo shows the appearance of the LED lamp driver from the side of the output elements installation. Since space allowed, to reduce the ripple factor of the luminous flux, the capacitor at the driver output was soldered by 6.8 uF instead of 4.7 uF.

If you have to remove the drivers from the body of this lamp model and you cannot remove the LED board, then you can use a jigsaw to cut the lamp body around the circumference just above the screw part of the base.

In the end, all my efforts to extract the driver turned out to be useful only for understanding the LED lamp design. The driver appeared to be working properly.

The flash of the LEDs at the moment of switching on was caused by a breakdown in the crystal of one of them as a result of a voltage surge when the driver was started, which misled me. First of all, it was necessary to ring the LEDs.

An attempt to test the LEDs with a multimeter was unsuccessful. The LEDs were off. It turned out that two light-emitting crystals connected in series are installed in one case, and in order for the LED to start flowing current, it is necessary to apply a voltage of 8 V.

A multimeter or tester, included in the resistance measurement mode, produces a voltage within 3-4 V. I had to check the LEDs using a power supply, supplying 12 V to each LED through a 1 kΩ current-limiting resistor.

There was no replacement LED available, so instead the pads were shorted with a drop of solder. It is safe for the driver to work, and the power of the LED lamp will decrease by only 0.7 W, which is almost imperceptible.

After repairing the electrical part of the LED lamp, the cracked body was glued with fast-drying super glue "Moment", the seams were smoothed by melting the plastic with a soldering iron and leveled with sandpaper.

For fun, I did some measurements and calculations. The current flowing through the LEDs was 58 mA, the voltage was 8 V. Therefore, the power supplied to one LED is 0.46 W. With 16 LEDs, 7.36 W is obtained, instead of the declared 11 W. Perhaps the manufacturer indicated the total power consumption of the lamp, taking into account the losses in the driver.

The service life of the LED lamp ASD LED-A60, 11 W, 220 V, E27, declared by the manufacturer, raises my doubts. In a small volume of a plastic lamp housing, with a low thermal conductivity, significant power is released - 11 watts. As a result, LEDs and the driver operate at the maximum allowable temperature, which leads to accelerated degradation of their crystals and, as a result, to a sharp decrease in their MTBF.

LED lamp repair
LED smd B35 827 ERA, 7 W on a BP2831A chip

A friend shared with me that he bought five bulbs as in the photo below, and all of them stopped working after a month. He managed to throw out three of them, and he brought two, at my request, for repair.

The bulb worked, but instead of bright light, it emitted a flickering faint light with a frequency of several times per second. I immediately assumed that the electrolytic capacitor had swollen, usually if it fails, then the lamp begins to emit light, like a stroboscope.

The light-scattering glass was removed easily, it was not glued. It was fixed by means of a slot in its rim and a protrusion in the lamp body.

The driver was secured with two solders to a PCB with LEDs, as in one of the above lamps.

A typical driver circuit on the BP2831A microcircuit taken from the datasheet is shown in the photo. The driver board was removed and all simple radio elements were checked, everything turned out to be in good order. I had to start checking the LEDs.

The LEDs in the lamp were installed of an unknown type with two crystals in the case and inspection did not reveal any defects. Using the method of serial connection between the leads of each of the LEDs, I quickly identified the faulty one and replaced it with a drop of solder, as in the photo.

The light bulb worked for a week and was repaired again. Shorted out the next LED. A week later I had to short-circuit another LED, and after the fourth I threw out the light bulb, as I was tired of repairing it.

The reason for the failure of bulbs of this design is obvious. LEDs overheat due to insufficient heat sink surface, and their resource is reduced to hundreds of hours.

Why is it permissible to short-circuit the leads of burned-out LEDs in LED lamps

The driver of LED lamps, in contrast to the constant voltage power supply, outputs a stabilized current value at the output, not voltage. Therefore, regardless of the load resistance within the specified limits, the current will always be constant and, therefore, the voltage drop across each of the LEDs will remain the same.

Therefore, with a decrease in the number of LEDs connected in series in the circuit, the voltage at the driver output will also decrease proportionally.

For example, if 50 LEDs are connected in series to the driver, and a voltage of 3 V drops on each of them, then the voltage at the driver output was 150 V, and if 5 of them are short-circuited, the voltage will drop to 135 V, and the current will not change.

But the efficiency (efficiency) of the driver assembled according to this scheme will be low and the power loss will be more than 50%. For example, for an MR-16-2835-F27 LED bulb, you need a 6.1 kΩ resistor with a power of 4 watts. It turns out that the driver on the resistor will consume power that exceeds the power consumption of the LEDs and it will be unacceptable to place it in a small LED lamp housing, due to the release of more heat.

But if there is no other way to repair the LED lamp and it is very necessary, then the driver on the resistor can be placed in a separate case, all the same, the power consumption of such an LED lamp will be four times less than that of an incandescent lamp. It should be noted that the more LEDs connected in series in the bulb, the higher the efficiency will be. With 80 series-connected SMD3528 LEDs, you will need an 800 Ohm resistor with a power of only 0.5 W. The capacitor C1 will need to be increased to 4.7 µF.

Finding faulty LEDs

After removing the protective glass, it becomes possible to check the LEDs without peeling off the printed circuit board. First of all, a careful examination of each LED is carried out. If even the smallest black dot is found, not to mention the blackening of the entire surface of the LED, then it is definitely faulty.

When examining the appearance of LEDs, you need to carefully examine the quality of the rations of their conclusions. In one of the bulbs being repaired, there were four LEDs poorly soldered at once.

The photo shows a light bulb that had very small black dots on its four LEDs. I immediately marked the faulty LEDs with crosses so that they could be clearly seen.

Defective LEDs may or may not have a change in appearance. Therefore, it is necessary to check each LED with a multimeter or a pointer tester, included in the resistance measurement mode.

There are LED lamps, in which standard LEDs are installed in appearance, in the case of which two crystals connected in series are mounted at once. For example, lamps of the ASD LED-A60 series. For continuity of such LEDs, it is necessary to apply a voltage of more than 6 V to its terminals, and any multimeter outputs no more than 4 V. Therefore, such LEDs can be checked only by applying a voltage of more than 6 (9-12) V to them from a power source through a 1 kΩ resistor ...

The LED is checked, like a regular diode, in one direction the resistance should be equal to tens of megohms, and if you swap the probes (this changes the polarity of the voltage supply to the LED), then small, while the LED may glow dimly.

When checking and replacing LEDs, the lamp must be fixed. For this, you can use a suitable size round jar.

It is possible to check the health of the LED without an additional constant current source. But this verification method is possible if the light bulb driver is working properly. To do this, it is necessary to apply a supply voltage to the base of the LED light bulb and sequentially short-circuit the terminals of each LED with a jumper made of a wire or, for example, with jaws of metal tweezers.

If suddenly all the LEDs light up, it means that the shorted one is definitely faulty. This method is suitable if only one of the LEDs in the circuit is faulty. With this method of verification, it must be taken into account that if the driver does not provide galvanic isolation from the mains, as for example, in the above diagrams, then touching the LED solders with your hand is unsafe.

If one or even several LEDs are faulty and there is nothing to replace them, then you can simply short-circuit the contact pads to which the LEDs were soldered. The light bulb will work with the same success, only the luminous flux will decrease slightly.

Other malfunctions of LED lamps

If the check of the LEDs showed their serviceability, then the reason for the inoperability of the light bulb is in the driver or in the soldering points of the current-carrying conductors.

For example, a cold soldering conductor was found in this light bulb that supplies power to the printed circuit board. The soot released from poor soldering even settled on the conductive paths of the printed circuit board. The soot was easily removed by wiping with a cloth moistened with alcohol. The wire was soldered, stripped, tinned and re-soldered into the board. We were lucky with the repair of this light bulb.

Of the ten failed bulbs, only one had a faulty driver, a diode bridge collapsed. The driver repair consisted in replacing the diode bridge with four IN4007 diodes designed for a reverse voltage of 1000 V and a current of 1 A.

Soldering SMD LEDs

To replace a faulty LED, it must be evaporated without damaging the printed conductors. You also need to remove the replacement LED from the donor board without damage.

It is almost impossible to solder SMD LEDs with a simple soldering iron without damaging their case. But if you use a special tip for a soldering iron or put on a nozzle made of copper wire on a standard tip, then the problem is easily solved.

The LEDs are polarized and must be correctly installed on the PCB when replaced. Typically, printed conductors follow the shape of the LED leads. Therefore, you can only make a mistake with carelessness. To seal the LED, it is enough to install it on the printed circuit board and warm it up with a 10-15 W soldering iron with its ends with contact pads.

If the LED is burnt to charcoal and the printed circuit board underneath is charred, then before installing a new LED, it is imperative to clean this place of the printed circuit board from burning, since it is a current conductor. When cleaning, you may find that the soldering pads for the LED are burnt or peeled off.

In such a case, the LED can be installed by soldering it to adjacent LEDs if the printed tracks lead to them. To do this, you can take a piece of thin wire, bend it in half or three, depending on the distance between the LEDs, tin it and solder it to them.

Repair of LED lamp series "LL-CORN" (corn lamp)
E27 4.6W 36x5050SMD

The device of the lamp, which is popularly called the corn lamp, shown in the photo below is different from the lamp described above, therefore the repair technology is different.

The design of lamps on LED SMD of this type is very convenient for repair, since there is access for the continuity of the LEDs and their replacement without disassembling the lamp body. True, I disassembled the light bulb anyway for interest in order to study its structure.

Checking the LEDs of the LED corn lamp does not differ from the technology described above, but it must be taken into account that three LEDs are located at once in the SMD5050 LED case, usually connected in parallel (three dark dots of crystals are visible on the yellow circle), and all three should light up when checking.

A defective LED can be replaced with a new one or short-circuited with a jumper. This will not affect the reliability of the lamp, only imperceptibly to the eye, the luminous flux will decrease slightly.

The driver of this lamp is assembled according to the simplest scheme, without an isolation transformer, therefore, touching the LED terminals with the lamp on is unacceptable. Lamps of this design must not be installed in luminaires that can be accessed by children.

If all the LEDs are working, then the driver is faulty, and to get to it, the lamp will have to be disassembled.

To do this, you need to remove the bezel from the side opposite to the base. With a small screwdriver or a knife blade, you need to try in a circle to find the weak spot where the rim is worst glued. If the bezel gives in, then working with a tool, like a lever, the bezel will easily move away around the entire perimeter.

The driver was assembled according to the electrical circuit, like that of the MR-16 lamp, only C1 was 1 µF, and C2 was 4.7 µF. Due to the fact that the wires leading from the driver to the lamp base were long, the driver was easily removed from the lamp housing. After studying its circuit, the driver was inserted back into the case, and the bezel was glued in place with transparent glue "Moment". The failed LED is replaced with a good one.

Repair of LED lamp "LL-CORN" (corn lamp)
E27 12W 80x5050SMD

When repairing a more powerful lamp, 12 W, the same design of failed LEDs was not found and in order to get to the drivers, I had to open the lamp using the above technology.

This lamp gave me a surprise. The wires leading from the driver to the base turned out to be short, and it was impossible to remove the driver from the lamp housing for repair. I had to remove the base.

The lamp base was made of aluminum, nibbled around the circumference and held firmly. I had to drill the attachment points with a 1.5 mm drill. After that, the base, which had been pushed with a knife, was easily removed.

But you can do without drilling the base, if you pry and slightly bend its upper edge with the edge of a knife around the circumference. Beforehand, a mark should be made on the plinth and the casing so that the plinth can be conveniently installed in place. To securely fix the base after repairing the lamp, it will be enough to put it on the lamp body in such a way that the punched points on the base fall into the old places. Then push through these points with a sharp object.

Two wires were connected to the thread with a clamp, and the other two were pressed into the central contact of the base. I had to eat these wires.

As expected, the drivers were two identical, feeding 43 diodes each. They were closed with a heat-shrinkable tube and taped together. In order to fit the driver back into the tube, I usually cut it neatly along the PCB on the side where the parts are installed.

After repair, the driver is wrapped in a tube, which is fixed with a plastic tie or wrapped in several turns of thread.

In the electrical circuit of the driver of this lamp, protection elements are already installed, C1 for protection against impulse surges and R2, R3 for protection against current surges. When checking the elements, resistors R2 were immediately found on both drivers in the open circuit. It looks like an over-voltage was applied to the LED lamp. After replacing the resistors, there was no 10 Ohm at hand, and I set it to 5.1 Ohm, the lamp worked.

Repair of LED lamp series "LLB" LR-EW5N-5

The appearance of this type of light bulb inspires confidence. Aluminum body, high quality workmanship, beautiful design.

The design of the light bulb is such that disassembly is impossible without significant physical effort. Since the repair of any LED lamp begins with checking the health of the LEDs, the first thing that had to be done was to remove the plastic protective glass.

The glass was fixed without glue on a groove made in the radiator with a collar inside it. To remove the glass, you need to use the end of a screwdriver that will pass between the fins of the radiator, lean on the end of the radiator and lift the glass up like a lever.

Checking the LEDs with a tester showed their serviceability, therefore, the driver is faulty, and you need to get to it. The aluminum board was secured with four screws which I removed.

But contrary to expectations, behind the board was the heatsink plane, smeared with heat-conducting paste. The board had to be returned to its place and continued to disassemble the lamp from the base side.

Due to the fact that the plastic part, to which the radiator was attached, held very tightly, I decided to go the proven way, remove the base and remove the driver through the opened hole for repair. I drilled out the punching spots, but the base did not come off. It turned out that he still held on to the plastic due to the threaded connection.

I had to separate the plastic adapter from the radiator. He held on, just like the protective glass. For this, it was washed down with a hacksaw for metal at the junction of the plastic with the radiator and by turning a screwdriver with a wide blade, the parts were separated from each other.

After unsoldering the leads from the LED printed circuit board, the driver became available for repair. The driver circuit turned out to be more complex than previous bulbs, with an isolation transformer and a microcircuit. One of the 400 V 4.7 µF electrolytic capacitors was swollen. I had to replace him.

Checking all the semiconductor elements revealed a faulty D4 Schottky diode (pictured below, left). There was an SS110 Schottky diode on the board, replaced with an existing analog 10 BQ100 (100 V, 1 A). The forward resistance of Schottky diodes is half that of ordinary diodes. The LED light is on. The second light bulb had the same malfunction.

Repair of LED lamp series "LLB" LR-EW5N-3

This LED lamp is very similar in appearance to the "LLB" LR-EW5N-5, but the design is slightly different.

If you look closely, you can see that at the junction between the aluminum radiator and the spherical glass, unlike LR-EW5N-5, there is a ring in which the glass is fixed. To remove the protective glass, it is enough to pick it up with a small screwdriver at the junction with the ring.

The aluminum PCB contains three nine crystal super bright LEDs. The board is screwed to the heatsink with three screws. Checking the LEDs showed their serviceability. Hence, the driver needs to be repaired. Having experience in repairing a similar LED lamp "LLB" LR-EW5N-5, I did not unscrew the screws, but unsoldered the lead wires coming from the driver and continued to disassemble the lamp from the base side.

The plastic connecting ring of the base / plinth with the radiator was removed with great difficulty. At the same time, part of it broke off. As it turned out, it was screwed to the radiator with three self-tapping screws. The driver was easily removed from the lamp housing.

Self-tapping screws screwing the plastic ring of the base cover the driver, and it is difficult to see them, but they are on the same axis with the thread to which the transitional part of the radiator is screwed. Therefore, you can reach them with a thin Phillips screwdriver.

The driver was assembled according to the transformer circuit. Checking all the elements, except for the microcircuit, did not reveal those that failed. Therefore, the microcircuit is faulty, I did not even find a mention of its type on the Internet. The LED light bulb could not be repaired, it will come in handy for spare parts. But I studied her device.

Repair of LED lamp series "LL" GU10-3W

At first glance, it turned out to be impossible to disassemble a burnt-out GU10-3W LED bulb with protective glass. An attempt to remove the glass led to its chipping. When applied with great effort, the glass cracked.

By the way, in the marking of the lamp, the letter G means that the lamp has a pin base, the letter U, that the lamp belongs to the class of energy-saving bulbs, and the number 10 is the distance between the pins in millimeters.

LED bulbs with GU10 base have special pins and are rotatable in the socket. Thanks to the expanding pins, the LED lamp is pinched in the holder and is securely held even when shaken.

In order to disassemble this LED bulb, we had to drill a hole with a diameter of 2.5 mm in its aluminum case at the level of the printed circuit board surface. The drilling location must be chosen so that the drill does not damage the LED when exiting. If you don't have a drill at hand, you can make a hole with a thick awl.

Next, a small screwdriver is threaded into the hole and, acting like a lever, the glass is lifted. I removed glass from two bulbs without any problems. If the test of the LEDs with a tester showed their serviceability, then the printed circuit board is removed.

After separating the board from the lamp body, it immediately became obvious that the current-limiting resistors burned out both in one and in the other lamp. The calculator determined their nominal value by the bands, 160 ohms. Since the resistors burned out in LED bulbs of different batches, it is obvious that their power, judging by the size of 0.25 W, does not correspond to the power released when the driver is operating at the maximum ambient temperature.

The driver PCB was solidly sealed with silicone, and I didn’t detach it from the LED board. I cut off the leads of the burned out resistors at the base and soldered more powerful resistors to them, which were at hand. In one lamp, a 150 Ohm resistor with a power of 1 W was soldered, in the second two parallel 320 Ohm with a power of 0.5 W.

In order to prevent accidental touching of the resistor terminal, to which the mains voltage with the metal lamp housing is suitable, it was insulated with a drop of hot melt glue. It is waterproof, excellent insulator. I often use it for sealing, insulating and securing electrical wires and other parts.

Hot melt glue is available in 7, 12, 15 and 24 mm diameter rods in different colors, from transparent to black. It melts, depending on the brand, at a temperature of 80-150 °, which allows it to be melted using an electric soldering iron. It is enough to cut off a piece of the rod, place it in the right place and heat it up. The hot melt glue will acquire the consistency of May honey. After cooling, it becomes solid again. When reheated, it becomes liquid again.

After replacing the resistors, the performance of both bulbs was restored. It remains only to fix the PCB and protective glass in the lamp housing.

When repairing LED bulbs, I used "Montage" liquid nails to fix PCBs and plastic parts in place. Odorless glue, adheres well to the surfaces of any materials, remains plastic after drying, has sufficient heat resistance.

It is enough to take a small amount of glue on the end of a screwdriver and apply it to the contact points of the parts. After 15 minutes, the glue will already hold.

When gluing the printed circuit board, so as not to wait, holding the board in place, since the wires pushed it out, I fixed the board additionally at several points with hot glue.

The LED lamp started flashing like a strobe

I had to repair a couple of LED lamps with drivers assembled on a microcircuit, the malfunction of which was the blinking of light with a frequency of about one hertz, like in a stroboscope.

One copy of the LED lamp started blinking immediately after turning on for the first few seconds and then the lamp began to shine normally. Over time, the duration of the lamp blinking after turning on began to increase, and the lamp began to blink continuously. The second copy of the LED lamp began to flash continuously suddenly.

After disassembling the lamps, it turned out that the electrolytic capacitors installed immediately after the rectifier bridges in the drivers were out of order. It was easy to identify the fault as the capacitor housings were swollen. But even if in appearance the capacitor looks without external defects, then all the same, the repair of an LED light bulb with a stroboscopic effect must begin with its replacement.

After replacing the electrolytic capacitors with serviceable ones, the stroboscopic effect disappeared and the lamps began to shine normally.

Online calculators for determining the value of resistors
by color coding

When repairing LED lamps, it becomes necessary to determine the value of the resistor. According to the standard, marking of modern resistors is made by applying colored rings to their cases. 4 colored rings are applied to simple resistors, and 5 to high precision resistors.