DIY LED lamp repair: step-by-step guide

How to make a driver for LEDs

The circuits below use the most common elements that can be purchased at any radio store. No special equipment is required during assembly; all necessary tools are widely available. Despite this, with a careful approach, the devices work for quite a long time and are not much inferior to commercial samples.

Make it yourself

Connecting LED strip

LED lamps are rarely made again. It is easier to make a lamp from a faulty one. In fact, it turns out that repair and production of a new product is one process. To do this, the LED lamp is disassembled and the burnt-out LEDs and driver radio components are restored. There are often original lamps on sale with non-standard lamps, which are difficult to find replacements in the future. A simple driver can be taken from a faulty lamp, and LEDs from an old flashlight.

The driver circuit is assembled according to the classic model discussed above. Only resistor R3 is added to it to discharge capacitor C2 when turned off and a pair of zener diodes VD2, VD3 to bypass it in case of an open circuit of the LEDs. You can get by with one zener diode if you choose the right stabilization voltage. If you select a capacitor for voltages greater than 220 V, you can do without additional parts. But in this case, its dimensions will increase and after the repair is done, the board with the parts may not fit into the base.

LED lamp driver

The driver circuit is shown for a lamp of 20 LEDs. If their number is different, it is necessary to select a capacitance value for capacitor C1 such that a current of 20 mA passes through them.

The power supply circuit for an LED lamp is most often transformerless, and care should be taken when installing it yourself on a metal lamp so that there is no phase or zero short circuit to the housing.

Capacitors are selected according to the table, depending on the number of LEDs. They can be mounted on an aluminum plate in the amount of 20-30 pieces. To do this, holes are drilled in it, and LEDs are installed on hot-melt adhesive. They are soldered sequentially. All parts can be placed on a printed circuit board made of fiberglass. They are located on the side where there are no printed tracks, with the exception of LEDs. The latter are attached by soldering the pins on the board. Their length is about 5 mm. The device is then assembled in the luminaire.

LED table lamp

Required materials and tools

In order to assemble a homemade driver, you will need:

Soldering iron with a power of 25-40 W. You can use more power, but this increases the risk of overheating of the elements and their failure. It is best to use a soldering iron with a ceramic heater and a non-burning tip, because... a regular copper tip oxidizes quite quickly and has to be cleaned.
Flux for soldering (rosin, glycerin, FKET, etc.). It is advisable to use a neutral flux - unlike active fluxes (phosphoric and hydrochloric acids, zinc chloride, etc.), it does not oxidize the contacts over time and is less toxic. Regardless of the flux used, after assembling the device, it is better to wash it with alcohol. For active fluxes this procedure is mandatory, for neutral ones - to a lesser extent.
Solder. The most common is low-melting tin-lead solder POS-61. Lead-free solders are less harmful when inhaling fumes during soldering, but have a higher melting point with lower fluidity and a tendency to degrade the weld over time.
Small pliers for bending leads.
Wire cutters or side cutters for cutting long ends of leads and wires.
Installation wires are insulated. Stranded copper wires with a cross-section of 0.35 to 1 mm2 are best suited.
Multimeter for monitoring voltage at nodal points.
Electrical tape or heat shrink tubing.
A small prototype board made of fiberglass. A board measuring 60x40 mm will be sufficient.

PCB development board for quick installation

LED lamp device

DIY LED lamp repair

The lamp contains:

frame;
base;
diffuser;
radiator;
LED block;
transformerless driver.

220 volt LED lamp device

The figure shows a modern LED lamp using SOV technology. The LED is made as one unit, with many crystals. It does not require wiring of numerous contacts. It is enough to connect just one pair. When a lamp with a burnt-out LED is repaired, the entire lamp is replaced.

The shape of the lamps is round, cylindrical and others. Connection to the power supply is made through threaded or pin sockets.

For general lighting, luminaires with color temperatures of 2700K, 3500K and 5000K are selected. The spectrum gradations can be any. They are often used for advertising lighting and for decorative purposes.

The simplest driver circuit for powering a lamp from the mains is shown in the figure below. The number of parts here is minimal, due to the presence of one or two quenching resistors R1, R2 and the back-to-back connection of LEDs HL1, HL2. This way they protect each other from reverse voltage. In this case, the flickering frequency of the lamp increases to 100 Hz.

The simplest diagram for connecting an LED lamp to a 220 volt network

The supply voltage of 220 volts is supplied through the limiting capacitor C1 to the rectifier bridge, and then to the lamp. One of the LEDs can be replaced with a regular rectifier, but the flickering will change to 25 Hz, which will have a bad effect on vision.

The figure below shows a classic LED lamp power supply circuit. It is used in many models and can be removed for DIY repairs.

Classic scheme for connecting an LED lamp to a 220 V network

The electrolytic capacitor smooths out the rectified voltage, which eliminates flicker at a frequency of 100 Hz. Resistor R1 discharges the capacitor when the power is turned off.

Simple driver circuit for 1 W LED

One of the simplest circuits for powering a powerful LED is shown in the figure below:

As you can see, in addition to the LED, it includes only 4 elements: 2 transistors and 2 resistors.

The powerful n-channel field-effect transistor VT2 acts here as a regulator of the current passing through the LED. Resistor R2 determines the maximum current passing through the LED and also acts as a current sensor for transistor VT1 in the feedback circuit.

The more current passes through VT2, the greater the voltage drops across R2, accordingly VT1 opens and lowers the voltage at the gate of VT2, thereby reducing the LED current. In this way, stabilization of the output current is achieved.

The circuit is powered from a constant voltage source of 9 - 12 V, a current of at least 500 mA. The input voltage should be at least 1-2 V greater than the voltage drop across the LED.

Resistor R2 should dissipate 1-2 W of power, depending on the required current and supply voltage. Transistor VT2 is n-channel, designed for a current of at least 500 mA: IRF530, IRFZ48, IRFZ44N. VT1 – any low-power bipolar npn: 2N3904, 2N5088, 2N2222, BC547, etc. R1 - power 0.125 - 0.25 W with a resistance of 100 kOhm.

Due to the small number of elements, assembly can be carried out by hanging installation:

Another simple driver circuit based on the LM317 linear controlled voltage regulator:

Here the input voltage can be up to 35 V. The resistor resistance can be calculated using the formula:

R=1.2/I

where I is the current strength in amperes.

In this circuit, the LM317 will dissipate significant power given the large difference between the supply voltage and the LED drop. Therefore, it will have to be placed on a small radiator. The resistor must also be rated for at least 2 W.

This scheme is discussed more clearly in the following video:

Here we show how to connect a powerful LED using batteries with a voltage of about 8 V. When the voltage drop across the LED is about 6 V, the difference is small, and the chip does not heat up much, so you can do without a heatsink.

Please note that if there is a large difference between the supply voltage and the drop across the LED, it is necessary to place the microcircuit on a heat sink.

Repair of 220 V LED lamp

They brought a light bulb ( Fig. 1 ). They say that it burned out after not even working for four months, but before throwing it away, I would like to see how it works.

Fig.1

The first disassembly is not optimal, but it is complete

First, take out the light-diffusing panel. It is snapped into grooves by four small protrusions and you need to “go around in a circle” with a thin clock screwdriver, alternately prying the panel ( Fig. 2, 3 ).

Fig.2

Fig.3

The inscription “RED” on the plate with LEDs indicates that LEDs of different glow colors can be attached to it ( Fig. 4 ). The soldering points for the power conductors are labeled (GND - “minus” and VCC - “plus”).

Fig.4

The plate sits tightly, but can be removed if you pry it with a screwdriver through the through holes ( Fig. 5 ) and then a view of a small printed circuit board with electronic filling opens. The board is wrapped in an insulating film, very similar to lavsan ( Fig. 6 ). It is impossible not to notice that one of the conductors is painted light red (it goes to VCC, i.e. “positive” power supply) and that some latches are visible at the place where the conical part of the housing is attached to the cylindrical one. The plate itself is made of aluminum alloy ( Fig. 7 ).

Fig.5

Fig.6

Fig.7

The filling moves freely from side to side, but cannot be removed - it looks like it is held in place by short conductors, so we unsolder the wires from the plate with LEDs ( Fig. 8 ). If you then turn the conical part clockwise with a little force, the white plastic case is disconnected ( Fig. 9 ), but you still can’t get to the printed circuit board.

Fig.8

Fig.9

Let's see how the metal base is attached to the plastic - thin sheet metal is pinched in several places ( Fig. 10 ). To disassemble, you need to either heat the metal (plastic) and pull off the base, or drill out the depressed areas with a drill of a slightly larger diameter. Since the thought “what if it gets repaired” is already in our heads, we choose the second option as it is less painful, and then during assembly we can press the edges of the holes inward or make the same fastenings in new places.

Fig.10

We assemble the plastic case (this makes it easier to hold in the hand) and using a 1.2 mm drill, carefully drill through the tin so that there are no very deep recesses in the plastic ( Fig. 11 ).

Fig.11

Then we insert a thick and not very sharp knife blade into the junction of the base with the plastic ( Fig. 12 ) and “stir” the connection until it completely separates ( Fig. 13, 14 ).

Fig.12

Fig.13

Fig.14

It can be seen that the wire going from the printed circuit board to the threaded part of the base is pressed against it only mechanically (without soldering), and the wire going to the central contact of the base is really short and in order to inspect the electronic filling, it should be unsoldered from the side of the printed circuit board ( Fig. 15, 16, 17 ).

Fig.15

Fig.16

Fig.17

A circuit diagram was drawn from the printed circuit board ( Fig. 18 ) - it is very similar to the one found on the network at the request “LED driver SM7513” (available in the appendix to this text), but the filtering of the rectified high-voltage power supply is simplified and there are no protection elements against pulses of reverse polarity, standing parallel to the primary winding of the transformer.

Fig.18

The reason for the malfunction was found immediately - when the tester started “dialing”, a short circuit was detected in the V+/V- output power circuit. At first, suspicion fell on the VD US1D diode, but it turned out to be serviceable, and the ceramic capacitor C was short-circuited ( Fig. 19 ) - to get to it, you need to unsolder the converter transformer.

Fig.19

According to the description for the circuit from the network, this capacitor should have a capacity of 10 μF and an operating voltage of 16 V, but it was not possible to find exactly the same replacement, so 5.6 μF, 16 V was installed. After returning the transformer to the board and soldering the LED panel to it and the network cable, a test switch-on was carried out ( Fig. 20 ). Everything worked well and after about an hour the temperature conditions were checked - the aluminum plate was heating up (but this is how it should be), and the electronics on the board were slightly warm.

Fig.20

The wires were soldered off, the light bulb was assembled and screwed into a table lamp, where it worked for several more hours without any problems - no failures were noticed.

When I returned the light bulb to the delighted owner, it turned out that he had two more burnt ones in his stock ( Fig. 21 ). Well, well, let's take a look at them...

Fig.21

Description of incomplete disassembly (quick option)

Knowing the design features of the lamp, it is enough to drill holes in the base, remove it and then unsolder the conductor going to the central contact from the board. And the long power wires going to the LED panel allow you to remove the board with electronics. Now the tester can “ring” the LED power circuit and, if a short circuit is detected, find its cause and eliminate it. These two lamps also had "shorted" capacitors. One of them was replaced with 5.6 µF at 16 V, but since there were no more such capacitors, two capacitors were soldered into the third lamp - a ceramic 0.1 µF at 16 V and parallel to it, but on the other side of the board , electrolytic 10 uF at 16 V (unfortunately, there are no photographs, but from Figures 16 and 17 it is clear where to solder and that there is enough space in the base for this). After installing an additional capacitor, the overall size of the board increased slightly and the length of the old lavsan insulating tape was no longer enough, so the board had to be additionally wrapped with fluoroplastic tape of a suitable size.

Of course, the “breakdown” of the capacitors described here is not the only possible breakdown in the lamp; there are other elements in it, but since there are not many of them, the repair cannot be difficult. You just need some time to “ring” the diodes in the high-voltage rectifier bridge MB6S and US1D in the secondary circuit, check the integrity of the resistors, evaluate the appearance of the SM7513 microcircuit for case overheating (or even its destruction) and, of course, pay attention to electrolytic capacitor 4.7 uF 400 V (is it leaking or swollen). If you have a power supply with an adjustable output voltage, you can check the operation of the LED panel by supplying power to it through a resistor with a resistance of 1...10 Ohms.

When checking the lamp in a disassembled state, do not neglect the safety rules when working with a voltage of 220 V!

Andrey Goltsov, r9o-11, Iskitim, January 2018

Attached files:

Scheme for SM7513 from the network.rar (57 Kb)

Power driver circuit with PWM input

Below is a circuit for powering high-power LEDs:

The driver is built on a dual comparator LM393. The circuit itself is a buck-converter, that is, a pulse step-down voltage converter.

Driver Features

Supply voltage: 5 - 24 V, constant;
Output current: up to 1 A, adjustable;
Output power: up to 18 W;
Output short circuit protection;
The ability to control brightness using an external PWM signal (it will be interesting to read how to adjust the brightness of an LED strip using a dimmer).

Operating principle

Resistor R1 with diode D1 form a source of reference voltage of about 0.7 V, which is additionally regulated by variable resistor VR1. Resistors R10 and R11 serve as current sensors for the comparator. As soon as the voltage across them exceeds the reference one, the comparator will close, thus closing the pair of transistors Q1 and Q2, and they, in turn, will close the transistor Q3. However, inductor L1 at this moment tends to resume the flow of current, so the current will flow until the voltage at R10 and R11 becomes less than the reference voltage, and the comparator opens transistor Q3 again.

The pair of Q1 and Q2 acts as a buffer between the output of the comparator and the gate of Q3. This protects the circuit from false positives due to interference on the Q3 gate, and stabilizes its operation.

The second part of the comparator (IC1 2/2) is used for additional brightness control using PWM. To do this, the control signal is applied to the PWM input: when TTL logic levels (+5 and 0 V) are applied, the circuit will open and close Q3. The maximum signal frequency at the PWM input is about 2 KHz. This input can also be used to turn the device on and off using the remote control.

D3 is a Schottky diode, rated for current up to 1 A. If you cannot find a Schottky diode, you can use a pulse diode, for example FR107, but the output power will then decrease slightly.

The maximum output current is adjusted by selecting R2 and turning on or off R11. This way you can get the following values:

350 mA (1 W LED): R2=10K, R11 disabled,
700 mA (3 W): R2=10K, R11 connected, nominal 1 Ohm,
1A (5W): R2=2.7K, R11 connected, nominal 1 Ohm.

Within narrower limits, adjustment is made using a variable resistor and a PWM signal.

The nuances of connecting to a 220 V network

When connecting an LED to a 220V network, there are some features related to the amount of current passing. For example, in common backlit light switches, the LED is turned on according to the circuit shown below:

As you can see, there are no protective diodes here, and the resistor value is chosen in such a way as to limit the forward current of the LED to about 1 mA. The lamp load also serves as a current limiter. With this connection scheme, the LED will glow dimly, but enough to see the switch in the room at night. In addition, the reverse voltage will be applied mainly to the resistor when the switch is open, and the light-emitting diode will be protected from breakdown.

If you need to connect several LEDs to 220V, you can turn them on in series based on a circuit with a quenching capacitor:

In this case, all LEDs must be designed for the same current for uniform illumination.

You can replace the bypass diode with a back-to-back LED connection:

In both cases, it will be necessary to recalculate the capacitance value of the capacitor, because The voltage on the LEDs will increase.

Parallel (not back-to-back) connection of LEDs to the network is unacceptable, since if one circuit fails, double the current will flow through the other, which will cause the LEDs to burn out and a subsequent short circuit.

Several more options for the unacceptable connection of light-emitting diodes to a 220V network are described in this video:

Here's why you can't:

turn on the LED directly;
connect LEDs designed for different currents in series;
turn on LED without reverse voltage protection.

Assembling and configuring the driver

The driver components are mounted on a breadboard. First, the LM393 chip is installed, then the smallest components: capacitors, resistors, diodes. Then transistors are installed, and lastly a variable resistor.

It is better to place elements on the board in such a way as to minimize the distance between the connected pins and use as few wires as jumpers as possible.

When connecting, it is important to observe the polarity of the diodes and the pinout of the transistors, which can be found in the technical description for these components. You can also check diodes using a multimeter in resistance measurement mode: in the forward direction, the device will show a value of about 500-600 Ohms.

To power the circuit, you can use an external DC voltage source of 5-24 V or batteries. 6F22 (“crown”) and other batteries have too small a capacity, so their use is impractical when using high-power LEDs.

After assembly, you need to adjust the output current. To do this, LEDs are soldered to the output, and the VR1 engine is set to the lowest position according to the diagram (checked with a multimeter in the “testing” mode). Next, we apply the supply voltage to the input, and by rotating the VR1 knob we achieve the required brightness.

List of elements:

What are the types of drivers for LEDs by device type?

Drivers for LEDs are classified by device type into linear and pulsed. The structure and typical driver circuit for linear-type LEDs is a current generator on a transistor with a p-channel. Such devices provide smooth current stabilization under the condition of unstable voltage on the input channel. They are simple and cheap devices, but they are low efficient, generate a lot of heat during operation and cannot be used as drivers for high-power LEDs.

Pulse devices create a series of high-frequency pulses in the output channel. Their operation is based on the PWM (pulse width modulation) principle, when the average output current is determined by the duty cycle, i.e. the ratio of the pulse duration to the number of its repetitions. The change in the average output current occurs due to the fact that the pulse frequency remains unchanged, and the duty cycle varies from 10-80%.

Due to the high conversion efficiency (up to 95%) and compactness of the devices, they are widely used for portable LED designs. In addition, the efficiency of the devices has a positive effect on the duration of operation of autonomous power devices. Pulse-type converters are compact in size and have a wide range of input voltages. The disadvantage of these devices is the high level of electromagnetic interference.

LED driver efficiency reaches 95%

Before choosing a driver for LEDs, you need to know the conditions of its operation and the location of the LED devices. Pulse-width drivers, which are based on a single microcircuit, are miniature in size and are designed to be powered from autonomous low-voltage sources. The main application of these devices is car tuning and LED lighting. However, due to the use of a simplified electronic circuit, the quality of such converters is somewhat lower.