Saving coolness, or heat sink for powerful LED matrices

January 2, 2014

CreearticleLED

Thanks to their high luminous flux and long lifespan (on the order of tens of thousands, or even hundreds of thousands of hours), LED lamps are a very competitive solution. However, many suppliers and manufacturers of LED luminaires have difficulties when working with new high-power LEDs (from 20 W). And a particularly common problem is the design of correct and reliable heat dissipation. An incorrectly selected thermal operating mode of the LED can lead to undesirable consequences. First of all, overheating can lead to LED failure. All CREE LEDs have a critical transition temperature of 150°C; exceeding this threshold will lead to burnout of the LED crystal and a long repair process.

Secondly, operation at elevated temperatures significantly reduces the service life of LEDs (Figure 1). The graph shows the dependences for three temperatures at the “solder point” of the LED: 55, 85 and 105°C. The graphs marked LM-80 show the time during which the tests were performed. Graphs marked TM-21 show the decline in luminous flux from the original level as a function of time. As can be seen from the graphs, at elevated operating temperatures, the service life of LEDs is significantly reduced: at 105°C, the service life of LEDs is 200 thousand hours less than at a temperature of 85°C.

The following LED parameters also depend on temperature:

The amount of luminous flux. Figure 2 shows the dependence of the relative luminous flux on temperature for the MKR series LEDs from CREE. As can be seen from the graph, as the LED junction temperature increases, the luminous flux decreases, and vice versa - with good cooling, the flux increases.

Forward voltage drop. As the temperature changes, the forward voltage drop across the LED (Vf) also changes. As the temperature increases, the voltage decreases. The amount of voltage change depends on the specific model. Table 1 shows the coefficients of voltage versus temperature for the MKR and MKR2 series of LEDs. It is important to take into account the value of this parameter and select a driver for the lighting system so that it can provide the required voltage over the entire operating temperature range of the LED.

Table 1. Values ​​of coefficients of voltage versus temperature for the MKR and MKR2 series of LEDs

As can be seen from the graphs (Figures 1, 2), at temperatures below 100°C the luminous flux decreases slightly, and at a temperature of 85°C it is 100%. Recently, LEDs have been tested at a transition temperature of 85°C, so at temperatures below 85°C the graphs show an increase in luminous flux. This temperature will be considered the operating temperature for CREE LEDs.

Rice. 1. Lifetime of XPG LEDs, depending on temperature

Rice. 2. Dependence of luminous flux on transition temperature using the example of an MKR series LED

Now let’s describe the method for calculating and selecting a heat sink for high-power LEDs. An LED, like any other electronic device, is not 100% efficient, which means that part of the power consumed is converted into heat. Modern LEDs have an efficiency of about 30...40%, that is, on average 60...70% of power consumption is converted into heat. For example, when using a 20-watt LED matrix, 12 watts of heat must be dissipated, which is quite a lot. CREE in its “XLampThermalManagement” document recommends using the assumption that 75% of the power consumed is converted into heat, this assumption allows you to play it safe when designing a heat sink. The power that needs to be dissipated can be calculated using the formula:

(1)

Where:

Pt—thermal power (W);

Vf - forward voltage drop across the LED (V);

If is the current through the LED (A).

Before describing the methodology for calculating the cooling system, let's say a few words about the theory of heat transfer.

The main contributions to the cooling of LED lamps are thermal conduction and convection.

Thermal conduction is the process of transferring heat from a more heated body to a less heated one. In luminaires, due to direct contact, heat is transferred from the LED to the printed circuit board, and then to the radiator, or, if the LED is installed directly on the radiator, then directly to the radiator. To calculate the amount of heat transferred due to thermal conductivity, you can use the formula:

(2)

Where:

Qcond is the amount of heat transferred through the material (W);

k is the thermal conductivity coefficient of the material (W/(m*K));

A is the area of ​​intersection of materials through which heat passes (m2);

DT—temperature gradient (K);

Dx is the distance that heat travels (m).

Convection is transmission due to the movement of flows of liquids or gases. Usually in LED lamps this is the transfer of heat from the radiator to the environment (usually air). There are two options for convection: natural and forced. With natural convection, heat is transferred due to pre-existing air currents caused by temperature differences. In forced convection, the movement of liquid or gas flows is created by additional devices such as a fan, pump, etc.

The amount of heat dissipated by convection can be calculated using the formula:

(3)

Where:

Qconv is the amount of heat dissipated by convection (W);

h—heat transfer coefficient (W/(m2*K));

A is the surface area of ​​the radiating element (m2);

DT is the difference between the temperature of the radiating element and the ambient temperature (K).

The main problem in calculating the amount of heat dissipated by convection is determining the coefficient h. The value of the coefficient h can vary significantly, depending on the geometry of the radiator, boundary conditions and other parameters. For example, with natural convection, the coefficient h is in the range of 5...20 W/(m2*K). And for systems with forced convection, the heat transfer coefficient can reach values ​​of 100 W/(m2*K) with air cooling, and up to 1000 W/(m2*K) with liquid cooling. LED lighting usually uses natural air cooling; for calculations of such systems, the heat transfer coefficient can be taken equal to 10 W/(m2*K).

The LED cooling system can be represented as an equivalent circuit of thermal resistances connected in series and parallel. As an example to create an equivalent circuit, let's take a matrix of n LEDs mounted on a printed circuit board attached to a heatsink (Figure 3).

Rice. 3. Equivalent circuit of thermal resistances

In this case, the equivalent circuit will consist of n thermal resistances “LED junction - contact” (indicated in the diagram as Qj-sp), connected in parallel. Then - from n thermal resistances “contact - printed circuit board” (Qsp-pcb). It is also necessary to take into account the thermal resistances between the printed circuit board and the thermally conductive material (Qpcb-tim), between the thermally conductive material and the heatsink (Qtim-hs) and, finally, between the heatsink and the environment (Qhs-a).

At the nodes of this equivalent circuit, the temperature can be measured, for example, at the Theatsink point, the temperature of the radiator can be measured.

If the lighting device uses only one LED, the equivalent circuit will be a chain of thermal resistances connected in series. In turn, the thermal resistance of the entire cooling system is the sum of all thermal resistances. For a lamp consisting of one LED installed on a printed circuit board and on a radiator, the thermal resistance of all cooling systems is calculated using the following formula:

The lower the total thermal resistance value, the better the heat is dissipated from the LED. The thermal resistance between elements a and b is calculated by the formula:

(4)

Where:

Qa-b is the thermal resistance between elements a and b (°C/W);

Ta—temperature of element a (°C);

Tb—temperature of element b (°C);

Pt is power calculated using formula 1.

In the documentation for its LEDs, CREE offers a graph of maximum current versus temperature. An example of such a graph is shown in Figure 4. Knowing the maximum current and the estimated ambient temperature, you can calculate the value of the power that needs to be dissipated, and, accordingly, you can obtain the value of the maximum thermal resistance of the cooling system, which will allow you to select a radiator and heat-conducting materials.

Rice. 4. Dependence of maximum current on temperature for MKR LEDs

Let's take a closer look at how elements such as a printed circuit board, heat-conducting materials and a radiator contribute to the overall thermal resistance.

Printed circuit board. Most CREE LEDs must be mounted on a board (for LED power and mechanical wiring). The thermal resistance largely depends on the choice of PCB material and topology. For example, for standard FR4 boards the thermal resistance can be 20...80 °C/W, while for boards on a metal substrate the thermal resistance will be a few °C/W. CREE offers an "optimizing pcb thermal performance" guide for LED PCB design that provides recommendations for PCB layout to reduce thermal resistance. You can also use LEDs mounted directly on the radiator. In this case, the PCB will not contribute to the total thermal resistance.

Thermal conductive materials are required to create good thermal contact between the PCB and the heatsink or between the LED and the heatsink. In addition to providing reliable thermal contact, some thermally conductive materials, depending on the design of the cooling system, can perform other functions, such as insulating electrical components of the circuit or providing mechanical support. Table 2 below presents the characteristics of the main heat-conducting materials.

Table 2. Characteristics of heat-conducting materials

When choosing a heat-conducting material, it is necessary to take into account many parameters, not only the value of thermal conductivity. The thickness of the adhesive layer of the material is often overlooked, and as follows from formula (5) below, the thermal resistance directly depends on this parameter. Manufacturers of thermally conductive materials provide information about the main parameters in the documentation, and in order to correctly select a thermally conductive material, it is very important to understand the influence of each of these parameters on the operation of the cooling system. Sometimes a thinner adhesive layer with poor thermal conductivity value will have lower thermal resistance compared to a thicker layer with better thermal conductivity value. Both of these conditions must be taken into account when choosing materials. The thermal resistance of a heat-conducting material is described by the formula:

(5)

Where:

Qtim is the thermal resistance of the heat-conducting material (°C/W);

L—layer thickness (m);

K—thermal conductivity (W/m*K);

A is the contact area (m2).

The heatsink is perhaps the most important element in the LED cooling system; it removes heat from the PCB or directly from the LED, and dissipates the heat into the air. The following requirements are imposed on the radiator: the radiator material must have a high thermal conductivity value, the surface area of ​​the radiator must be maximum. In addition to cooling, a radiator can perform other functions; most often it can act as a housing or holder. Table 3 shows the thermal conductivities of some materials. Moreover, radiators made of the same material, but made using different surface treatment methods, may have different thermal conductivity coefficients. For example, an anodized aluminum radiator, due to radiation, has a higher thermal conductivity coefficient than a conventional aluminum radiator.

Table 3. Thermal conductivity of some materials

Often, LED lamps are subject to quite serious dimensional requirements, which may result in the need to design a radiator for specific requirements. When designing a radiator, it is necessary to take into account the weight of the final product, cost, thermal parameters, and the possibility of further production.

Typically cast or forged aluminum radiators are used. The anodized aluminum radiator has a high emissivity.

Designing a radiator can be a rather complex task, in which it is necessary to take into account dimensional restrictions, cost, weight, and the possibility of mass production. Below are some guidelines for radiator design:

• The surface area of ​​the radiator should be as large as possible;
• As a rough estimate, we can take the following assumption: for 1W of dissipated heat, a radiator with an area of ​​32...65 cm2 is required;
• for the correct location of the radiator, to ensure good air flow between its fins, it is necessary to have a good idea of ​​how the LED lamp will ultimately be mounted;
• a material with good thermal conductivity is required;
• use radiators with good emissivity. Anodizing dramatically increases the heat emissivity of an aluminum radiator;
• use programs to simulate cooling systems;
• select the radiator production method. Some methods of producing radiators may impose restrictions on the thickness and length of radiator fins and the materials used. The most common production methods: stamping, casting, forging. Each production method has its pros and cons.

Solving the cooling problem

Low-power LEDs, for example: 3528, 5050 and the like, give off heat due to their contacts, and the power of such specimens is much less.
When the power of the device increases, the question of removing excess heat arises. For this purpose, passive or active cooling systems are used. Passive cooling is a regular radiator made of copper or aluminum. There is debate about the benefits of cooling materials. The advantage of this type of cooling is the absence of noise and the almost complete absence of the need for maintenance.

Installation of LED with passive cooling in a spotlight

Active cooling is a cooling method that uses external force to improve heat dissipation. As the simplest system, we can consider a radiator + cooler combination. The advantage is that such a system can be much more compact than a passive one, up to 10 times. The disadvantage is the noise from the cooler and the need to lubricate it.

Purpose of radiators

The purpose of radiators is to remove heat from semiconductors, this helps reduce the influence of temperature on the operating parameters of devices. For this purpose, plate, fin, and pin radiators are used. The larger the radiator, the better it removes heat from radio components, and thus they heat up less.

To improve heat dissipation, it is best to attach the semiconductor device directly to the radiator. If electrical isolation of the semiconductor device from the chassis is necessary, the radiator is mounted on the chassis through insulating gaskets.

The thermal insulation capacity of a radiator depends on the degree of blackness of the material from which it is made. The more blackness, the more efficient the heat removal will be. Let's look at several types of different radiators to understand which radiator to choose for the semiconductor you need.

There are several technologies for making aluminum radiators. But mainly two technologies are used - casting and extrusion. The choice of radiator production is determined by the question: price - quality. Thus, it is cheaper to make finned radiators using the extrusion method, and needle radiators using the casting method.

For example, ribbed radiators will be cheaper and more efficient when produced by extrusion, but needle-shaped radiators can only be produced by casting.

How to fix the LED

There are two main methods of fastening, let's consider both of them.

The first method is mechanical. It consists of screwing the LED with self-tapping screws or other fasteners to the radiator; for this you need a special “star” type substrate (see star). A diode, pre-lubricated with thermal paste, is soldered to it.

On the “belly” of the LED there is a special contact patch with the diameter of a slim cigarette. After that, power wires are soldered to this substrate, and it is screwed to the radiator. Some LEDs go on sale already mounted on an adapter plate, as in the photo.

The second method is glue. It is suitable for mounting through a plate or without it. But it’s not always possible to attach metal to metal; how to glue an LED to a radiator? To do this, you need to purchase a special thermal conductive glue. It can be found both in hardware stores and in radio parts stores.

The result of such fastening looks like this:

DIY cooling

The simplest example of a radiator is a “sun” cut out of tin or aluminum sheet. Such a radiator can cool 1-3W LEDs. By twisting two such sheets together through thermal paste, you can increase the heat transfer area.

This is a banal radiator made from improvised means; it turns out to be quite thin and cannot be used for more serious lamps.

It will be impossible to make a radiator for a 10W LED with your own hands this way. Therefore, you can use a radiator from the computer’s central processor for such powerful light sources.

If you leave the cooler, active cooling of the LEDs will allow you to use more powerful LEDs. This solution will create additional noise from the fan and require additional power, plus periodic maintenance of the cooler.

The radiator area for a 10W LED will be quite large - about 300cm2. A good solution would be to use finished aluminum products. You can purchase an aluminum profile at a hardware or hardware store and use it to cool high-power LEDs.

By assembling the required area from such profiles, you can get good cooling; do not forget to coat all joints with at least a thin layer of thermal paste. It is worth saying that there is a special profile for cooling, which is commercially produced in a wide variety of types.

If you do not have the opportunity to make an LED cooling radiator with your own hands, you can look for suitable copies in old electronic equipment, even in a computer. There are several located on the motherboard. They are needed for cooling chipsets and power switches of power circuits. An excellent example of such a solution is shown in the photo below. Their area is usually from 20 to 60 cm2. Which allows you to cool a 1-3 W LED.

Another interesting option for making a radiator from aluminum sheets. This method will allow you to gain almost any required cooling area. Watch the video:

How to attach semiconductors to a heatsink

Semiconductors are attached to the radiator using special flanges. If it is necessary to isolate radioelements from the radiator, insulating gaskets must be used. So, of course, the efficiency will decrease, but when using different microcircuits and transistors this is necessary. Another option is to isolate the radiator itself from the board, so as not to use insulating gaskets.

The surface at the point of contact with the radiator must be smooth and clean for better heat transfer. You can also use special thermal pastes to increase efficiency and reduce heat transfer resistance. If you have transistors in a protected housing, then they can be placed on the radiator without gaskets. But you shouldn’t forget about thermal paste.

Aluminum fixtures

The most popular LED heat sink is made of aluminum. The main disadvantage of the device is that it consists of a number of layers. This inevitably causes transient thermal resistance, which can be overcome with the help of additional heat-conducting materials: adhesive substances, insulating plates, materials for filling air gaps.

Aluminum radiator for LEDs is used more often than others. It is subject to pressing and does an excellent job of dissipating heat.

An active level of cooling typically requires a flat sheet of aluminum no larger than the size of the luminaire. The sheet is blown by a fan.

The suitable temperature for LED operation is 65 °C. However, the lower the temperature, the higher the efficiency level of the device and the longer its resource. The optimal temperature of the radiator surface is considered to be 45 °C, but not higher. For a 1 W diode, it must be installed on an aluminum radiator. The radiator area is 30-35 cm2. A 3 W LED radiator will require a doubling of area and will be 60-70 cm2.

An aluminum device is best suited as a radiator as it is the lightest and relatively inexpensive. When calculating a device for LED matrices, a proportion of 35 cm per 1 W is taken.

For active cooling systems, the radiator area can be 10 times smaller. A 1 W LED is enough for 3-3.5 cm2.

For example, consider a star radiator for LEDs. The device is used to remove heat from the LED and is a small radiator. It is based on a plate made of a composite material - aluminum is used, which removes heat from the LED, and copper foil with contact pads. The radiator is mounted on LEDs with a high power rating (1-3 W).

LED lamps with cooling radiator: types of radiators

LED lamps with a cooling radiator: types of radiators
LED lamps have firmly entered our lives, almost completely replacing incandescent lamps and energy-saving compact fluorescent lamps (CFLs). It can be assumed that the whole point is their cost-effectiveness, excellent technical characteristics (such as luminous flux, CRI, scattering angle), as well as their long service life.

Why do you need a radiator in an LED lamp?

The service life of the product is primarily influenced by the quality of the LEDs, as well as the driver, the correct operation of which directly affects the stability of the diodes.

However, during the operation of an LED lamp, its surface becomes dirty, which negatively affects the removal of the generated heat. Over time, the problem of overheating appears, which is associated with a decrease in the light output of the diodes until they fail.

To avoid this, increase the stability of the light sources. For this purpose, a radiator is provided in the design of each of them.

Types of radiators

A radiator is a structural element that serves to remove and dissipate heat from LEDs.

LED lamps with radiator cooling

LED lamps with a cooling radiator come in the following types:

• with aluminum radiator;
• ceramic;
• composite;
• plastic.

LED lamps with aluminum heatsink

These lamps belong to the standard or high class. The aluminum radiator in such products can be either a strip of metal or a structurally more complex aluminum base. Hence the division of such lamps into two types:

1. with finned radiator;
2. with a flat radiator.

LED lamps with finned aluminum heatsink

The most effectively protected lamps, the radiator of which is presented in the form of a multilayer structure with ventilation ducts. Due to them, the area of ​​heat dissipation increases, which significantly increases the service life of LEDs, and also prevents their degradation over time due to overheating.

LED lamps with aluminum heatsink

Lamps with flat radiator

A flat radiator is less efficient than a finned one. Such a cooling element is used mainly in low-power lamps. Often, for more efficient heat removal, it has ventilation channels, and its dielectric surface is covered with a layer of special paint or varnish.

Composite radiator

LED lamps with a cooling radiator made of composite material are distinguished primarily by their affordable price. In such lamps, the element is a two-layer structure made of an aluminum strip coated with heat-conducting plastic.

Due to their low price, lamps with a composite radiator are the most widely represented on the market in the economy class segment.

However, such radiators cannot effectively remove heat, so the warranty period for products with them rarely exceeds 1 year.

Plastic radiator

The simplest option, it would be more correct to call it an imitation of a radiator. The element is a housing made of heat-dissipating plastic. The main differences between such lamps: low price, short warranty period, short service life (10,000-15,000 hours). In high-power lamps, to increase heat dissipation, the plastic radiator is made with additional massive fins and ventilation holes.

LED lamps with plastic radiator

Ceramic radiator

LED lamps with a ceramic cooling radiator are distinguished by high heat resistance, and the dielectric properties of the material allow LED modules to be mounted directly on the surface of such a radiator. The most common type of lamp with a ceramic radiator without a diffuser bulb is the so-called corn lamp.

LED lamps with ceramic cooling radiator

LED lamps with which cooling radiator to choose?

From all of the above, we can conclude that the quality of any LED lamp depends, among other things, on the quality of the radiator, or rather on the material from which it is made.

The most reliable, with a long actual service life, are considered to be LED lamps with an aluminum cooling radiator, as well as with a ceramic one (if such a light source does not have a diffuser bulb).

Plastic options can be preferred if such lamps will be used only occasionally and for a short time, for example, in closets and utility rooms.

Existing types of radiators

Radiators for LEDs
Cooling devices are divided according to design features into 3 main types and can have a round, square or rectangular shape, regardless of whether it is a plate, rod or ribbed radiator.

When choosing a cooler or making it yourself, you need to pay special attention to the thickness of its base, because it will take on the main heat, which will then be evenly distributed over other parts of the radiator.

The choice of the shape of the cooling device is influenced by the design of the future device itself, namely how it will be cooled, whether the ventilation will be forced or natural.

The distance between the plates depends on this. Provided there is no forced ventilation, it cannot be less than 4 millimeters. If the condition is not met, then such a cooling device will be of no use.

But the shape does not matter for cooling. An example would be LED lamps. Designers probably have to work hard to come up with an option in which the heat sink will not extend beyond the size and shape of the light bulb itself, will not spoil the appearance, and will still do its job effectively. Sometimes in such cases the cooling device is attached with a special heat-conducting adhesive directly to the printed circuit board.

Cooling features of high-power LEDs

As mentioned earlier, effective heat removal from the LED can be achieved by organizing passive or active cooling. It is advisable to install LEDs with a power consumption of up to 10 W on aluminum (copper) radiators, since their weight and size indicators will have acceptable values.

The use of passive cooling for LED arrays with a power of 50 W or more becomes difficult; The dimensions of the radiator will be tens of centimeters, and the weight will increase to 200-500 grams. In this case, it is worth considering using a compact radiator together with a small fan. This tandem will reduce the weight and size of the cooling system, but will create additional difficulties. The fan must be provided with the appropriate supply voltage, and care must also be taken to protect the LED lamp in the event of a cooler failure.

There is another way to cool powerful LED matrices. It consists of using a ready-made SynJet module, which looks like a cooler for a medium-performance video card. The SynJet module features high performance, a thermal resistance of no more than 2 °C/W and a weight of up to 150 g. Its exact dimensions and weight depend on the specific model. Disadvantages include the need for a power source and high cost. As a result, it turns out that a 50 W LED matrix must be mounted either on a bulky but cheap radiator, or on a small radiator with a fan, power supply and protection system.

Whatever the heatsink, it can provide good, but not the best, thermal contact with the LED substrate. To reduce thermal resistance, heat-conducting paste is applied to the contacted surface. The effectiveness of its impact has been proven by its widespread use in cooling systems for computer processors. High-quality thermal paste is resistant to hardening and has low viscosity. When applied to a radiator (substrate), one thin, even layer over the entire contact area is sufficient. After pressing and fixing, the layer thickness will be about 0.1 mm.

Finned radiator

Ribbed (plate) radiator. The area of ​​these radiators is equal to the sum of the areas of all sides. The larger the surface area, the higher the heat transfer efficiency. To improve cooling of the elements, it is necessary to increase the radiator area. And in order not to increase the size, since this is not always possible due to the design of the device, radiators are made with ribs and micro-relief on them. The number of fins and their sizes significantly increase the area of ​​the radiators.

Also, radiators with fins differ in the distance between the fins. On radiators where the distance between them is small, you need to install forced cooling. Such radiators are found in computers to cool processors. Radiators with a large distance between the fins are used without fans. Although the former can also be used without forced cooling. Here you can see how efficiently they will work with and without cooling.

An improved plate heat sink is a set of several plates bent in different directions. This radiator, with a surface area equal to the simplest plate one, has smaller dimensions. This heat sink is installed similarly to a plate heat sink. The number of plates may vary depending on the required surface. The dissipation area of ​​such a radiator is equal to the sum of the areas of all curved sections of the plates, plus the surface area of ​​the central part. This type of radiator also has disadvantages: reduced efficiency of heat removal from all plates, as well as the impossibility of obtaining a perfectly straight surface at the junction of the plates.

HOW TO CALCULATE A RADIATOR FOR AN LED

The calculation of a radiator for an LED is carried out not by the surface area, but by the useful dispersion area. The larger it is, the more intensely the device will transfer heat to the air. It is also necessary to take into account the power input. If the LED is used at full power, then it will need more cooling

It is equally important to consider where the device will be located: outdoors or indoors

The professional calculation method takes into account several important factors:

• ambient air indicators;
• radiator modification;
• heat sink material;
• dispersion area.

But such characteristics are usually taken into account by designers who develop a heat sink. At home, you can use a simpler formula. It involves calculating the maximum power dissipation of the heat exchanger.

Ф = а · S · (Т1 – Т2),

where F is the magnitude of the heat flow, S is the surface area of ​​the radiator (all heat-removing surfaces), T1 and T2 are the temperature of the medium that removes heat and the temperature of the heated surface, respectively, and is the heat transfer coefficient (conventionally assumed to be 6-8 W/m2 K) .

When calculating the surface area of ​​a heat sink, the following must be taken into account:

• Plate and fin radiators have 2 surfaces for heat removal, so in the formula it will not be S, but 2S.
• For needle radiators, the heat sink surface area is determined as the circumference (π · D) multiplied by the height (H).

There is a simpler formula for calculating the radiator area for an LED, which is popular among Internet users as experimental. It is applicable for aluminum radiators and looks like this:

Soх = (22 – (M 1.5) W,

where Sox is the area of ​​the cooler, M is the unused power of the LED (W), W is the supplied power (W). The area obtained from the formula is sufficient for natural cooling of the LED without the use of a cooler. Using the formula for calculating a copper radiator, the area must be reduced by 2 times.

You can avoid calculating the LED cooling radiator, but use average data that reflects the dependence of area on power. For aluminum radiators, the following values ​​are relevant:

• 1 W – 10-15 cm2;
• 3 W – 30-50 cm2;
• 10 W – 1000 cm2;
• 60 W – 7000-7300 cm2.

The indicated area of ​​​​the LED heatsink has a fairly large scatter, so the data is considered approximate, which must be taken into account when choosing a suitable device

Pin radiator

A pin (needle) heat sink is a very effective heat sink for semiconductor devices. They are made of duralumin with a thickness of 4-6 mm and aluminum wire with a diameter of 3-5 mm. On the surface of the pre-processed radiator plate, you need to mark with a center punch the locations for the holes for the pins, terminals of transistors, diodes and mounting screws.

The distance between centers should be equal to 2-2.5 times the diameter of the aluminum wire used. The diameter of the holes for the pins is chosen so that the wire fits into them with the smallest possible gap. On the reverse side, the holes for the pins are countersunk to a depth of 1-1.5 mm. With a minimum volume, such a radiator has a maximum effective dissipation area. The surface area of ​​such a heat sink is equal to the sum of the areas of each pin and the area of ​​the main body.

You can also buy ready-made radiators both in the store and on the market. I buy them at the market - flea market, where they cost five times cheaper than in the store. There is a large selection and you will not have any problems which radiator to choose for your craft.

How to connect a HPS lamp

Here is a compact shield assembled with your own hands, according to the connection diagram.

You can, of course, assemble all this in the overall body of the lamp, if the dimensions allow.

It is very important, before assembling such a circuit yourself and using any components, using a conventional multimeter in the maximum resistance measurement mode, check the insulation of the inductor and capacitor. Is there a hole in the body? Is there a hole in the body?

Is there a hole in the body?

To supply and disconnect 220V power, use a two-pole input circuit breaker.

For one lamp with a power of up to 400W, a machine with a nominal value of 5-6A is quite suitable. In addition to on-off switching operations, it will also play the role of a protective device.

A circuit breaker is mounted at the very beginning of the circuit. Do not forget to also ground the entire panel body.

Two neutral wires come out of the machine. According to the diagram, one of them is connected directly to the lamp, and the second one is connected to the corresponding terminal labeled “N” on the starter.

Keep in mind that the choke must be installed only in the open phase wire going to the lamp, and not in the neutral wire.

Otherwise, you can accidentally burn the product if during operation the neutral wire after the ballast choke accidentally shorts out. Next, turn off the phase. Mount one wire from the machine to the incoming contact of the throttle.

And connect the wire from the output contact to terminal “B” (Balast) of the ballast.

After that, connect the middle terminal Lp (Lampa) to the light bulb socket.

IZU switching circuits

Let's consider the parallel launch scheme of the IZU. In such a circuit, the lamp current does not pass directly through the IZU, which virtually eliminates any power loss. The circuit of the ignition device for such inclusion is quite simple, the devices themselves are inexpensive, easy to operate and quite reliable. However, high-frequency pulses generated by the ignition device in such a circuit affect, in addition to the lamp, also the choke, which necessitates the use of chokes with increased insulation, resistant to voltages of 2–5 kV.

Since standard chokes for metal halide and sodium lamps do not support this voltage value, the parallel IZU switching circuit is used only with lamps whose ignition voltage is less than 2 kV. First of all, such lamps include high-power metal halide lamps (from 2000 to 3500 W).

We calculate the radiator area

Please note that to correctly calculate the radiator area, the parameters of the useful dissipation area, and not the surface area, are taken into account. When calculating the usable area (S), the sum of the areas of the ribs and the substrate in square meters is included

It should be taken into account that each rib has two outlet surfaces. In this case, S of a rectangular heat sink S - 1 cm2 is - 2 cm2

When calculating the usable area (S), the sum of the areas of the ribs and the substrate in square meters is included. It should be taken into account that each rib has two outlet surfaces. In this case, S of a rectangular heat sink S - 1 cm2 is - 2 cm2.

As a result of the experiments, a formula was derived for calculating the required heat sink area:

S = (22 – (M x 1.5)) x W, in which

S – radiator heat sink area; W – power supplied (W); M – LED power. For plate radiators made of aluminum, the following approximate data calculated by experts from Taiwan can be used:

• 1 W: 10 ÷ 15 cm2;
• 3 W: 30 ÷ 50 cm2;
• 10 W: approximately 1000 cm2;
• 60 W: 7000 73000 cm2.

Since the range of the indicated data has a wide range and they were determined in conditions for the climate of a southern country, the values ​​are not absolutely accurate and are suitable for preliminary calculations.

More detailed information on calculating the radiator area can be obtained by watching the video.

How to fix the LED

There are two main methods of fastening, let's consider both of them.

First way

- it's mechanical. It consists of screwing the LED with self-tapping screws or other fasteners to the radiator; for this you need a special “star” type substrate (see star). A diode, pre-lubricated with thermal paste, is soldered to it.

On the “belly” of the LED there is a special contact patch with the diameter of a slim cigarette. After that, power wires are soldered to this substrate, and it is screwed to the radiator. Some LEDs go on sale already mounted on an adapter plate, as in the photo.

Second way

- it's adhesive. It is suitable for mounting through a plate or without it. But it’s not always possible to attach metal to metal; how to glue an LED to a radiator? To do this, you need to purchase a special thermal conductive glue. It can be found both in hardware stores and in radio parts stores.

The result of such fastening looks like this:

DIY: Water Cooled Powerful LED

• Jurei-678
• October 6, 2016
• Homemade productsLight

We have a new addition to the section of useful homemade crafts: a powerful DIY water-cooled LED. Hi all! Sometimes you want to build a powerful LED lamp, but there is no suitable radiator or they are expensive and bulky.

Today I will show you how to cool a powerful 10-30 watt LED with a three by three centimeter radiator. We take the radiator and glue the LED onto it; as soon as the glue has dried, we apply colorless silicone to the LED and glue a piece of glass or lens onto it.

We also seal the negative and positive terminals with silicone; when the silicone dries, we lower the structure into a glass of water and check the water resistance between the conductors with an ohmmeter - it should be very high. We lower the LED into a 200 gram jar of water or oil, make a hole in the lid and turn on the power. After an hour of operating the lamp at 1000 mA, the water temperature rose from 19 to 21 degrees. An LED with a radiator can be glued directly to the bottom of the can, and an E 14 or E 27 socket can be attached on top - you can screw it into the chandelier by first converting the power supply to 12 volts. With such cooling, the LED does not blind as much as without water! Very pleasing to the eyes. If you add aromatic liquid to the oil, it will smell pleasant when illuminated.

Author of the article “Do it yourself: powerful water-cooled LED” Jurei-678

Heat

To measure the heating of Philips Z ES LEDs, I heat the samples for 60 minutes. Unlike the original Philips Ultinon car lamps, Smart H7 also has a flexible radiator. Therefore, the service life of Smart will be approximately 2 times longer than that of real Philips Ultinon.

I take measurements near the LEDs on the plate. There were already holes for installing a temperature sensor on the copper board. I just used a file to make a groove for the thermocouple wire to come out.

The assembly and materials turned out to be very good, they did not skimp on thermal paste, this rarely happens. The diodes are on a copper plate, not an aluminum one. The thermal conductivity of copper is almost 2 times higher, and it is much more expensive than aluminum. The flexible part of the radiator is crimped with copper and placed on heat-conducting glue, so it is glued well. The petals are pressed very tightly, I tried to loosen them, but it didn’t work.

In cheap car lamps, the petals sometimes dangle, are connected to the body loosely or without thermal paste. The flexible petals even fell out of the car light bulb when trying to install them in the headlight.

For the most part, I am a supporter of a passive cooling system for LED lamps for cars. Passive cooling is more universal and is applicable not only to closed low-beam or high-beam headlights. Passive can be installed in foglights (abbreviated PTF) and other unprotected places. The simpler the cooling, the more reliable it is. Car LED lamps with a passive radiator are of course more difficult to install, but this can be solved with a set of additional covers. Enlarged rubber covers hide a massive rigid or flexible radiator without interfering with the operation of the electric corrector or hydraulic headlight angle corrector.

I have higher requirements for cooling with a fan, because problems with heat dissipation are much more common. Usually such models look decent on the outside, but when you figure it out, there are big problems inside,

Raw material

Aluminum is most often used to make heat sinks these days. The thing is that this material is very convenient for such purposes, and at the same time quite cheap. But if the dimensions of the product matter, then it is unlikely that you will be able to find anything better than copper, since it has greater heat conductivity, which means the heat sink will be 2 times smaller in size.

But not only these two materials are suitable for making a cooling device? It makes sense to understand what other raw materials a heat sink can be made from and what their differences are.

Aluminum

Aluminum radiator
In terms of thermal conductivity, the average value ranges from 200 to 240 W/m*K, which is almost 3 times higher than the same parameter for brass and iron. It mainly depends on the presence and amount of impurities in aluminum. Of course, this is an easy-to-process metal, which is why it is so common, but still, provided that the device body is small and decent cooling is required, an aluminum radiator is inferior to copper.

Copper

The indicator of this metal is 2 times higher than the thermal conductivity of aluminum, second only to such a noble metal as silver, and amounts to 400 W/m*K. But despite the fact that copper cools so well, such radiators are quite rare. The thing is that it is quite expensive when compared with aluminum, and besides, it is difficult to machine and has a large mass.

Copper radiator

It turns out that if copper coolers are installed in an LED lamp, its price will increase, and this is unacceptable, because as a result, the company will become uncompetitive in a tough market.

Ceramics

The thermal conductivity parameter is close to the parameters of aluminum and amounts to 175–235 W/m*K. Ceramics are convenient because they themselves are dielectrics, which is important in electronic and electrical circuits.

And yet, with such thermal conductivity, it loses to another material that is very easy to handle.

Thermoplastic

Of course, the thermal conductivity parameters of thermoplastic are slightly lower than those of aluminum (from 5 to 40 W/m*K), but it has some advantages. In addition to its dielectric properties, it is also very light and low cost. Only when designing LED lamps more powerful than 10 watts, it clearly loses to aluminum and copper.

Why is it needed?

Along with other semiconductor devices, the LED is not an ideal element with 100% efficiency. Most of the energy it consumes is dissipated into heat. The exact efficiency value depends on the type of emitting diode and its manufacturing technology. The efficiency of low-current LEDs is 10-15%, and for modern white ones with a power of more than 1 W, its value reaches 30%, which means that the remaining 70% is spent as heat. Whatever the LED, for stable and long-term operation it requires a constant removal of thermal energy from the crystal, that is, a radiator. In low-current LEDs, the radiator function is performed by the leads (anode and cathode). For example, in SMD 2835 the anode lead takes up almost half of the bottom of the element. In high-power LEDs, the absolute value of power dissipation is several orders of magnitude greater. Therefore, they cannot function normally without an additional heat sink. Constant overheating of the light-emitting crystal significantly reduces the service life of the semiconductor device and contributes to a gradual loss of brightness with a shift in the operating wavelength.

What kind of heatsink is needed to cool an LED?

LEDs, which appeared on the radio electronics market relatively recently, have already firmly taken a leading position in relation to other light sources. They are the most economical in terms of energy consumption, more compact and convenient to use and have less heat generation.

And yet, no matter how high-tech the LED is, an increase in temperature during its operation cannot be avoided. In addition, when heated, such an LED element, due to its design features, begins to lose its luminous flux.

Of course, if it is a regular DIP LED with two contact legs, external cooling is quite enough for it. But if we take more powerful elements, then it’s worth thinking about a cooling radiator for LEDs, which would help remove heat from the light source.

If you pay attention to similar cooling devices in stores, you can understand how high their cost is. What to do then?

It remains to figure out whether it is possible to make a radiator for a specific LED or group of LEDs yourself, how to do it, and how difficult it is. Now we will try to resolve this issue.

Do you need a radiator?

To begin with, it makes sense to understand whether a cooling radiator is needed for an LED and, if so, why.

The fact is that in terms of efficiency, if we take low-current diode emitters, their efficiency is only 15–17%. It is clear that the rest of the energy will be spent on heat generation. Of course, the efficiency of more powerful LEDs (more than 1 watt) is 2 times higher, but they also consume more energy.

So any such lighting device ultimately releases a certain amount of heat, which must go somewhere. For example, in the SMD2835 light diode, the anode contact makes up slightly less than half of the component; it provides the necessary heat dissipation, and this despite the fact that it is low-current. It turns out that it already has a radiator. But powerful LEDs require more attention.

When the crystal temperature is constantly increased, the wavelength of the radiation shifts, resulting in reduced brightness and greatly reduced service life. It turns out that there is no way to do without a radiator when independently installing a circuit using powerful LEDs.