# Cooling systems - from the radiator to liquid nitrogen! Part 1

Even from the school course in physics, we know that any conductor through which electric current flows generates heat. This leads to the fact that all the components of the computer through which current flows (from the processor to the connecting wires) heat the air surrounding them.

I wrote this article almost a year ago for one contest ... I decided to post it, maybe someone will be interested. It is terribly huge (no kidding - the hubr did not accept it in one piece, so there will be two parts / Part 2 /). Traffic .

## Boring theory

The amount of heat released depends on the contents of your system unit, on its energy consumption. This does not mean that you need to cool absolutely all involved components of the system. You don’t need to hang fans on sockets at all, but modern processors and video cards can’t do without cooling!

Alas, there is no escape from heat dissipation, but this problem has many solutions. Another question is how to cool. At the moment, there are quite a lot of cooling systems, all of them use the general principle of action - heat transfer from a hotter body (cooled object) to a less hot one (cooling system). We will consider only the following systems:

- Radiator + fan = cooler;
- Liquid cooling system;
- Cooling system on Peltier elements;
- Phase transition system (freon);
- The system of extreme cooling on liquid nitrogen;

You can use the most effective installations, which combine different types of these systems, but this is beyond the scope of this article.

In order, we will begin consideration of the main cooling systems and begin with the very first - the radiator.

Radiator (novolat. Radiator, "radiator") - a heat exchanger, is used to dissipate heat from the cooled object. The mechanism of heat transfer here is thermal conductivity, the ability of a substance to conduct heat inside its volume. All that is needed is to create a physical contact of the radiator with the object being cooled, which is why it is always in close contact with what it cools. After the radiator receives part of the heat from the cooled object, its task is to dissipate it into the surrounding air.

But it’s not enough just to make physical contact! After all, sooner or later, the cooling system itself will heat up from a constantly heated cooled object. And the heat exchange process in a system of bodies with the same temperature, as we know, cannot be. To find a way out of this situation and not to face the problem of overheating, it is necessary to organize the supply of some cold substance in order to cool the cooling system itself. Such a substance is commonly called a refrigerant (refrigerant, a special case of a coolant)

The radiator is an air cooling system, i.e. the refrigerant in this case is cold air from the environment. The heat from the cooled object goes to the base of the radiator, then it is evenly distributed over all its ribs, and after that it goes into the surrounding air. This process is called thermal conductivity. The air around the radiator gradually heats up, which makes the heat exchange process less and less efficient. The heat transfer efficiency in can be increased by constantly supplying cold air to the fins of the radiator. Simply put, effective cooling requires free circulation of cold air.

Physical quantities such as thermal conductivity (rate of heat propagation through the body) and heat capacity (the amount of heat that must be communicated to the body in order to increase its temperature by 1 degree) at the radiator must be at a high level. From the same school course, we know that metals have the highest thermal conductivity. In fact, this is not so - diamond has the highest thermal conductivity :), and it lies in the range from 1000 to 2600 W / (m · K). Of metals, silver conducts heat best of all - its thermal conductivity is 430 W / (m · K). After silver comes copper [390 W / (m · K)], then gold [320 W / (m · K)]. The chain ends with aluminum [236 W / (m · K)].

Having thrown back the jewelry, it becomes clear that the most applicable are two materials - aluminum and copper. The first is due to the low cost and high heat capacity (930 versus 385 for copper), the second is due to the high thermal conductivity (the disadvantages of copper include a higher melting temperature and the complexity of its processing). Silver, for its high thermal conductivity, is sometimes used to make the base of a radiator. Even for the manufacture of radiators, an alloy of aluminum with silicon, silumin, can be used. The advantage of its use is cheaper than aluminum.

If the radiator is made of highly heat-conducting material, then the temperature at any point will be the same. Heat will be equally effective over the entire surface area. Because Since the object transfers heat from its surface, this means that in order to achieve the best heat dissipation, the surface area of ​​the cooled object must be maximum. There are two ways to increase the area of ​​the radiator - increase the area of ​​the ribs while maintaining the size of the radiator and increase the geometric dimensions of the radiator. The second option, of course, is preferable, but it introduces a number of inconveniences - for example, it increases the weight and dimensions of the radiator, which can complicate the installation of the device. Well, the price, respectively, grows in proportion to the amount spent on manufacturing material.

There are a lot of types of designs for radiator fins. They can be thick if created by extrusion. Or vice versa, thin - if the ribs were cast. They can be straight along the entire length of the radiator, and can be cut across. They can be flat, bent from plates, pressed into the base. But the best in today's work are needle-type radiators - square or cylindrical needles instead of ribs in such radiators.

At the moment, I know 6 methods for the production of radiators:

1. Pressed (extrusion) radiators are the cheapest and most common on the market. The main material used in their production is aluminum. Radiators of this type are made by pressing (extrusion), which allows you to get quite complex profiles of the surfaces of the ribs and achieve good heat dissipation properties.

2. Folded (tape) radiators- obtained when a thin metal tape rolled into an accordion is soldered (or using adhesive conductive pastes) to a radiator base plate. The folds of the accordion tape in this case play the role of ribs. Such manufacturing technology allows to obtain compact products in comparison with extruded radiators, but with approximately the same thermal efficiency.

3. Forged (cold-deformed) radiators - radiators obtained as a result of using cold pressing technology. This technology allows you to create a radiator surface in the form of rods of arbitrary section, and not just standard rectangular fins. As a rule, they are more expensive than radiators of the first two types, but their efficiency is often much lower.

4. Compound radiators- close relatives of the "folded" radiators. Despite this, they are distinguished by a significant point: in this type of radiators, the surface of the fins is formed not by an accordion tape, but by thin separate plates, which are fixed by soldering or butt welding on the base of the radiator. Radiators of this type are slightly more efficient than extrusion and folded.

5. Cast radiators - in the production of products of this type, injection molding technology is used. The use of this technology allows to obtain profiles of the costal surface of almost any complexity, significantly improving heat transfer.

6. Turned radiators- are the most expensive and advanced radiators. Products of this type are created by precision machining (on special high-precision CNC machines) of monolithic billets and are characterized by the highest thermal efficiency. If not for the production cost, then radiators of this type would have long been able to supplant their competitors in the market.

### Heat pipes

In modern systems ceased to be a rarity used in radiators and coolers - heat pipes or just heat pipes.

It is a hermetic heat transfer device that operates in a closed evaporation-condensation cycle in thermal contact with external heat sources and sinks. Thermal energy is taken on the cooled object and expended on the evaporation of the coolant, which is located inside the heat pipe body. Further, thermal energy is transferred by the vapor in the form of latent heat of evaporation further, at a certain distance from the place of evaporation, where steam is released into the drain during condensation. The condensate formed again returns to the place of evaporation, either under the action of capillary forces (which are ensured by the presence of a specialized capillary structure inside the heat pipe) or due to the action of mass forces (this design is usually called a thermosiphon).

It turns out that instead of the usual electronic mechanism of heat transfer (through heat conduction, which takes place in a continuous metal heat conduit), a molecular transfer mechanism is used in the heat pipe (more precisely, the transfer process of the kinetic and vibrational energy of the random motion of vapor particles).

### There is a contact! What is the area?

It is necessary to strive to ensure that the contact area between the radiator and the cooled object is as large as possible - after all, it is through this area that the heat from the object will go to the radiator. But keep in mind that when two even the smoothest surfaces come into contact, the smallest cavities and gaps filled with air still remain between them [I remind you that the thermal conductivity of air is 0.026 W / (m · K)] - this can play a trick.

To get rid of harmful air and allow the radiator to work with maximum efficiency, various thermal interfaces are used, most often it is thermal paste (thermal paste). It has a high thermal conductivity [due to the use of substances such as aluminum and silver (up to 90% of the content)] and due to the fluidity it fills all the irregularities in the contacting surfaces.

Thermal grease comes complete with most branded coolers and radiators. It happens in the form of a syringe or a small tube-bag. It is recommended that thermal paste not get on the computer's electrical components.

One of the parameters of thermal grease is the length of the period when it reaches its maximum efficiency. On average, this time is about a week. Coolink recently made its first nanoparticle-based thermal grease - its advantage is that there is no waiting period.

In addition to thermal paste, there is another type of thermal interface - conductive gaskets. The essence of their work is the same, but they are used in a different way - they are laid on the contact surface and, when exposed to heat, change their state of aggregation, filling in irregularities and displacing air.

Despite all kinds of variations, the most important advantage of the radiator is that it is not a source of any noise. The disadvantages include relatively low efficiency, lack of potential for overclocking the system and often large dimensions.

If you trust the cooling of modern video cards and processors to passive radiators is quite dangerous, then the cooling of memory modules, hard drives, chipset, power circuits - you can rely on.

# Coolers

Cooler (eng. Cooler - cooler) is a combination of a radiator and a fan installed on electronic components of a computer with increased heat emission. The main task of the device is to reduce the temperature of the cooled object and maintain it at a certain level. This is achieved due to the continuous flow of air blowing around the radiator. That is, a less efficient radiation process turns into a more efficient process - convection. Coolers are the easiest, fastest, most affordable, and, in most cases, sufficient way to cool computer components - everything is cooled by air.

There are a lot of options for execution. If we talk about the appearance for a long time, then you can’t tell a lot about the functional differences.

Coolers come in different sizes - usually from 40x40mm to 320x320mm.

### Roller Balls

The most important part of any cooler is its fan. It is he who makes noise in your system unit. And to be more precise, this noise appears when the air stream collides with the radiator. This noise is especially noticeable on cheap cooler models, as nobody works on their design.

The fan consists of an impeller (a magnet is located inside it) and an electric motor that rotates this magnet together with the impeller. Through the center of the fan there is an axial pin, which is located in the center of the motor. For more smooth operation of the impeller, three types of bearings can be used (the service life of which is indicated by the manufacturers in thousands of hours on the package):

- Sleeve bearing - the cheapest and least reliable option that creates high noise levels during operation.
- 1 sleeve bearing + 1 ball bearing - a combined bearing - a more durable design, working on average two times longer than on a sliding bearing.
- 2 or 4 ball bearings are the most reliable options with a low noise level, but such fans are significantly more expensive than the first two.
- Needle and NCB (nanomillimeter ceramic) bearings - are installed in fans by a limited number of manufacturers. They are characterized by low noise, low cost and a very long service life.

By the way, about the service life (uptime). If the service life is specified in 40-50 thousand hours (almost 5 years. Although it happens more - up to 300,000 hours!), This does not mean that you will have to remember the cooler next time only after this time. No! This number needs to be divided by two or three, and still from time to time take preventive measures - dust, blow, lubricate. If you do not take care of the cooler, it may start to make noise, and if you really forget - and then stop.

U fan roizvoditelnost (metering characteristic)- perhaps his main characteristic. It is measured in the number of cubic feet of air distilled by it per minute, abbreviated as CFM (Cubic Feet per Minute). This characteristic mainly depends on the area of ​​the fan, the profile of the blades and the speed of their rotation. The higher this value, the higher the cooling efficiency and, as a rule, the higher the noise level created by the fan during operation.

### Healthy eating

The cooler can distill cubic meters of air with its blades at speeds up to 8000 rpm (for comparison, the engine of an ordinary passenger car produces 5-8 thousand revolutions. The engine of the Formula-1 car is up to 22 000 revolutions). But it’s clear that at this speed the noise from the cooler will be noticeable. Therefore, it is preferable to take coolers with temperature sensors - which “analyze” the temperature and, depending on the situation, can increase or decrease the number of revolutions. Most often, this positively affects the noise from work.

All computer coolers are powered by direct current, the voltage of which is most often 12V. They use Molex connectors (for Smart fans) or PC-Plug connectors to connect to power.

The PC-Plug has four wires: two black (ground), yellow (+ 12V) and red (+ 5V).

Molex connectors on the motherboard are used so that the system itself can control the fan speed by supplying a different voltage (usually from 8 to 12 V) to the red wire. From the yellow (signal) wire, the system learns from the cooler information about the speed of rotation of its blades. Using Molex has one significant drawback: it is dangerous to hook fans with a power consumption of more than 6W.

The situation is different with the PC-Plug - it can withstand tens of watts. But it was not without tar - when connected to it, you will not be able to find out if your fan is working or not.
Finding an adapter from one connector to another is now easy - they often come in the kit.

Also, to reduce noise, the cooler is sometimes switched to 5V or 7V. The loops are rounded, the wires are braided or tied with a braid and put away in a secluded place - so as not to interfere with a thoughtful air circulation.

All coolers are classified by the noise level emitted from their work in the following classes (the lower the noise level, the more comfortable the computer will be):

- Conditionally silent . The noise level of such a cooling system is less than 24 dB. This indicator is lower than typical background noise in a quiet room (in the evening or at night). Thus, the cooler makes almost no significant contribution to the noise picture. Typically, this value is achieved with a minimum fan speed for systems with speed control.

- low noise . The noise level from such a cooling system lies in the range from 24 to 30 dB inclusive. The cooler makes a barely perceptible contribution to the acoustics of the PC.

- Ergonomic. The noise level of such a cooling system lies in the range from 37 to 42 dB inclusive. Noise from such a cooler is likely to be noticeable in most user computer configurations.

- Not ergonomic . The noise level of the cooling system in question is more than 42 dB. In such conditions, the cooler will be the main "generator" of computer noise in almost any configuration. The home use of such a cooler is unjustified - it is more suitable for industrial and office premises with a background noise of more than 45 dB.

### Cooler Summary

The advantages of coolers include their prevalence, versatility, accessibility. A small cost can also be attributed to a plus, but it should be borne in mind that you should not be greedy for a good cooler - after all, this, in fact, is the second heart of the computer - you can’t stop it.

By cons, I will attribute possible noises that sooner or later will appear on any cooler.

To summarize the above. At the moment, the cooler is the most common cooling system, which can be cooled with anything - from the processor to the hard drive and memory. The question is the selection and selection of the right cooler - because there are a great many of them from dozens of manufacturers.

Someone needs a middle ground between silence and performance. Someone needs gigahertz and spit on the noise, someone on the contrary, prefers silence.