State of the art In electric circuits a non-negligible heat generation often takes place. If the generated heat is not taken care of in a suitable way, the circuits run the risk of being overheated. To control the temperature of electric circuits, it is therefore common to arrange different forms of cooling systems in connection with the circuits.

Some of these cooling systems are active in the sense that extra power must be supplied for the cooling to be able to take place. Fans are an example of active cooling systems, also cooling systems, which use the same physical principles as refrigerators and freezers, are active. A disadvantage of active cooling systems is of course that they require a supply of power in order to work. If the power supply would for some reason be interrupted, the circuits run the risk of being overheated.

Active cooling systems are often relatively complicated and therefore require a larger amount of maintenance. These features of the active cooling systems result in that they are not regarded suitable for circuits in outdoor equipment, such as radio base stations in mobile telephone systems.

Other such cooling systems are passive, and no extra power needs to be supplied for these to work. The passive cooling systems are often constructed to conduct in an efficient way the generated heat in the circuits to the surrounding air. Different

physical processes are hereby used to achieve this, such as natural convection, heat radiation and evaporation. To achieve an effective thermal connection to the surrounding air, the passive cooling systems often comprise heat exchangers with heat sinks, e. g. in the form of cooling flanges.

An example of a passive cooling system is given in US Patent 5 529 115. In this patent document, devices for cooling electric circuits are shown. A chamber is partly filled with a liquid which is vaporized by the heat which develops in the electric circuits. The vaporized liquid is condensed against the ceiling of the chamber, and the condensation drops fall back into the liquid. Special means are provided for guiding the condensation drops back in a suitable way so that stagnation is avoided.

Another example of a passive cooling system is given by US Patent 5 390 077. In this patent document, a device for cooling electric circuits is shown. The device comprises a container with an upper part (cold plate) and a lower part. The container is filled with liquid. The upper part is thermally connected to a heat exchanger with a number of cooling flanges. The electric circuits are arranged on the lower part and are completely submerged in the liquid, to maximize the heat transfer. The dimensions of the container are such that a Raleigh number of at least 1700 is achieved, whereby an effective heat transport between the upper part and the lower part is achieved because of a well-established natural convection in the container.

A large volume of modern electronics are not specified for temperatures below 0°C.

However, electric circuits in outdoor equipment must work even if the temperature of the surrounding air falls towards-35°C. A disadvantage of the passive cooling systems is in this case that these cooling systems cool the electric circuits very well, also at low air temperatures, whereby the temperature in the electric circuits runs the risk of becoming lower than 0°C. One way to avoid this disadvantage is to install heaters which heat the electronics at low air temperatures, so that the temperature in the electronics will not fall below 0°C. However, this solution has several

disadvantages, such as: high costs (heater, control equipment, mechanical equipment, wiring etc.), space requirements (heater and associated equipment) and high electric power consumption (it is not unusual that the heater consumes more electric power than is normally consumed by the equipment itself).

Description of the invention The present invention tackles the problem of achieving a device for controlling the temperature of a heat generating object, e. g. electronics, so that the temperature of the object will not exceed a predetermined maximum value or fall below a predeter- mined minimum value when the temperature of a surrounding region varies within a predetermined interval.

In short, the problem stated above is solved according to the following. The device according to the invention comprises a heat exchanger, which is arranged for emit- ting heat to the surrounding region. The device further comprises a container, which is arranged to contain a medium with temperature-dependent properties. The con- tainer is thermally connected to the object as well as to the heat exchanger. When the temperature of the surrounding region exceeds a first value, the medium has a low viscosity, and heat generated by the object is transported in the container in a relatively effective way, from the object to the heat exchanger by natural convection in the medium. Since the heat transport by natural convection is relatively effective, the temperature of the object is kept low and thus does not exceed the maximum value. The medium is chosen to have such temperature dependent properties that the natural convection in the medium mainly is prevented when the temper-ature of the surrounding region falls below a second value. The heat generated by the object is then transported in the container from the object to the heat exchanger mainly by thermal conduction in the medium. However, the heat transport by thermal conduction is considerably less effective than that by natural convection, for which reason the temperature of the object will not fall below the minimum value. Thus,

the container has a heat transferability which varies relatively strongly in depend- ence upon the temperature of the surrounding region. At higher temperatures of the surrounding region, the heat transferability is good, while the heat transferability at lower temperatures of the surrounding region is relatively poor, whereby the con- tainer isolates the object from the heat exchanger.

Thus, the purpose of the invention is to control the temperature of the object when the temperature of the surrounding region varies in a predetermined interval, and the invention comprises a device for achieving this.

A main advantage of the invention, apart from solving the problem stated above, is that the temperature control of the object is completely passive.

The invention will now be described more closely with the aid of preferred embodi- ments and with reference to the attached drawings.

Description of the figures Figure 1 shows, in a sectional view, as an example according to the invention, a device for controlling the temperature of electronics.

Figure 2 shows, in a diagram, a curve showing how a temperature difference (the vertical axis) between the temperature of the electronics and the temperature of a surrounding region varies in dependence upon the temperature of the surrounding region (the horizontal axis).

Preferred embodiments In Fig. 1, a device 1 for controlling the temperature of electronics is shown in a sec- tional view, as an example according to the present invention. Here the expression

'electronics'comprises all forms of electric equipment, such as separate electric components, electric circuits, radio equipment, microwave equipment, and so on.

The device 1 in Fig. 1 can, of course, also be used to control the temperature of other objects than electronics.

The device 1 in Fig. 1 comprises a container 2 with a first mainly flat wall 3 and a second mainly flat wall 5, which is opposite to the first wall 3. In the example shown in Fig. 1 the first and the second walls 3 and 5 are arranged mainly vertically.

The container 2 further comprises an upper part 7 and a lower part 9. The container 2 additionally comprises further walls (not shown), which together with the first and the second walls 3 and 5 and the upper and lower parts 7 and 9 enclose a space containing a medium 11. In the example shown in Fig. 1, the medium 11 is a liquid with a temperature-dependent viscosity. Electronics in the form of a printed circuit board 13 with electric circuits 15 is arranged on the outside of the first wall 3 so that the electric circuits 15 have good thermal connection to the first wall 3. The device 1 comprises a heat exchanger 17 with a base plate 19 on which heat sinks in the form of cooling flanges 21 are arranged. The heat exchanger 17 is arranged with the base plate 19 against the outside of the second wall 5 so that the heat exchanger 17 has good thermal contact with the second wall 5. The heat exchanger 17 is arranged for emitting heat to a surrounding region 18 with a temperature Ts The surrounding region 18 in Fig. 1 constitutes a surrounding atmosphere, but is alternatively some other form of heat absorber or something else to which heat can be emitted from the heat exchanger 17. The invention is not limited to the particular design of heat exchanger 17 shown in Fig. 1, rather the heat exchanger 17 is alternatively some other means by which heat can be emitted to the surrounding region 18. For example, the heat exchanger 17 instead constitutes a wall, the surface and thermal conductivity of which being sufficient for effectively emitting heat to the surrounding region 18.

The device 1 in Fig. 1 is arranged to control a temperature Te of the electric circuits 15 when the temperature Ts of the surrounding region 18 varies within a predeter- mined interval [Tsl, between a lower and an upper temperature Tsl and Ts2. The temperature Te of the electric circuits 15 is in this case controlled so that it does not exceed a predetermined maximum value TCmax and also so that it does not fall below a predetermined minimum value Tcmin. In the example shown in Fig. 1, Tsi =-33°C, Ts2 = 55°C, Tanin = 0°C and Tcna, = 75°C, which are values often stated in technical specifications, which is well known for a person skilled in the art.

The invention is, of course, not limited to the temperature values given above, rather the person skilled in the art has the possibility of choosing these, according to the circumstances, in another way and to adapt the device 1 thereafter. For example, T is chosen between-40°C and-25°C, TCmjn between-10°C and 20°C, Ts2 between 40°C and 55°C, and TCmax between 60°C and 75°C.

In Fig. 1, the device 1 is shown when the temperature Ts of the surrounding region 18 exceeds a predetermined first value T1, which is less than Tus2. Then the medium 11 in the container 2 has a low viscosity (is very liquid). Heat generated by the elec- tric circuits 15 heats the first wall 3, which in turn heats the medium 11 in contact with the first wall 3. The heating of the medium 11 at the first wall 3 causes an upwardly directed natural convection flow indicated by the upwardly directed arrows in Fig. 1. The upwardly directed natural convection flow turns at the upper part 7 of the container 2, whereby heated medium 11 approaches at the second wall 5. Thereby the heated medium emits heat, via the second wall 5, to the heat exchanger 17. The heat exchanger 17 in turn emits the heat to the surrounding region 18. The heated medium 11 is cooled at the second wall 5, which causes a downwardly direction natural convection flow, as indicated by downwardly directed arrows in Fig. 1. The downwardly directed natural convection flow turns at the lower part 9 of the container 2, whereby the cooled medium 11 approaches at the first wall 3. Thus, in the container 2, there exists a circulating natural convection

flow which in a relatively effective way transports the heat generated by the electric circuits 15 through the container 2 to the heat exchanger 17, where the heat is emitted to the surrounding region 18. A plate 23 is arranged in the container 2 between the first and the second wall 3 and 5. However, the plate 23 does not reach all the way up to the upper part 7 of the container 2 and also does not reach all the way down to the lower part 9 of the container 2. The plate 23 keeps the upwardly directed natural convection flow at the first wall 3 separated from the downwardly directed natural convection flow at the second wall 5, which reinforces the circulating flow in the container 2.

The upper part 7 and the lower part 9 are arranged to thermally isolate the first and the second walls 3 and 5 from each other. For example, the upper part 7 and the lower part 9 are thin or made of a thermally isolated material, or are both thin and made of a thermally isolated material.

Fig. 2 shows, in a diagram, a curve 25, which schematically describes how a temperature difference AT (the vertical axis) between the temperature Te of the electric circuits 15 and the temperature Ts of the surrounding region 18 varies in dependence of the temperature T, of the surrounding region 18 (the horizontal axis).

When the temperature Ts of the surrounding region 18 exceeds the first value T1, the medium 11 is, as mentioned, very fluid and the heat generated by the electric circuits 15 is transported mainly by natural convection in the container 2 to the heat exchanger 17. The temperature difference AT is therefore relatively small when the temperature Ts of the surrounding region 18 exceeds the first value T1. Additionally, when the temperature T, of the surrounding region 18 exceeds the first value T1, the temperature difference AT is mainly constant and thus corresponds approximately to a fixed first difference value AT1. The first difference value AT1 is dependent upon the properties of the medium 11, such as viscosity, density, thermal conductivity and specific heat capacity, the shape of the container 2, primarily the areas of the first and the second walls 3 and 5 and the distance between the first and the second walls

3 and 5, as well as the dimensioning of the heat exchanger 17. In this case, the container 2 and the heat exchanger 17 are designed with consideration taken to the properties of the medium 11 so that for the first difference value AT1 it is true that AT1 < TCmax-Ts2 for which reason the temperature Te of the electric circuits 15 does not exceed the maximum value Tcmax In order for the temperature Te of the electric circuits 15 not to fall below the minimum value Tcmin, the medium 11 is chosen so that its temperature-dependent viscosity is such that the viscosity is sufficiently high to mainly prevent the natural convection flow in the container 2, when the temperature Ts of the surrounding region 18 falls below a predetermined second value T2 (which is less than Tl and higher than Tel). In this connection, the expression'is mainly prevented'means that the natural convection is prevented completely or to such a degree that the heat transport by natural convection in the container 2 is negligible. Thus, when the temperature Ts of the surrounding region 18 falls below the second value T2, the heat generated by the electric circuits 15 is transported mainly by means of thermal conduction through the container 2 to the heat exchanger 17. However, the heat transport by thermal conduction is considerably less effective than the heat transport by natural convection. Thereby the container 2 causes merely an isolating effect between the electric circuits 15 and the heat exchanger 17, which causes the temperature difference AT to be considerably larger than what is the case when the natural convection is well established. When the temperature Ts of the surrounding region 18 falls below the second value T2, the temperature difference AT is mainly constant and thus corresponds approximately to a fixed second difference value AT2. The second difference value AT2 depends upon the thermal conductivity of the medium 11, the shape of the container 2, primarily the areas of the first and the second walls 3 and 5 and the distance between the first and the second walls 3 and 5, as well as the dimensioning of the heat exchanger 17. The container 2 and the heat exchanger 17 are designed with consideration taken to the thermal conductivity of the medium 11 so that for the second difference value AT2 it is true that AT2 >

Tcmin-T, the reason for which the temperature Tc of the electric circuits does not fall below the minimum value Tcmin- The second value T2 is adapted so that the natural convection is mainly prevented in sufficient advance when the temperature Ts of the surrounding region 18 falls, for the temperature Tc of the electric circuits 15 not to fall below the minimum value Tcmin. but for other aspects the choice of the second value T2 is not critical. In a corresponding way the first value T1 is adapted so that a well established natural convection flow is obtained in the container 2 sufficiently in advance, when the temperature Ts of the surrounding region 18 increases, for the temperature of the electrical circuits Te not to exceed the maximum value Temax. but apart from that also the choice of the second value T2 is not critical. Fig. 2 shows that, when the temper- ature Ts of the surrounding region 18 varies in the interval between the second and the first values T2 and T1, there is a gradual transition from heat transport mainly by thermal conductivity in the container 2 to heat transport mainly by natural convection in the container 2. This is, of course, because the medium 11 gradually obtains a lower viscosity when the temperature Ts of the surrounding region 18 increases.

The medium 11 is in the example shown in Figs. 1 and 2, a fluid with temperature- dependent viscosity. The fluid is, for example, a suitably chosen oil. Another alter- native is that the fluid is a mixture of water and glycol. As a suggestion, the percen- tage by weight of glycol comprised in the mixture is 5 to 60, which gives a viscosity which varies relatively strongly with the temperature. Especially, the percentage by weight of glycol comprised in the mixture is approximately 60.

Another alternative is that the medium 11 is a substance which is fluid and has low viscosity when the temperature of the surrounding region 18 exceeds the first value T1, whereby the heat generated by the electric circuits 15 is transported mainly by natural convection through the container 2 to the heat exchanger 17, in a manner

corresponding to that in Fig. 1. However, the temperature-dependent properties of the substance are adapted so that the natural convection in the container 2 is mainly prevented when the temperature Ts of the surrounding region 18 falls below the second value T2 by the substance solidifying. For example, the substance consists of different forms of salts with melting points suitable for the circumstances, such as sodium acetate trihydrate (C2H3NaO2H603) or glauberite (Na2Ca (SO4) 2). Another alternative is that the substance is paraffin or similar.

Yet another alternative is that the medium 11 comprises a liquid which has low viscosity. In the liquid with low viscosity, there is an additive which, in dependence upon the temperature Ts of the surrounding region 18 forms a precipitate in the liquid, whereby the natural convection in the container 2 is mainly prevented by the precipitate when the temperature T, of the surrounding region 18 falls below the second value T2.

In the example shown in Fig. 1, the container 2 is oriented so that the first and second walls 3 and 5 are mainly vertical. However, the invention is not limited to only this orientation of the container 2, rather, alternatively, the container 2 can be oriented in another way. For example, the container 2 is oriented so that the first and the second walls 3 and 5 are mainly horizontal, whereby the plate 23 is suitably omitted.

Neither is the invention limited to only the shape of the container 2 shown in Fig. 1, but, alternatively, the container 2 has some other shape. For example, the first wall 3 is given a shape in which its contour is adapted relative to the shape of the object the temperature of which is to be controlled, whereby the device 1 takes up less room and the thermal connection between the object and the container 2 is enhanced.