WO2012032307A1

WO2012032307A1 - Planar transformer

Info

Publication number: WO2012032307A1
Application number: PCT/GB2011/001331
Authority: WO
Inventors: Danny Ireland; Colin Day; Alan Hopkin
Original assignee: Himag Solutions Limited
Priority date: 2010-09-09
Filing date: 2011-09-09
Publication date: 2012-03-15
Also published as: GB201015054D0

Abstract

A planar transformer and a method of dissipating heat from a planar transformer comprise a metal thermally conductive member disposed between two of a plurality of electrically conductive planar members forming the primary and secondary coils of the transformer. The metal thermally conductive member is electrically insulated from the plurality of electrically conductive planar members, and comprises a portion extending beyond said two electrically conductive planar members, to conduct heat away from the transformer.

Description

PLANAR TRANSFORMER

The present invention relates to planar transformers, and in particular how to provide a planar transformer with improved heat dissipation.

Conventional wire-wound transformers operate at relatively low frequencies

(of the order of Hz), which allows the use of relatively thick conductors without incurring unacceptable skin effect losses. Such transformers can use insulated wire windings or, for example, can be made from thicker copper "bars" (approximately rectangular cross section "wire" with a thickness of a few millimetres ) with an enamel coating, or "foil" with an insulated backing that are shaped into coils.

Planar transformers provide advantages over standard, wire- ound, transformers by being able to provide a transformer with a higher power rating in a smaller volume. This is achieved by creating the windings from flat "lead frames" or "stampings" connected into a spiral or by using traces on a PCB as the windings, whilst operating at high frequencies (i.e. 1 kHz or more). The compact nature of planar transformers makes them attractive for use within printed circuit boards, and in the growing technical fields of electric automobiles and solar panels.

As planar transformers are often used in enclosed spaces where space is limited, it can be difficult to cool them. The transformers generate heat, for example due to resistance in the windings and core losses, and if this heat is not dissipated the transformer heats up. Overheating of the transformer can eventually lead to failure of one of the components of the transformer. In any case, for a particular operation, it may be undesirable for a transformer to operate above a certain temperature.

Therefore, prior art planar transformers have been provided with external heat sinks which conduct heat away from the transformer core. Whilst this does provide some cooling effect, planar transformers are still liable to overheating.

The present invention aims to at least partially overcome the above- mentioned problem of overheating in planar transformers.

According to the invention there is provided a planar transformer comprising: a plurality of electrically conductive planar members, connected together to form primary and secondary coils; and a metal thermally conductive member disposed between two of the plurality of electrically conductive planar members, wherein the metal thermally conductive member is electrically insulated from the plurality of electrically conductive planar members, and comprises a portion extending beyond said two electrically conductive planar members.

The coils of the planar transformer generate heat, due to resistance in the coils, when the transformer is in use. The metal thermally conductive members are inserted between the turns or electrically conductive planar members of the coils to conduct heat away from the coils. In this way, heat is conducted away and dissipated, allowing the transformer to run cooler at a given power, or to run at a higher power without over heating. This method of dissipating heat is more efficient than relying on heat to be conducted away by the transformer core, because the primary and secondary coils usually extend outside of the core, and so there are some areas of the coils which will not be in close thermal contact with the core. In contrast, the present invention allows for areas of the coils distant from the core to be cooled. This alleviates hot spot formation, and therefore reduces the likelihood of failure in the transformer due to overheating.

Preferably, the metal may be copper or aluminium. Although most metals would be suitable for conducting the heat away from the coils, copper is a particularly good thermal conductor. Aluminium does conduct heat as well as copper, it is cheaper and so may be preferable if cost is an issue. Preferably, the metal thermally conductive member is solid, to provide the greatest amount of heat conducting material per unit volume.

The portion of the metal thermally conductive member extending beyond the two electrically conductive planar members may comprise a means for connecting to a heat sink, and/ or cooling fins. The planar transformer may additionally comprise a heat sink thermally connected to the metal thermally conductive member. The thermally conductive member removes heat from the transformer, and subsequently dissipates the heat itself by any suitable means. In particular, the conductive member may be connected to a heat sink, to which the heat is conducted, or it may dissipate the heat to the surroundings. When dissipating heat to the surrounds, the provision of cooling fins increases the surface area for heat dissipation, which will help increase the rate of heat dissipation. The rate of heat dissipation may be further increased by providing a flow of cooling air or other gas over the fins.

When a heat sink is thermally connected to the metal thermally conductive member, the metal thermally conductive member is preferably only in electrical contact with the heat sink at one point. This is because heat sinks are often electrically, as well as thermally, conductive. Therefore, this ensures that the conductive member is only terminated at one point on the heat sink avoids the risk of forming a complete conductive loop around the transformer windings, and thereby creating a short circuit in the transformer.

The planar transformer may further comprise a ferromagnetic or ferrimagnetic core, preferably made of a ferrite, powdered metal, or a nanocrystalline material. Two windows extend through the ferromagnetic or ferrimagnetic core, defining a core centre limb therebetween, and the primary and secondary coils may extend around the centre limb and through the two windows. The provision of the ferromagnetic or ferrimagnetic core helps increase the strength of the magnetic fields generated by the transformer coils. In particular, when the coils are formed around the central limb of a core comprising two windows, the coils are surrounded on all sides, other than the ends wrapping around between the two windows, by a closed loop of core.

The aforesaid two electrically conductive planar members may each comprise a portion extending outside of the ferromagnetic or ferrimagnetic core, and the metal thermally conductive member may be disposed between the two portions extending outside of the ferromagnetic or ferrimagnetic core. In this configuration, the thermally conductive member provides conductive cooling to an area of the coils that is only in distant thermal contact with the core, which may also provide some conductive cooling.

The metal thermally conductive member may be at least partially disposed within one of the two windows. That is, the conductive member may extend into one of the windows, so that it also provides conductive cooling to the coils within the window. This allows more direct conductive cooling of the coils in the window, especially those turns or windings towards the centre of the coils.

The metal conducting member may be partially disposed within both windows. That is, it may extend around the centre limb and project into both windows. However, it preferably does not form a continuous loop around the core centre limb, as that could cause an electrical short circuit in the transformer.

The primary and secondary coil may each comprise one or more groups of electrically conductive planar members, and the groups of the primary coil may be interleaved with the groups of the secondary coil. The interleaving of groups of conductive planar members or turns results in more efficient transformer operation, due to a reduction in parasitic losses.

The metal thermally conductive member may be disposed between electrically conductive planar members within one of the groups or between electrically conductive planar members of adjacent groups. In both cases, heat generated by the groups can be conducted away by the thermally conductive member, thereby reducing the temperature of the groups.

The planar transformer may further comprise two or more metal thermally conductive members, each respectively disposed between a respective two electrically conductive planar members. That is, the invention is not limited to the use of one metal thermally conductive member, as has been previously discussed. Multiple thermally conductive members may be used, at different positions and locations within the transformer, to remove heat from different potions of the transformer.

The electrically conductive planar members of one of the primary and secondary coils may be lead frames. The electrically conductive planar members of one of the primary and secondary coils may be conductive tracks on printed circuit boards. Accordingly the invention may be applied to any sort of planar transformer.

According to the invention there is also provided a method of cooling a planar transformer, the transformer comprising a plurality of electrically conductive planar members, connected together to form primary and secondary coils, the method comprising: disposing a metal thermally conductive member between two of the plurality of electrically conductive planar members, and electrically insulating the metal heat conductive member from the plurality of electrically conductive planar members, wherein the metal thermally conductive member comprises a portion extending beyond said two electrically conductive planar members.

The invention will now be described, by way of example only, with reference to the Figures in which:

Fig. 1 shows the building blocks of a lead frame group for a planar transformer, in particular Fig. la shows a plan view of a lead frame, Fig. lb shows a plan view of an insulator, Fig 1 c shows a plan view of an insulator and lead frame combination, Fig. 1 d shows a perspective view of a group of lead frames separated by insulators and Fig. le shows two lead frames connected by a continuous connecting tab;

Fig. 2 is a cross section through a planar transformer;

Fig. 3a is plan view of a constructed lead frame planar transformer and Fig. 3b is a side view of the constructed lead frame planar transformer;

Fig. 4a is a cross section view through a planar transformer of the invention, parallel to a lead frame, with a first configuration of heat dissipating inserts and Fig. 4b and Fig. 4c are end views of a planar transformers according to Fig 4a;

Fig. 5 is a cross section view through another planar transformer of the invention, parallel to a lead frame, with another configuration of heat dissipating inserts;

Fig. 6 is a cross section view through a planar transformer of the invention, parallel to a lead frame, with a configuration of a heat dissipating insert extending through both core windows;

Fig. 7a is perspective view of a printed circuit board planar transformer group, and Fig. 7b is a side view of a planar transformer constructed from groups as shown in Fig. 7a; and

Fig 8 is a cross section view through a printed circuit board planar transformer of the invention, parallel to a lead frame, the configuration of heat dissipating inserts shown in Fig. 4a.

Fig. 1 shows the building blocks of a lead frame group for a planar transformer. Planar transformers typically have a low profile compared to standard transformers, because of the use of thin elements such as lead frames or etched tracks on a PCB to create the coils of the transformer. The planar members of a planar transformer are preferably 2.0 mm or less, more preferably 1.0 mm or less and still more preferably 0.8 mm or less in thickness. Further, planar transformers typically operate at high frequencies (i.e. 1 kHz or above) compared to the lower mains frequencies (of the order of Hz) for conventional transformers. This is because this provides a size advantage: the faster the switching frequency the smaller the component/product size, because the higher frequency allows a smaller number of turns for a given winding/voltage.

Fig. la shows a lead frame 10, which is used to construct the coils of the transformer. The lead frame 10 is an electrically conductive planar member. Typically the lead frame 10 is made of metal, most often copper due to its high electrical conductivity. However, any suitable conducting material could be used, such as other metals. In particular aluminium, which is cheaper than copper, could be used (although it is not as conductive as copper).

The lead frame 10 may be formed by cutting, stamping or etching it from a metal sheet. That is, they are formed separately from any insulating members. As a result, lead frames 10 are typically self-supporting, although they may remain flexible. The lead frames 10 may have a thickness in the range of from 0.1 to 2 mm, and are preferably in the range of from 0.2 to 1 mm, and more preferably in the range of from 0.2 mm to 0.8 mm. It is undesirable for the lead frames 10 to be too thick, as electromagnetic skin effects will dictate the upper limit of metal thickness that will be utilised by a transformer. These skin effects are in turn dictated by the frequency of operation of the transformer, becoming more pronounced as the frequency of operation of a transformer is increased and therefore reducing the maximum useful thickness. The lead frame 10 is approximately ring shaped, with a central hole 1 1. The central hole 1 1 allows the lead frame 10 to be inserted over the limb of a transformer core. The copper lead frame 10 is not a complete ring. The copper lead frame 10 has end parts 12 and 13, which do not contact each other. The end parts 12 and 13 can be soldered to the end parts of further lead frames to create a series of turns or windings. In this way, the primary and secondary coils of a planar transformer may be constructed. As a result, each lead frame constitutes a single turn or winding of the overall coil.

The lead frame 10 may optionally further comprise a terminal 14. In Fig. la, the terminal 14 is shown by a dotted line, to show it is an optional feature. The terminal 14 allows for the connection of the lead frame, for example, to an external connection or to a further winding group. Typically, a group of connected lead frames 10 may have terminals 14 on the first and last turn. The terminals 14 are typically integral to the lead frame 10, and may be solder-dipped or plated to enable the creation of a good contact with an external connector.

Figure lb shows an electrical insulator 20 used to separate the lead frames 10 in a transformer. Insulators 20 may be used to separate the turns of the lead frames 10 within a group or coil, and also to separate the primary and secondary coils of the transformer from each other. Insulators 20 are also used to separate the coils from the transformer core. The insulator 20 has a central hole 21 to allow the insulator 20 to be positioned over the limb of a transformer core, in the same way as the lead frame hole 1 1.

When used between the turns of a single coil, the insulator 20 may be provided with a further hole 22 or 23, to allow the lead frames 10 on either side of the insulator to be connected. In Fig. lb, holes 22 and 23 are shown by dotted lines, to show they are optional features.

The insulator 20 may be of any thickness and made of any suitable material to meet the insulating requirements for the transformer, which will depend on the voltage at which the transformer operates and the industry standards to be complied with. Suitable materials include Nomex (RTM) and polyimide films such as Kapton (RTM). Different thicknesses of insulator, or a different number of pieces of insulator, may be required between groups of windings compared to between turns of a particular winding. A different thickness of insulator, or a different number of pieces of insulator, may also be required to separate the coils from the core. Once again, these will depend on the industry standards to be complied with.

In a coil comprising several turns or lead frames 10, an insulator 20 with a hole 22 will be positioned between a first and second lead frame 10, to allow the lead frames 10 to be connected via soldering, for example, between respective end points 12 of the respective lead frames 10. An insulator 20 with a hole 23 will be positioned between the second and a third lead frame 10, to allow the second and third lead frames 10 to be connected via their respective end points 13. In this way, by alternating the position of the hole 22 or 23 in the insulator 20 used between layers, a winding or coil formed from lead frames 10 can be built. Such a winding is also referred to as a "group" 15. It is not necessary to use lead frames 10 of the same thickness for each turn of a group 15 built in this way. Indeed, for strength, it may be preferable to use thicker lead frames 10 for the end turns of the winding, which comprise the extra tab, or tabs, 14 for connecting to the external circuit.

Figure lc shows how a lead frame 10 is positioned on an insulator 20. As can be seen, the hole 1 1 and hole 21 of the lead frame 10 and insulator respectively are aligned. The lead frame hole 1 1 is slightly larger than the hole 20 to keep the lead frame 10 away from the transformer core limb (which will pass through holes 1 1 and 21). To further ensure that the lead frame 10 is kept spaced away from the hole 21, a spacer member 24 is placed around the hole 21 on the insulator 20. The spacer member 24 is also made of insulating material, and serves to locate the lead frame 10 on the insulator 20 by providing a rim around the hole 21.

For clarity, a gap is shown between the spacer 24 and the lead frame 10 in Fig. lc. However, it is preferable to use as much space as available for the lead frame 10 (within the constraints imposed for separation from the core). As such, it is preferable for there to be little or no gap between the spacer 24 and the lead frame 10. This also helps ensure the lead frame 10 is located as securely as possible. As can also be seen from Figure lc, the overall width of the insulator 20 is wider than the overall width of the lead frame 10. This is to ensure that the lead frame 10 does not come into contact with the adjacent lead frames either side of the insulator 20. However, it is possible to leave less space around the outer edges of the lead frame 10 than around the central hole 21. This is because, once a series of lead frames and insulators 20 have been built up to form a group 15, final insulating papers (without holes 22,23) are placed at each end of the group and the three outer edges of the group from which tabs 14 do not project are taped up.

Fig. Id shows an example of such a finished group 15. As can be seen, two terminals 14 project from the group 15. Each of the terminals 14 are connected to one end of the winding formed by the connected lead frames 10. The top and bottom surfaces of the group 15 are covered in insulator 20. The three outer edges of the group 15 from which the tabs 14 do not project are covered over with tape 16, which is also an insulating material. Any suitable insulating material could be used, for example polyimide tape such as Kapton (RTM).

The inner edges of the hole through the group 15 are not taped, since the small size and closed loop nature of the inner edges would make reliable taping impossible. In particular, a single piece of tape would not be able to be folded around the internal corners without being split, which would then provide a short creepage path. Instead, the spacers 24 (as mentioned above) are used to ensure adequate separation of the lead frame 10 from the core limb and from lead frames 10 on adjacent windings.

In use, the group 15 is positioned over the central limb 31 of an E-shaped core 30a as shown in Fig. 2. Fig. 2 is a cross section through a transformer, in which the transformer comprises three groups 15.

The purpose of the core is to concentrate the magnetic field created by passing a current through a winding wound around the core. If no core were present, then the field created by the coil would be weak, because the surrounding air would effectively become the core. The permeability of air (which quantifies the ability of a medium/material to support a magnetic field) is not very high. Therefore, it is preferable to use a core material with a high permeability (which can be thousands to tens of thousands of times that of air).

A transformer uses alternating magnetic fields in its operation, and therefore the core material must be capable of magnetising in opposing directions and not retaining any magnetism. Steel has a relatively high permeability, but the highly conductive nature of steel means that eddy currents will be induced in a steel core, due to the alternating field. This creates core losses in the transformer. The higher the frequency of the alternating field (switching frequency) then the greater the losses in the core. Lamination can be utilised to reduce the eddy currents in a steel core, and therefore reduce losses. Even so, steel cores are typically only used at low frequencies, such as for connecting to a mains supply at around 50 Hz.

Ferrite materials are essentially ceramic and therefore suffer from eddy current losses much less than steel. As a result, ferrites are a preferable material for use in high frequency (i.e. 1kHz or above) planar transformers. Other new materials are emerging and being developed, such as powdered metal and nanocrystalline materials, that are similar to ferrite in terms of operating frequencies but have the benefit of much larger saturation limits. This allows the core to be utilised to a higher level, and any such increase can lead to reductions in the core size and therefore the overall transformer size.

In Figure 2, the first (lowest) group comprises two connected lead frames 10, separated by insulating layers 20a-20c as a first group 15 of an overall primary coil 50.

Above this group is a second group 15, comprising a single lead frame 10 enclosed by insulators 20d and e. That is, in this case, the group 15 consists of a single turn. This is the secondary coil 60. For clarity, the secondary coil 60 is shown spaced away from the lowest primary coil group 15. However, in practice, the groups 15 will normally sit directly on top of each other, perhaps separated by another layer of insulator 20 to provide further insulation between windings. In other constructions, primary and secondary groups 15 may be separated by a plastic 'tray', surrounding the inner and/or outer edges of the upper group 15, and possibly interlocking with another 'tray' holding the groups 15 above and below.

Above the secondary coil 60, there is provided a third group 15 comprising a single lead frame 10 enclosed by insulators 20f and 20g. Once again, this group consists of a single turn. This group 15 is a second group of the primary coil 50. That is, it is not necessary for the groups forming the primary (or secondary) coil to have equal numbers of turns. The first and second groups 15 of the primary coil 50 are connected, for example, using terminals 14 from the end turns of each group 15. However, other connection mechanisms are possible, such as the use of connecting pins from one group 15 to another or by pre- forming the two groups 15 with a continuous connection therebetween. That is, instead of a terminal 14, a lead frame 10 could be provided with a tab 14a integrally connected to another lead frame 10. An example of this is shown in Fig. le. The tab could subsequently be folded over, around a group of another winding.

The interleaved structure of primary and secondary windings of Figure 2 reduces the A.C resistance in the coils and parasitic elements in the transformer. The primary and secondary windings 50, 60 may each be split into several groups 15 that are interleaved. That is, it is not a requirement that only one of the primary or secondary windings 50, 60 (as shown in Figure 2) comprises more than one group 15. Both the primary and secondary windings 50, 60 may contain a plurality of groups 15 with either series or parallel connections between the groups 15.

The groups 15 are provided around the central limb 31 of the E-shaped core 30a. The E-shaped core also has outer limbs 34, 35. Once the groups have been inserted, an I-shaped piece of core 30b is provided over the top of the E-shaped core 30a, to enclose the groups within an overall core 30. The core 30 may be made of any suitably ferromagnetic or ferrimagnetic material, such as ferrite, in which a magnetic field may be induced by current passing through the windings of the transformer.

Once constructed, the overall core 30 has two windows 32 and 33 through which the primary and secondary windings 50, 60 pass.

In an alternative construction, another E-shaped core could be used, in an inverted orientation to core 30a, instead of the I-shaped piece 30b. This would result in larger windows 32 and 33 being formed.

Figure 3a shows a plan view of a completed planar transformer. Fig. 3b shows a side view of the same transformer. As in Fig. 2 the primary coil 50 consists of two groups 15 provided around a single group 15 forming the secondary coil 60. The primary groups 15 are connected via connection 17, which (as mentioned above) could be a connecting pin or wire or formed by connecting suitable positioned tabs similar to tabs 14 of the groups 15 or by an integral connection such as connection 14a shown in Fig le. The secondary and primary coils 50, 60 preferably have their terminals projecting from opposite ends of the planar transformer, as shown.

In Fig. 3a a lead frame 10 is shown in dotted lines within the uppermost group, to aid understanding. The lead frame 10 is actually covered by at least the insulator 20g, and so is not actually visible. Similarly, the tape 16 has been omitted from the drawings of Fig. 3, for the sake of clarity.

As can be seen from Fig. 3a and 3b, the lead frames 10 can project out of each end of the core 30. In use, when current passes through the lead frames 10 forming the windings of the primary and secondary windings 50 and 60, the resistance of the lead frames 10 results in heat being generated. In addition, the core itself generates heat when excitation current passes through the windings. Heat generated by the lead frames 10 within the windows of the core 30 is partially dissipated by being transferred to the core, which itself may be mounted to a metal chassis or heat sink. That is, the core acts a heat sink for the lead frames. However, heat generated in the lead frames 10 outside of the core 30 is not conducted away by the core 30. As a result, even though high frequency transformers such as planar transformers are more efficient, hot-spots occur at ends of the lead frames 10 protruding from the core 30.

Figure 4a is a cross section through a planar transformer of the present invention. The cross section is through the plane AA indicated in Fig. 4b. The cross section shows the three limbs 31 , 34 and 35 of a core 30 and an insulating paper 20 provided on top of a group 15. In dashed lines, a lead frame 10 is shown, which underlies the insulator 20.

Compared to the previously discussed planar transformers, the planar transformer of Figure 4a also has heat conducting inserts or shunts 71. Inserts 71 are heat conductors made of metal and are preferably solid (i.e. without any internal cavities) to provide the greatest amount of conducting material per unit volume.

The heat conducting inserts 71 may be positioned between the lead frames 10 of a primary or secondary winding, or between groups o primary and secondary winding themselves (i.e. with a primary group 15 on one side of the insert 71 and a secondary group 15 on the other side). The heat conducting inserts 71 project in between the windings of the planar transformer. That is, they are above the lead frame 10 shown in dotted lines in Fig. 4a, and below the next lead frame 10 Out of the page' as shown in Fig. 4a. In Fig. 4a, the inserts 71 are in the region outside of the core 30. Further, at least part of each insert 71 projects beyond the coils. That is, the inserts 71 projects outside of the groups 15, so as to be exposed to the

surrounding environment, and not totally covered on the top and bottom by the groups 15. As such, heat located in the windings 50 and 60 in the regions outside of the core 30 is conducted away through inserts 71.

Preferably, the inserts 71 are electrically insulated from the primary and secondary coils 50, 60. This could be achieved either by using extra insulators 20, or by enamelling, for example, or insulating one side of the insert 71 to provide two sheets of insulation between the heat sink 71 and the lead frame 10 of the coil 50 or 60.

In the arrangement of Fig. 4a, the inserts 71 do not project inside the windows 32, 33 of the core 30. This allows the maximum amount of space in the windows 32, 33 to be utilised by the groups 15. That is, as the groups 15 have some flexibility, the inserts 71 may be inserted between groups 15 outside of the windows 32,33 even if the groups 15 fill the windows 32, 33 and thus prevent inserts 71 from extending into the windows 32, 33.

As shown in Fig. 4b, a plurality of inserts 71 may be used between different windings of the transformer that is, inserts 71 may be provided at different levels or heights within a transformer.

Further, some or all of the inserts 71 may be provided with fins to dissipate the heat to the surrounding. This may be provided in combination with forced air cooling, in order to help remove heat from the transformer. Alternatively, as shown in Fig. 4b and Fig. 4c, some or all of the inserts 71 may be connected to an external heat sink 72 to further conduct the heat away. Optionally, the planar transformer can be mounted on the heat sink 72. The heat sink 72 itself may have fins, and may be cooled by forced air cooling or even liquid cooling.

In both the arrangements of Fig. 4b and Fig. 4c, heat is removed from the transformer in the regions external to the core by the inserts 71. The removal of heat from the transformer prevents the transformer from overheating, and therefore eventually failing or being unfit for its intended use. The removal of heat also allows the transformer to be driven at a higher power without reaching an unsafe or unsuitable temperature. It is noted that two inserts 71 are used at each end of the transformer in Fig. 4a-c. This is because it is undesirable to short the magnetic flux generated by providing a full loop around the transformer core, which could occur if a single insert 71 were used to cross each end of the transformer, and then the inserts at each end were connected on both sides to the same heat sink or to each other. Therefore, two separate inserts 71 are used at each end. Alternatively, to avoid such a short circuit being formed, the inserts could be electrically insulated from each other and the heat sink to which they are attached. However, as electrical insulators are often also thermal insulators, this may be less preferable as it could reduce the heat dissipation to the heat sink.

One way to avoid any short circuits by forming a loop around the coils 50, 60 is to only terminate each insert 71 at one position on a heat sink. That is, by only providing one connection between a heat sink and the insert 71, a complete loop formed by any one insert 71 and another insert 71 or an insert 71 and heat sink is avoided.

Other configurations of heat removing inserts 71 are possible and examples are shown in Figs 5 and 6. It is noted that a lead frame 10 is shown in dotted lines in Figs. 5 and 6 to aid understanding. However, the lead frame underlies the insulator 20 in each Figure. An insert 71 may extend within the core 30. For example, in Fig. 5, inserts 71 are arranged to extend through each window 32, 33 respectively of the core 30. In the arrangement of Figure 5 each insert 71 is only connected to an external heat sink at one end, to avoid causing a short circuit by forming a loop around the groups 15.

In Fig. 6, an insert that is similar in shape to the lead frames 10 is used. That is, the insert loops around the central limb 31 of the core 30. However, the insert 71 does not create a closed loop around the core limb 31 to avoid shorting the magnetic flux generated. Similarly, the insert 71 is only connected to an external heat sink at one end, to avoid causing a short circuit by forming a loop around the groups 15.

The inserts of Fig. 5 and Fig. 6 are in thermal contact with a greater proportion of the windings 10, and therefore improve the efficiency of heat dissipation from the transformer by even more than the inserts of Fig. 4.

The invention can also be used with PCB type planar transformers. Fig 7a shows a single PCB group of windings. The PCB group 80 has several layers 81 formed integrally together to provide an overall PCB. Each layer 81 has an etched track 84, usually made of copper, defining a loop around the central hole 82. The track 84 is a conductive planar member. The track 84 may form a single loop or form a spiral of several loops around the hole 82. The tracks of each layer within the group 80 are internally connected and two terminals 83 on the PCB connect to the end points of the overall winding formed by the connection of the spirals between layers. In Fig. 7a, the etched track 84 of the top layer 81 is shown as being visible, for clarity. However, in practice it is covered by an insulating layer, so that the top winding is not exposed. Further, the right hand terminal 83 appears to be unconnected to the winding, but is connected to the end of the lowermost layer of winding (not shown).

Fig. 7b shows a side view of a planar transformer constructed from PCB groups 80. Other than the fact that the groups 80 are constructed from PCBs, the construction is similar to that of Figure 3. That is, two groups 80 are joined together to form a primary coil 50 and a single secondary group (i.e. forming the entire secondary coil 60) is interleaved between the two primary groups 80. The two primary groups 80 are connected via a pin 85, soldered between the two PCBs, and connecting the copper trace coils within each group via a terminal 83 on each PCB. A terminal 83 on each PCB allows external connection to a power source. On the secondary coil 60, similar terminals 83 allow connection to a further circuit.

Although the windings of the primary and secondary coils 50, 60 of Fig. 7b are constructed differently to those of Fig. 1-6, the principles of how the transformer works remains the same. As such, the issues regarding heating are also the same, and any shape of heat conducting insert 71 suitable for use with the lead frame construction is also suitable for use with the PCB construction.

Of course, since the groups of a PCB planar transformer are constructed as a single board 80, it is easiest to position heat conducting inserts between the discrete boards of the primary and secondary coils 50, 60. Figure 8 shows an example corresponding to the arrangement of Figure 5 for the lead frame construction. In this case, it should be noted that there is no insulator between the PCB and the heat conducting insert 71, because etched trace 84 is already insulated within the PCB. However, if exposed traces were present on the top of the PCB, then it would be preferable to provide some further form of insulation between the trace and the inserts 71.

If it is desired to provide a heat shunt 71 within the windings of a group (i.e. between the internal layers of a PCB) one option would be to construct the PCB group from two sub-groups and position the heat shunt therebetween. Alternatively, the PCB could be designed with a layer specifically embedded as a heat dissipating insert, with separate terminals for connecting to an external heat sink. The same considerations regarding avoiding causing a short in the magnetic field would still apply.

Claims

A planar transformer comprising:

a plurality of electrically conductive planar members, connected together to form primary and secondary coils; and

at least one metal thermally conductive member disposed between two of the plurality of electrically conductive planar members, wherein the at least one metal thermally conductive member is electrically insulated from the plurality of electrically conductive planar members, and comprises a portion extending beyond said two electrically conductive planar members.

The planar transformer according to claim 1 , wherein the portion of the at least one metal thermally conductive member extending beyond the two electrically conductive planar members comprises a means for connecting to a heat sink.

The planar transformer according to clam 1 or claim 2, wherein the portion of the at least one metal thermally conductive member extending beyond said two electrically conductive planar members further comprises cooling fins.

The planar transformer according to any one of the previous claims, further comprising a heat sink thermally connected to the at least one metal thermally conductive member.

The planar transformer according to claim 4, wherein the at least one metal thermally conductive member is in electrical contact with the heat sink at only one point.

The planar transformer according to any one of the previous claims, wherein the planar transformer further comprises a ferromagnetic or ferrimagnetic core.

The planar transformer according to claim 6, wherein two windows extend through the ferromagnetic or ferrimagnetic core, defining a core centre limb therebetween, and wherein the primary and secondary coils extend around the centre limb and through the two windows.

The planar transformer according to claim 7, wherein the two electrically conductive planar members each comprise a portion extending outside of the ferromagnetic or ferrimagnetic core, and the at least one metal thermally conductive member is disposed between the two portions extending outside of the ferromagnetic or ferrimagnetic core.

The planar transformer according claim 6 or 7, wherein the at least one metal thermally conductive member is at least partially disposed within one of the two windows.

The planar transformer according to claim 9 wherein the at least one metal conducting member is partially disposed within both windows, but does not form a continuous loop around the core centre limb.

The planar transformer according to any one of claims 6 to 10, wherein the core is made of a ferrite, a powdered metal, or a nanocrystalline material.

The planar transformer according to any one of the preceding claims, wherein the metal thermally conductive member is solid.

13. The planar transformer according to any one of the preceding claims, wherein the electrically conductive planar members have a thickness of 2.0 mm or less, optionally 1.0 mm or less, and further optionally 0.8 mm or less.

The planar transformer according to any one of the preceding claims, wherein the primary and secondary coil each comprise one or more groups of electrically conductive planar members, and the one or more groups of the primary coil are interleaved with the one or more groups of the secondary coil.

The planar transformer according to claim 14, wherein the at least one metal thermally conductive member is disposed between electrically conductive planar members within one of the groups.

The planar transformer according to claim 14, wherein the at least one metal thermally conductive member is disposed between electrically conductive planar members of adjacent groups.

The planar transformer according to any one of the preceding claims, wherein the at least one metal thermally conductive member comprises two or more metal thermally conductive members, each respectively disposed between two electrically conductive planar members.

The planar transformer according to any one of the preceding claims, wherein the electrically conductive planar members of one of the primary and secondary coils are lead frames.

The planar transformer according to any one of the preceding claims, wherein the electrically conductive planar members of one of the primary and secondary coils are conductive tracks on printed circuit boards.

The planar transformer according to any one of the preceding claims, wherein the metal is copper or aluminium.

21. A method of cooling a planar transformer, the transformer a plurality of electrically conductive planar members, connected together to form primary and secondary coils, the method comprising:

disposing at least one metal thermally conductive member between two of the plurality of electrically conductive planar members, and

electrically insulating the at least one metal heat conductive member from the plurality of electrically conductive planar members,

wherein the at least one metal thermally conductive member comprises a portion extending beyond said two electrically conductive planar members.

22. A planar transformer or a method of cooling a planar transformer

substantially as hereinbefore described with reference to and as illustrated in the accompanying drawings.