This invention relates to optoelectronic devices and, in particular, to devices which are encapsulated. The use of encapsulation is well established and many complicated systems are built up by interconnecting encapsulated units.

The object of this invention is to simplify the building up of a complicated system which includes botii electronic and optoelectronic components. The important features of this invention are that a complete optoelectronic circuit is encapsulated as a single unit and the operational waveband is above 1200 nm, e.g. 1200 - 1600 nm.

The "complete circuit" comprises at least one optoelectronic component and electronic circuitry for the appropriate provision of power and the control of the optoelectronic component.

An optoelectronic component may be an optical signal generator such as an LED or laser but surface emitting devices are preferred to edge emitters. For use with signal generators, the electronic circuitry preferably provides drive current and optical control can be achieved by control of the drive current, e.g. by means of control networks. More complicated control is appropriate if desired, e.g. feedback loops which may include optical detectors. The provision of data to an optical transmitter is important and the electronic circuitry preferably carries out the modulation. In preferred embodiments the encapsulated device has an external terminal or terminals for receiving data and the encapsulated circuitry modulates this data onto die optical signal for transmission, e.g. by coupling to an external waveguide, e.g. to a monomode fibre.

Alternatively, an optoelectronic component may be an optical detector, e.g. a photodiode which is serviced by the electronic circuitry which, for example, provides power to the detector and it also separates the data, e.g. it demodulates. The data is provided to an external terminal.

The encapsulation may include more than one optoelectronic device, e.g. a transmitter and a receiver. Such an encapsulation is suitable for

connection to an optical highway which comprises a plurality of waveguiding structures and/or for duplex operation.

The electronic circuitry usually performs a variety of functions. Thus it provides interfaces for the or all of the optoelectronic components, e.g. it provides data to a transmitter and it receives data from a receiver. Other functions may include:-

(i) the control of the optoelectronic components;

(ii) the provision of electric power to the contents of the encapsulation; (iii) the provision of alarms, e.g. loss of light alarms.

In order to provide these functions the electronic circuitry usually has a plurality of external electrical connectors.

In addition, the electronic circuitry may include a control network for defining the operational parameters of the device. Preferably, the control network is adjustable from outside the encapsulation. It may include resistors whose resistance can be altered by heat spattering, e.g. by laser beams. Alternatively or in addition, it may include fusible links with active components such as transistors.

It is emphasised that an encapsulated device is complete. Thus, an encapsulated transmitting device is adapted to receive external data in electrical form and to provide an optical signal modulated with the data. In the receive mode the device receives data in the form of a modulated optical signal and it provides the data as an external electrical output. It will be appreciated that the external connections extend outside the encapsulation.

Devices according to the invention are conveniently assembled by securing the various optoelectronic and electrical components to a lead frame which provides both mechanical support and electrical interconnection. The various components can be secured to the lead frame by electrically conductive attachments and extra connections can be formed by wire bonding. Some components are provided with

"input/output" in the form of external "wires" so that external connections can be made, e.g. by establishing electrical continuity between the "wires" of two (or more) components. The electrical continuity can be established either by directly bonding the two "wires" to one another or

binding each wire separately to the same part of the leadframe (which part is electrically isolated in the final product). During assembly, more difficult manipulations are needed to bond two wires than to bond the separate wires to a leadframe and, therefore, bonding to d e leadframe is preferred.

After assembly, encapsulation is applied, e.g. by transfer moulding.

In use the encapsulated devices are usually positioned in housings which either include optical waveguides or sockets adapted to receive and locate optical waveguides, e.g. fibre waveguides, especially single mode fibres. Optionally, the encapsulation may include lens surfaces to improve the imaging between the waveguides and the optoelectronic component. Lens surfaces are conveniently produced during the transfer moulding process.

The invention will now be described by way of example wid reference to the accompanying drawings in which:-

Figure 1 is a circuit diagram of a circuit selected to exemplify encapsulation in accordance with the invention;

Figure 2 shows the circuit of Figure 1 mounted on a leadframe in preparation for encapsulation; Figure 3 shows the encapsulated circuit of Figure 2 mounted for use;

Figure 4 shows a cross section through an encapsulation having a lens;

Figure 5 illustrates a first method of interconnecting components, and

Figure 6 shows an alternative method of interconnecting components when weak signals are involved.

The encapsulation of the circuit illustrated in Figure 1 will now be described by way of example. It will be understood that a wide variety of circuits are suitable for encapsulation but the optical receiver illustrated in Figure 1 was chosen for the purposes of illustration.

As shown in Figure 1, the optoelectronic component 10 is a PIN photodiode sensitive to radiation between 1200nm and 1600nm. The PIN diode 10 has its active region implemented in InGaAs with an active

diameter of 90μm. The output of PIN diode 10 is amplified in a transimpedance preamplifier 11 coupled by capacitors 12 and 13 to a post-amplifier 14. Post-amplifier 14 is a single ended differential converter and the amplifier chain 11/14 has a high but limited gain. The output from the amplifier 14 is provided on terminals 15 and 16 for the electrical output of the optical input, i.e. for the demodulated data of an optical signal received by PIN diode 10.

In addition, the amplifier 14 provides a loss-of-light alarm on terminals

17 and 18, the threshold for this signal being set by the resistors 19 and 20 (which can be regarded as a control network). Power for the circuit is provided on rails 21 and 22 with smoothing by resistor 23 and capacitor

24.

Figure 2 shows the circuit of Figure 1 mounted on a leadframe 27 which includes an inner metallic skeleton 25 (which in the final device constitutes the metallic skeleton which provides electrical connections and mechanical support for the components shown in Figure 1). Leadframe 27 includes scaffolding 26 which surrounds the metallic skeleton 25. It is an important feature of the arrangement shown in Figure 2 that the optoelectronic component 10 is located in the centre of the frame. During use the device is coupled to an optical signal, usually provided by a fibre, and it is most convenient to provide the signal to the centre of the device.

In the finished device the skeleton 25 provides electrical connections for the various components but short circuits must be avoided and, therefore, the skeleton 25 cannot have mechanical integrity. However the leadframe 27 must have mechanical integrity in order to support the components during assembly. Therefore the leadframe 27 also includes scaffolding 26 which surrounds the skeleton 25. The lead-frames are conveniently produced by photochemical etching or stamping or laser cutting and large sheets can be etched to provide substantial numbers of individual items. To assemble the device, the various components are placed in their correct positions and attached by means of a silver loaded epoxy adhesive.

The assembly shown in Figure 2 contains all that is necessary to operate the circuit but the Figure 2 arrangement is inoperable because the scaffolding creates short circuits.

After assembly, the arrangement shown in Figure 2 is placed in the moulding chamber of a transfer moulding machine. This machine, which is not shown in any drawing, also includes a source chamber into which is placed a measured pellet of thermosetting plastic. When the mould has been closed the pellet is heated, e.g. to temperatures in the range 80 - 100 ^β C, so that the plastic becomes mobile. The mobile material is transferred into the moulding chamber where it surrounds the skeleton. After transfer, the moulding chamber is heated to approximately 200 ^β C whereby a chemical reaction occurs in the fluid moulding material whereby it sets into a solid form. Conveniently the device is removed to an oven wherein it is heated to complete the curing process. After curing the skeleton 25 and all the components are fully encapsulated in a medium which provides protection and mechanical support. (It should be noted that the mechanical support is provided mostly by the plastic encapsulation but the metallic skeleton substantially assists.)

After the curing is completed the scaffolding 26, which is outside die encapsulation, is cut away so that the short circuits are removed leaving an operable, encapsulated device in accordance with the invention.

The plastics material used for encapsulation is required to transmit the operational optical signals and it must have low attenuation at this wavelength. It is convenient for the encapsulation to have high attenuation outside the operational waveband to reduce the risk of interference by extraneous optical signals (unless it is desired to check the contents by visual inspection, when adequate transparency in the visible region of the spectrum is needed).

Figure 3 shows an encapsulated device 33 according to the invention mounted for use in a moulded plastic housing 39. The housing comprises a socket 31 which is adapted to receive a terminated fibre (not shown in the drawing) and to locate the said fibre for optical coupling to the encapsulated device 33 (and, more particularly, to its optoelectronic component 10). The housing 39 is shown attached to a panel 30. In addition to the socket 31, the housing 39 includes a receptacle 32 in which the encapsulated device 33 is located. If desired the housing 39 may include an integral lens 34 which facilitates coupling between the fibre and the device.

The device 33 is introduced into the housing 39 using a conventional active technique. That is to say the terminated optical fibre (not shown) is introduced into the socket 31 said fibre being attached to an optical source (not shown). The encapsulated device 33 is introduced into the receptacle 32 having its external leads connected so tiiat its performance may be measured. The position of the encapsulated device 33 is manipulated until maximum reception is achieved wherein the location of the device 33 wid in the receptacle 32 is finalised e.g. by means of adhesive or a weld. It will be appreciated that this mounts the device for future use.

It is important that the device 33 be adjusted for maximum transmission because, at the maximum, small displacements of d e fibre in the socket 31 have the least effect upon the optical coupling. This is important because the socket is intended to receive a fibre plug and, although the socket is accurate, small errors may occur and it is important that the effect of these small errors be minimised. It is emphasised that the device 33 is adjusted for maximum transmission because it is desired to minimise these errors even though maximum optical transmission may be less important. Many transmit applications specify a maximum power transmission and, when maximum coupling has been selected for the reason given above, the power transmission may be in excess of the maximum. Receive applications may specify a threshold for a loss of light alarm and it is desirable to set this after mounting to take into account couplers efficiency. The inclusion of a control (resistive) network in the encapsulated circuit may make it possible to adjust die performance parameters of the circuit in order to place them within specified limits. For example, the values of resistance can be adjusted, even after encapsulation by the local application of heat, e.g. by laser beams, in order to evaporate resistive material into the encapsulating matrix. Fusible links can be broken after encapsulation.

Figure 4 shows a modified version of the invention in which a lens is moulded into the encapsulating matrix. Figure 4 shows an optoelectronic component 10 (e.g. a PIN photodiode) which is enclosed in an encapsulating matrix 42. Figure 4 also shows the end 40 of an optical fibre which is optically linked to the PIN photodiode 10. The beam

passes through the encapsulant 42 which includes a lens surface 41 to assist imaging of the fibre end 40 on the active surface of the PIN photodiode 10. It will be apparent that the lens surface 41 is conveniendy moulded into the encapsulant by die transfer moulding process. The lens 41 could be a substitute for or in addition to die lens 34 which is integral with the housing 39.

As mentioned, die lens surface 41 assists imaging of the fibre end 40 on die active surface of the PIN photodiode 10. It will be apparent diat it requires accurate relative location between die mould (which produces and therefore defines die position of) the lens and d e optical component.

It is necessary to locate the the active surface on the axis of the lens and at die correct distance for focusing. A sharp focus is not necessarily the best location because a larger spot size, i.e. slight defocus, may illuminate more of the active surface and it usually more desirable to illuminate a larger than a smaller part of the active surface. A small tolerance can be allowed in die distance because this only means a smaller or larger spot size.

It is, of course, important that the spot illuminates die active surface and lateral displacement of the optical component is therefore strictly limited. During moulding d e leadframe, with all its components mounted tiiereon, is clamped in die mould and tiiis effectively sets the distance so tiiat spot size is appropriately controlled. In addition, it is also desirable to control the lateral position of die optical component and this invention provides an extra feature which facilitates the precise location of said component. According to this feature of the invention die leadframe is provided with a first index mark for defining d e location of die optical component and with a second index mark for defining die position of die leadframe witiiin the mould. These two index marks are produced from die same photographic mask and, tiierefore, their relative location are accurately defined. During assembly, it is relatively easy to place the optical component accurately on the first index mark. The second index mark conveniently takes die form of locational apertures which engage with pins in die mould so that the transverse position of die leadframe is accurately defined. Therefore, the position of d e optical component is accurately defined witiiin die mould and, when d e lens in moulded, d e active surface is sufficientiy close to the lens axis and at die right distance

from the lens. The invention includes leadframe having a first index mark for defining die position of an optical component and a second index mark for locating position of d e leadframe within a mould, the relative positions of said first and second index marks being accurately defined.

After it has been fully assembled and encapsulated, die optoelectronic device according to die invention is usually mounted experimentally for optimum transmission of light. Therefore adequate performance can be achieved even if the lens and the optoelectronic component are not similarly co-located but it is highly desirable to achieve the correct actual position and die correct actual distance to achieve the best performance.

The description given above explains tiiat it is convenient to connect components togetiier by connecting tiieir lead wires through the same parts of the leadframe. Some components, especially optical detectors, provide very weak signals as tiieir output and, if these weak signals are subject to interference, undesirably high signal-noise ratios may result. It has been recognised tiiat the leadframe may pick up interference and, in the special circumstances mentioned, die use of part of die leadframe to establish electrical connections can have unwanted side effects. A modification which reduces tiiis effect will now be described. According to this modification a connector pad is located on die leadframe and secured, e.g. by an epoxy adhesive. The connections are made to die connector pad instead of directly to the leadframe. The connector pad has a conductive layer to which the connections are actually made and an insulation layer which separates the conductive layer from die leadframe. The insulating layer effectively isolates the conductive layer from interference picked up by die leadframe and, d erefore, the level of interference in the transferred signals is reduced.

The two systems for making connections will be further described with reference to Figures 5 and 6.

As shown in Figure 5, a component 100 provides output signals on wire 101 and die output signals are received on component 102 via wire 103. Component 100 is mounted on a first part 104 of the leadframe by means of silver loaded epoxy adhesive 107. Similarly component 102 is mounted on portion 105 by silver loaded epoxy adhesive 110. As mentioned the two components are connected together and this is

achieved by bonding wire 101 to a portion 106 of the leadframe using epoxy loaded adhesive 108 and wire 103 is similarly bonded by silver loaded epoxy adhesive 109. Thus components 100 and 102 are connected togetiier via 106. Provided the signals in wires 101 and 103 are sufficiently strong there will be no adverse effects from the very weak interference which may be picked up in leadframe 106.

Figure 6 shows a modified version in which component 100 is an optical detector and component 102 is the first amplifier. Under these circumstances die signal level in wires 101 and 103 is very low and it is possible that tiiere could be adverse effects from even weak interference picked up in leadframe 106. Therefore the connection as shown in Figure 5 is slightly modified as shown in Figure 6. Parts which are effectively unchanged in Figure 6 retain the same reference numerals as in Figure 5.

As shown in Figure 6 as connector pad is bonded to leadframe 106 by means of epoxy resin 123. The connector pad 120 comprises a conductive layer 121 and an insulating 122. The connections 108 and 109 which, in Figure 5, were made directly to 106 are now made to die conducting layer 121. In this arrangement any interference which is picked up by leadframe 106 does not reach its conductive layer 121 because of the insulating layer 122. Conducting layer 121 is far smaller in extent that leadframe 106 so its own ability to pick up interference is substantially less and, therefore, even weak signals passing through wires 101 and 103 are not subjected to strong interference. Hence there is a substantial improvement in die performance for the transmission of weak signals. It should be noted tiiat die configuration illustrated in Figure 6 is no more difficult to assemble tiian that shown in Figure 5. The only extra step is the attachment of the connector pad 120 and this is one trivial operation.

It would be highly desirable for the encapsulating resin to have the same coefficient of thermal expansion as the leadframe but it will usually be difficult, or impossible, to achieve this. In general, it has been found tiiat differential diermal expansion does not have a major effect. However it would be desirable, as far as possible, to avoid the use of very large unbroken areas of metal in the leadframe and, if it is necessary to use long runs of metal it would be appropriate to incorporate flexible links to allow for differential diermal expansion. It is believed that the only

adverse effect worth considering is the possibility that a component could be displaced from its leadframe. However, at die sizes and temperature ranges likely to be encountered tiiis possible is considered remote.