The electronic circuitry is convoluted despite being intrinsically simple and the whole design is quite clever.

Yet in practice it seems to be acutely sensitive to water ingress, inherently susceptible to early electronic failure (because of the use of X2 suppresser capacitors as voltage droppers) and let down by suspect quality pcb design and construction.

Apart from that - it shows real promise!

Although only 18 months old, the connectors are extensively corroded owing to having been immersed in high humidity as a result of a small quantity of water leaking into the unit.

The corrosion rate would have been increased because of the mixed metal combination - steel connectors pushed onto brass lugs.

The crimped copper wires can be seen on the connectors serving the blue and grey cables.

There are two areas where the printed circuits have vaporised or burnt away - one at the top of the picture (in an area carrying only mains voltage) and (to the middle right of the picture) where there are 24volt and 240 volt pcb tracks in close proximity.

The degree of corrosion is quite remarkable for a unit that is only 18 months old - it was manufactured in late 2008 according to the date stamps and failed in December 2010.

One of the BZT 24 volt zener diodes (centre right of picture) is so covered in corrosion products as to be unrecognisable.

The housing itself shows signs of distortion - maybe a consequence of overheating as a result of the motor running continuously - itself a consequence of the electronic and pcb failures, which in turn resulted in part from water ingress.

Because the board is secured tightly to the location plate, water is channelled by the U shape onto the top (component side) of the circuit board where it can either evaporate or act directly to short out both 24volt and 240 volt components.

The rust stains are from the steel screws used to secure the microswitches (and hence the circuit board) into position.

It would seem to be advisable for these actuators only to be positioned above valves, so that water leaking from a valve spindle would be denied any opportunity to affect the electronics.

It is also advisable for vertical pipework to have the actuator positioned with its cable pointing upwards - but 3 port design MUST be installed with its cable to the PORT A side of the 3 port valve (central heating side).

This board shows extreme examples of a type of degradation that has been seen in other Unishare actuators. Where there is a high voltage gradient (voltage per unit length between adjacent tracks) the pcb material seems subject to early breakdown and carbonation. This leads to current passing across and burning out the pcb material completely, as well as leading to a possible electrical connection between tracks that should be isolated from each other.

The example of burning at the top (highlighted in red) is extreme - the pcb is burnt nearly through its depth and had perforated at one point. The purple lines show the three sets of microswitch connections.

The connections in the black circle are 24 volt, yet mains voltage has tracked across in two areas (shown within blue circles) causing a capacitor to explode and the bridge rectifier to be short circuited. The areas in blue are where 24 volt and 240 volt tracks come quite close to one another.

Amazingly, the microswitches were in good electrical condition. The zener diodes also often survive when boards are damaged in this way.

Localised carbonisation of circuit boards is a recognised problem but is often not as extensive as shown here.

There may be a fundamental design weakness here - low quality pcb boards may simply not be suitable for use where closely adjacent tracks carry substantially different voltages. Water vapour within the unit may in time soak into the fabric of the pcb degrading its electrical resistance and leading to current flow. However, water ingress cannot be blamed in cases where no valve spindle or other leakage was evident - therefore the pcb quality and or track layout and design seems to be suspect. This circuit board was only 18 months old.

I would always recommend that domestic hot water and central heating systems use 2 two-port valves rather than a single Unishare of this type.

Two-port valve actuators are easier to repair, it is easier to diagnose faults and the valves themselves can more easily be operated manually if the actuator has to be removed for repair.

The mains neutral line is shown blue - this connects only to one side of the 240VAC motor and to one pin of each of the bridge rectifiers. All other tracks carry mains live voltage at various times during cycles.

The motor live connection and tracks to microswitches are shown yellow: the motor can receive power from many different places on the board, depending on where in the cycle it is and the state on incoming signals.

At three points, the purple areas come very close to tracks carrying mains voltage: these are shown by the small red circles. The two areas to the right of the picture seem prone to burning through - especially the one shown by the thicker small red circle.

At the top of the board there is another design problem: within the two concentric red circles the linked microswitch pins are extremely close to another track carrying mains voltage.  The next picture (below) shows this area in close-up.

When this area does burn through (as has happened on the board that is the subject of most of the photos on this webpage) the motor may receive power when it should not and so it runs continually - and overheats.

The track right at the top of the board is thicker than most because it carries the full load current to the orange wire - and it is this that tells the boiler to fire. In these 3 port valves, power supplied to the white wire (asking for central heating) flows via a relay contact and a microswitch to the orange wire.

Failure of either relay or any of the operative microswitch points (5 out the possible 6) can cause actuator malfunction. However, intermittent failure is usually owing to capacitor problems - which are much easier to diagnose and remedy!

The pcb was manufactured with one of the link wires standing proud of the pcb. This link carries mains voltage to the motor.

When the motor was screwed into place during assembly, the smoothing capacitor was compressed on to the link wire.

The red circle shows where the capacitor was pressed down by the motor casing.

The green circle shows where the link wire was eventually forced through the insulating plastic sleeve around the capacitor, thus imposing mains voltage on its metal outer case. In turn the whole motor body may have become 'live' but this seemed not to have caused any lasting damage.

As the capacitor was destroyed its casing ruptured, leaking chemical by-products - these can be seen as a black deposit on the top of the capacitor.

A pcb track on the other side of the board also failed owing to the sudden imposition of mains voltage onto a part of the circuit that was designed to work at only 24 volts.

In all, it is an unusual example of poor quality control - if a whole batch of Unishare actuators have been manufactured with the same fault then many if not all may be expected to fail in a similar manner.

This actuator was manufactured in 2009 and failed in January 2011. A manufacturing fault of this type may cause failure at any time - after a few weeks or many years' of operation, depending on how much pressure is exerted on the thin insulation around the capacitor casing.

In both cases the circuit boards exploded. One or more of the relay connecting pins were destroyed, thus rendering the units beyond economic repair.

In any case, both circuit boards were breaking down internally with carbonation evident within the thickness of the board. This damage seems to lead to high voltage tracking and subsequently to explosions.

The vaporised metal has been safely contained inside the actuator casing. In the unit on the right, one of the connecting pins has been blown away from the board by the force of the explosion, and its remains can be seen adhering to the side of the casing (top right corner).

These units were 8 years old - many of this age seem now to be failing in a similar manner as very slight water seepage causes dampness within the circuit boards.