This third contribution completes my efforts on this topic and is the simplest and cheapest to execute. Parts I and II used a super capacitor to store power for the hours of darkness whereas here we are using some form of rechargeable cell. The first version of this stage uses a nickel/metal hydride cell.
As a caveat I must say that I have some uncertainty regarding the longevity of the rechargeable cells whereas there I have hope that super capacitors will have a long life.
As with the previous sections this cannot be guaranteed to work everywhere in the world it depends on the nature, length and levels of the daylight.
You will need:
2 X 2 Volt 100mA solar cells. Very common but see:
A diode (not critical and a rectifier diode from the IN4xxx series will be fine.)
Soldering iron, solder, hook-up wire and a resistor to be chosen from the range 220 Ohm to 1K.
Small squares of single sided printed circuit board around 1 cm square. I used a junior hacksaw to cut these but please do not cut yourself! This might be a better solution:
A single Ni-mH rechargeable cell.
Step 1: The Clock
This was a present for which we had no immediate use so it has been languishing in a garden shed for some years. The dial is a bit faded but the standard battery movement is sound. The extra woodwork is a moveable date display system which is not used.
Step 2: The Circuit
The two solar cells are wired in series and the positive output is routed through a silicon diode to a ballast resistor which is connected to the rechargeable cell and the clock. The diode is vital and is to prevent the battery leaking power back through the solar cell when it is not being illuminated. The ballast resistor limits the current through the cell and we aim for around five milliamp or so. When the cell is fully charged the available current continues to flow through the cell acting as a very light float charge.
Step 3: Fix Solar Cells to Clock
The solar cells have to be on the face of the clock to pick up light. How you affix the cells depends on the clock. I opted to attach the two cells using MS polymer glue in a symmetrical configuration as shown in the picture like a pair of ears and this did save drilling the clock to pass the wires through.
Step 4: Attach Printed Circuit Board Pads
Our circuit is so simple that it can be accommodated on a few PCB pads and the picture show these stuck into position using MS polymer glue. It is best to leave the glue to cure overnight.
Step 5: Circuit Wiring
Simply follow the circuit diagram. The picture shown the completed circuit with a nickel/metal hydride cell installed. The resistor is 470 Ohm. Try to solder quickly and efficiently--prolonged heat may damage the glue under the pads.
Step 6: Discussion
The assembly worked from the word go. Nickel/metal hydride cells are sold partially charged and so you will probably not have to wait until the cell receives charge and this should mean also that the cells are good at holding the charge for the long term. The clock was placed near a south facing window and after three days which included some sunshine the battery was reading 1.38 Volts indicating a satisfactory state of charge. The charge current can be measured by reading the voltage across the ballast resistor and using I = V/R; in this case with 470 Ohms and a reading of 2 Volts the current is 4.25 mA. This will drop on dull winter days but for this location if we can reach autumn with a reasonably full charge then there is every chance that the battery can take us through the winter. Bear in mind that cold can adversely affect the cell performance. This easy way of checking the charge current is a good way of checking if there is sufficient light to run our clock in various locations.
Cell manufacturers do not claim a life of more than five years for this type of cell but that is for deep charge/discharge duty. Here we have a very strange duty with the cell very lightly loaded and probably spending summer with a reasonably full state of charge whilst receiving the lightest of float charge from the solar cells so how long will a cell last? I just do not know :-) but it will not cost a fortune to find out.
Step 7: The Last Word--Using a Dead Alkaline Cell
I have to own up to some embarrassment here and admit that I had to edit this step on May 18th 2016. The original idea of simply substituting an alkaline AA cell for the nickel/metal hydride cell in the previous steps showed enormous promise but the voltage steadily rising to 1.8 Volts or so resulted in a very messy leak after six weeks operation. I think that I have solved the problem at the cost of a slight increase in complexity.
There is some slight controversy over whether you can charge ordinary alkaline cells or if you should do so. A reasonable consensus might be that you can to an extent. The demands of the battery clock movement are so small that the 'extent' might be more than enough for our purpose. What we must not do is float charge the AA cell and allow the voltage to rise above the normal working voltage as this can cause leakage.
All that we have to do to try the idea is to take out the nickel/metal hydride battery from the project so far and substitute an alkaline cell for the nickel/metal hydride one. However we must take steps to limit the voltage to the alkaline cell working voltage of 1.5 Volts and this requires the charging circuit shown in the diagram above. A picture of the construction is also shown above.
The alkaline cell is now placed in the emitter circuit of a general purpose NPN transistor--I used an 2N3904. The base circuit consists of a 22k resistor feeding a red light emitting diode (LED) and a general purpose silicon diode in series. This places a voltage of around 2.05 volts on the transistor base resulting in around 1.45 Volts at the emitter and across the AA cell. The cell is simply being maintained at its normal working voltage with the clock taking very little power out of the battery and the solar circuit having to feed very little in--there is no element of float charging involved.
This last section is very much one for enthusiastic experimenters and I make no claims. A dead battery could last for the long term in lieu of a proper rechargeable cell--I just do not know.