Some more information....
HOW THE CIRCUIT WORKS
The circuit is a single transistor oscillator called a feedback oscillator, or more accurately a BLOCKING OSCILLATOR. It has 45 turns on the primary and 15 turns on the feedback winding. There is no secondary as the primary produces a high voltage during part of the cycle and this voltage is delivered to the output via a high-speed diode to produce the output. The output voltage consists of high voltage spikes and should not be measured without a load connected to the output. In this case, the load is the battery being charged. The spikes feed into the battery and this prototype delivered 30mA as a starting current and as the battery voltage increased, the charging current dropped to 22mA.
The transistor is turned on via the 1 ohm base resistor. This causes current to flow in the primary winding and produce magnetic flux. This flux cuts the turns of the feedback winding and produces a voltage in the winding that turns the transistor ON more. This continues until the transistor is fully turned ON and at this point, the magnetic flux in the core of the transformer is a maximum. But is is not EXPANDING FLUX. It is STATIONARY FLUX and does not produce a voltage in the feedback winding. Thus the "turn-on" voltage from the feedback winding disappears and the transistor turns off slightly (it has the "turn-on effect of the 1 ohm resistor).
The magnetic flux in the core of the transformer begins to collapse and this produces a voltage in the feedback winding that is opposite to the previous voltage. This has the effect of working against the 1 ohm resistor and turns off the transistor even more.
The transistor continues to turn off until it is fully turned off. At this point the 1 ohm resistor on the base turns the transistor on and the cycle begins.
At the same time, another amazing thing occurs. The collapsing magnetic flux is producing a voltage in the primary winding. Because the transistor is being turned off during this time, we can consider it to be removed from the circuit and the winding is connected to a high-speed diode. The energy produced by the winding is passed through the diode and appears on the output as a high voltage spike. This high voltage spike also carries current and thus it represents ENERGY. This energy is fed into the load and in our case the load is a battery being charged.
The clever part of the circuit is the high voltage produced. When a magnetic circuit collapses (the primary winding is wound on a ferrite rod and this is called a magnetic circuit), the voltage produced in the winding depends on the QUALITY of the magnetic circuit and the speed at which it collapses. The voltage can be 5, 10 or even 100 times higher than the applied voltage and this is why we have used it.
This is just one of the phenomenon's of a magnetic circuit. The collapsing magnetic flux produces a voltage in each turn of the winding and the actual voltage depends on how much flux is present and the speed of the collapse.
The only other two components are the electrolytics.
The 100u across the solar panel is designed to reduce the impedance of the panel so that the circuit can work as hard as possible. The circuit is classified as very low impedance. The low impedance comes from the fact the primary of the transformer is connected directly across the input during part of the cycle.
The resistance of the primary is only a fraction of an ohm and its impedance is only a few ohms as proven by the knowledge that it draws 150mA @ 3.2v. If a battery is connected to the circuit, the current is considerably higher. The 150mA is due to the limitation of the solar panel.
Ok, so the circuit is low-impedance, what does the 100u across the panel do?
The circuit requires a very high current for part of the cycle. If the average current is 150mA, the instantaneous current could be as 300mA or more. The panel is not capable of delivering this current and so we have a storage device called an electrolytic to deliver the peaks of current.
The 10u works in a similar manner. When the feedback winding is delivering its peak of current, the voltage (and current) will flow out both ends of the winding. To prevent it flowing out the end near the 1R resistor, an electrolytic is placed at the end of the winding. The current will now only flow out the end connected to the base of the transistor. It tries to flow out the other end but in doing so it has to charge the electrolytic and this take a long period of time.
These two components improve the efficiency of the circuit considerably.
You will notice the battery is receiving its charging voltage from the transformer PLUS the 3.2v from the solar panel. If the battery voltage is 12.8v (the voltage during charging) the energy from the transformer will be equivalent to 9.6v/12.8v and the energy from the solar cell will be equivalent to 3.2v/12.8v. In other words the energy into the battery will be delivered according to the voltage of each source.
THE BLOCKING OSCILLATOR
The operation of the circuit has been covered above but the term BLOCKING OSCILLATOR needs more discussion. By simply looking at the circuit you cannot tell if the oscillator is operating as a sinewave or if it is turning on and off very quickly. If the circuit operated as a sinewave, it would not produce a high-voltage spike and a secondary winding would be needed, having an appropriate number of turns for the required voltage.
A sinewave design has advantages. It does not produce RF interference and the output is determined by the number of turns on the secondary. The disadvantage of a sinewave design is the extra winding and the extra losses in the driving transistor, since it is turned on and off fairly slowly, and thus it gets considerably hotter than a blocking oscillator design.
The factor that indicates the circuit is a blocking oscillator is the absence of a timing capacitor. The circuit gets its timing from the inductance of the transformer. It takes time for the current to start to flow in an inductive circuit, once the voltage has been applied. In technical terms CURRENT LAGS IN AN INDUCTIVE CIRCUIT.
The timing feature is hidden in the circuit, but it has nothing to do with the feedback winding or the transistor. If we simply place the 45 turn coil (the transformer) across a voltage source, current will flow in the coil and this will produce magnetic flux. This flux will cut all the turns of the coil and produce a back-voltage in each turn that will OPPOSE the applied voltage and reduce the voltage being applied to the coil. This will cause less current to flow. During the time when the magnetic flux is increasing (expanding) the current is also increasing and the full current does not flow until the magnetic flux is STATIONARY. When this effect is viewed on a set of voltmeters and ammeters, it appears that the current is LAGGING. In other words it is taking time to reach full value.
This is the delay that creates the timing for the oscillator. The voltage generated across the primary winding at the instant WHEN THE TRANSISTOR IS TURNED OFF, is called a FLYBACK VOLTAGE. The value of this voltage is determined by the inductance of the transformer (coil), the number of turns and the strength of the magnetic flux. In our case we are taking advantage of this energy to charge a battery but if we did not "tap-off" this energy, it would enter the driver transistor as a high-voltage spike and possibly damage it. (A reverse-biased diode can be placed across the winding to absorb this energy).
WHAT? NO VOLTAGE REGULATION?
This simple circuit does not employ voltage regulation. This feature is not needed with a trickle charger. The charging current is so low the battery will never suffer from overcharge. To be of any benefit at all, voltage regulation must be accurately set for the type of battery you are charging. For a 12v jell cell, it is 14.6v. For a 12v Nicad battery, it is 12.85.
This is the way it works: When a battery is charging, its voltage rises a small amount ABOVE the normal voltage of the battery. This is called a "floating charge" or "floating voltage" and is due to the chemical reaction within the cells, including the fact that bubbles are produced. When the battery gets to the stage of NEARLY FULLY CHARGED, the voltage rises even further and this rise is detected by a circuit to shut-down the charger.
A voltage regulated charger is supposed to have the same results. When the voltage across the battery rises to it fully charged state, the output voltage does not rise above this and thus no current is delivered. Ideal in theory but in practice the voltage must be very accurately maintained. If its not absolutely accurate, the whole concept will not work.
In our case we don't need it as the charging current is below the "14 hour rate" and the battery is capable of withstanding a very small trickle current.
WINDING THE TRANSFORMER
The primary winding consists of 45 turns of 0.7mm wire on a 10mm dia ferrite rod. Wind 40 close-wound turns on the rod then 5 spiralling turns to get back to the start. Twist the two ends together to keep the coil in position.
The feedback winding must also be wound in the same direction if you want to keep track of the start and finish as shown in the circuit diagram. It consists of 15 turns spiral wound so that it takes 8 turns across the rod and 7 turns back to the start. Twist the two ends together to keep the coil in position.
The result is called a transformer. It's a feedback or blocking oscillator transformer with a flyback feature. The output is taken across the primary via a high-speed diode. The oscillator will only work when the feedback winding is connected around the correct way. The correct way is shown in the diagram, with the start of the primary and secondary as shown in the diagram. For this to work, both windings must be wound in the same direction. You can keep track of the start and finish of each winding or simply connect the transformer and see if it works. If it doesn't work, reverse the feedback winding (reverse only one winding - NOT both).
Nothing can be damaged by trying this method as the solar panel does not deliver enough current to damage the transistor.
One of the special features of this design is the driver transistor. It is one of the new style of transistors, having a very low collector-emitter resistance (voltage drop) when saturated. It is also capable of handling a very high current (3 amps) and peaks of 20 amps. When used in a high-speed saturation mode such as this, the losses in the transistor are extremely small and it does not require heat-sinking. Other transistors will work but the ZTX 851 transistor added 6mA to the output current due to its characteristics.
Wind the transformer as explained above and have it ready for fitting to the PC board. Fit the other components according to the overlay on the board making sure the transistor and diode are around the correct way. The two electrolytics must also be fitted around the correct way.
Now comes the transformer. As we have already mentioned, the easiest way to fit the transformer is to solder it in position and try the circuit. If it is around the wrong way, the circuit will not produce an output. Reverse one of the windings and the job's done.
A FEW NOTES ON TRANSFORMERS
Transformers are one of the versatile components in electronics. They can be large, small, high-frequency, low-frequency, single winding, multi-winding, step-up or step-down (voltage) high-current, isolating, extremely-high voltage, voltage-reversing or even a combination of any of the above. They can be technically very complex, or very simple to design and you could spend a life-time studying their construction.
On the other hand you can learn how to construct them very quickly. Simply copy a design and maybe modify it a little. By copying a design you "home-in" on the essential features such as wire-size, core size, number of turns etc and you can change any of the features to suit your own requirements.
Before we start, let's point out the two main mis-conceptions of a transformer. Firstly, a transformer only operates on a voltage that turns on and off. This is commonly called AC (it stands for Alternating Current but this also means the voltage is ALTERNATING). The voltage can also be a DC voltage that turns on and off - commonly called chopped DC.
A battery cannot be connected directly to a transformer. It will not work. An oscillator (an oscillator circuit) is needed to convert the DC into pulses.
Secondly, the energy into a transformer (called watts) is equal to the watts output of the transformer (minus some losses). If a transformer on 240v AC (or 110v) produces 240 AMPS output, the output voltage must be low because the maximum input wattage for 240v is 2400 watts. This means the maximum output voltage is 2400/240 = 10 volts. Even though a transformer performs amazing things, it abides by the laws of physics. In general terms, if an output voltage is higher than the input voltage, the current will be lower.