Pulse Charger Circuit for PVC to Battery

Picture is low resolution and component values cannot be seen.

The low resolution of the image makes it difficult to say for sure but I suspect the series of signal diodes provide a temperature stability or response characteristic. The pulse sequence looks like charge gets stored on the output capacitor until Vgs crosses the threshold voltage of the MOSFET. This discharges the capacitor and then repeat.

Circuit diagram is here: https://www.dropbox.com/s/hlo6e8f82n2kw2o/pulse%20charger.jpg

Sorry, I must be missing something, but I don't see how it works.

The circuit DOES NOT WORK.

Made some mistakes reverse engineering the pcb. Q3 is the other way up and the configuration of Q4 is different. Please see revised diagram at https://www.dropbox.com/s/hlo6e8f82n2kw2o/pulse%20charger.jpg

When the supply rail rises to 5v6 + 0.6 + 3v = 9v2 Q3 will start to turn on.
Q1 will never have any effect.
Q2 will turn on and Q4 will turn on.
Q4 will finish up with a voltage across it that is approx 9.2v minus about 5.5v of the battery.
There is no current-limiting and the charging current is unknown. All this can be done with a simple RESISTOR.

Q1 will certainly have an effect since it is part of the feedback path with Q3 that linearly regulates the current through the emitter of Q2. You do the OP a disservice by trivializing this circuit. This is a more sophisticated battery charger circuit than just a resistor. For instance your explanation does not mention at all Vgs of the MOSFET Q4 and how the transistor works in its linear or saturated region.

I don't have the time nor need to analyze this circuit fully but it is more sophisticated than you claim.

The supply will have to reach 12v for Q1 to have any effect. It will never reach 12v.
The circuit can be replaced with a single RESISTOR, as I said.
The circuit simply delivers all the current the solar panel can supply. It has no current-limiting. At least a resistor will provide current-limiting, if required.

One can easily obtain 12V with multiple solar cells in series. Notice also the slope of the VI curve line in the fourth quadrant is not parallel to the current axis because as more (usable) flux (photons per second) strike the surface the voltage does increase. As for your current limiting concerns, a MOSFET basically acts as either a resistor, constant current device or an OFF switch depending on an N channel MOSFET Vgs being above Vgsth and the voltage and current combination. between drain and source.

This is one of the worst battery charger circuits I have seen.
To start with, the PNP transistor, 5v6 and string of diodes puts a zener voltage of 0.6 + 5v6 + 3v on the supply rail and if the battery is removed and the supply rises above 9.2v the transistor will be DESTROYED !!
The whole circuit is equal to a current-limiting resistor of unknown value and the circuit will effectively deliver all the current from the solar panel to the battery.
If the solar panel can deliver 20 amps, this current will flow into the battery and BLOW IT UP.
Let's see what the circuit does:
As soon as the supply reaches 9.2v, the PNP transistor turns ON and this turns ON the emitter-follower transistor Q2 and raises the gate voltage on the MOSFET.
The supply only has to rise a few more millivolts and the PNP transistor is fully turned ON and the MOSFET will be fully turned ON. So the difference between the MOSFET being not-turned-ON and fully turned ON is only a few millivolts change in the voltage on the supply rail.
Now let's look at the MOSFET.
When (if) it turns on fully, the battery is connected directly to the solar panel and the supply rail is 4.8v plus the "floating voltage" produced by the battery when it is charging.
This means the battery voltage can be as high at 5.5v.
This is not high enough to get the circuit to turn on, so the MOSFET allows the difference between 5.5v and 9.2v to be dropped across it.
This is purely a VOLTAGE DROP and you can consider the voltage drop to be identical to a zener and the MOSFET will not limit the current.
We do not know the voltage of the solar panel or its current capability and this will be the limiting factor of the circuit.
But as soon as you remove the battery, the transistor will blow up.
The first transistor only starts to come into operation when the supply reaches 5v6 + 5v5 + 0.2 + 0.6v = 12v and it is clear that the supply can NEVER go above 9.2v without blowing up the circuit.
Thus the whole circuit can be replaced with a current limiting resistor and the battery will not be destroyed.

I suspect there are additional drawing errors. As the PV supply exceeds the Q3 BEvoltage of .7 volts, plus the zener D7 and diode string approx 3.5 volts or a supply voltage of .7+5.6+3.5 = 9.8 volts, then the BE current on Q3 is uncontrolled.

If it was a NPN instead?

The circuit as shown looks non-functional to me.

Thanks to all for comments so far.

I need to reverse engineer the pcb again, just for everyone's peace of mind. The last time I did this I 'discovered' a third layer inside the board that made a connection I hadn't noticed before ...

I'll also be wiring it up to a storage 'scope to do some measurements...

Multi layer boards are really hard to "trace".

I have enough problems with 2 layer - and house numbers on parts never helps.