I have built the attached circuit and its being working fine but has some hick ups.

The solar panel is connected to J1 and J2, U1 the LM317 is set to 18 Volts. Q1 is so the current to U2 PB137 is limited to 2 Amps.

MINRLY is to isolate the Battery when the Solar panel is not charging.

The Low voltage cut out made of TL431 and IRF9540N is to make sure the battery does not go below 10.6 Volts.

The terminals J4 and J5 are connected to a relay which the load is switched through.

I am experiencing 2 problems that I cant understand.

The first is the low voltage cut out does not always work when I bench test it there is never a problem, However the battery went flat and when I tested it the voltage was as low as 5 volts.

When the panel starts to Charge the TIP41 gets very hot, in fact so hot the heat sink discolours, the PCB also burns. Again I am not sure why

 

When the panel starts to Charge the TIP41 gets very hot, in fact so hot the heat sink discolours, the PCB also burns. Again I am not sure why

I think you know the answer! There's too much ∆V across the TIP41 and too much current, causing more heat dissipation than the design can tolerate. Some of that current is base current, so perhaps the LM317 is getting hot too? Anyway, you may have to add another transistor or two in parallel.

But I'm a little confused. Your arrangement sends a regulated ~17.3V to the PB137. How does that limit current to 2A? I think you need to arrange a current limiter and not worry about the voltage regulation.

 

I think you know the answer! There's too much ∆V across the TIP41 and too much current, causing more heat dissipation than the design can tolerate. Some of that current is base current, so perhaps the LM317 is getting hot too? Anyway, you may have to add another transistor or two in parallel

But I'm a little confused. Your arrangement sends a regulated ~17.3V to the PB137. How does that limit current to 2A? I think you need to arrange a current limiter and not worry about the voltage regulation.

I regulated the voltage across the PB137 as I was told that voltage above 18 volt would damage it.

The PB137 charges at 1.5 amps.

The TIP41 was there so that a larger current could go to the PB137.

I realise now I need to put a resistor in series with the emitter and input of the PB137.

Now to workout why the battery sometimes does go down to 5v

 

I think you should use a comparator to control the relay.
The PB137 looks like it will continue to output with an input voltage as low as 2 or 3 volts. The 12volt relay is probably not opening until the output of the PB137 goes to 5 or 6v.
Per the spec sheet, the PB137 should be good for input voltages up to 40V. You probably should also build some hysteresis into the relay on/off circuit.

 

I think you should use a comparator to control the relay.
The PB137 looks like it will continue to output with an input voltage as low as 2 or 3 volts. The 12volt relay is probably not opening until the output of the PB137 goes to 5 or 6v.
Per the spec sheet, the PB137 should be good for input voltages up to 40V. You probably should also build some hysteresis into the relay on/off circuit.

The PB137 needs at least 16 volts in order to conduct the purpose of the relay is to isolate the battery from the rest of the circuit when the PB137 is no longer charging the battery

 

Might be helpful to know SP & Battery specs? I've thrown in a low power solar panel battery control ckt. for possible ideas. This has been lighting a 1 W LED lighthouse for over a month now.

 

I can't see any hysteresis in you circuit; Most battery low-voltage disconnects don't reconnect the battery until its voltage is in excess of 12V. Otherwise, the output can oscillate off and on around the threshold.

I would only build a solar battery charger regulator if I had no other option because they are very inexpensive to buy now and include quite sophisticated battery management features; it isn't possible to build a better one for the same price.

 

I can't see any hysteresis in you circuit; Most battery low-voltage disconnects don't reconnect the battery until its voltage is in excess of 12V. Otherwise, the output can oscillate off and on around the threshold.

I would only build a solar battery charger regulator if I had no other option because they are very inexpensive to buy now and include quite sophisticated battery management features; it isn't possible to build a better one for the same price.

I prefer to build my own so I can fix when it goes wrong

 

I prefer to build my own so I can fix when it goes wrong

I know what you mean, it's good to build things