I have been building this regulator for some years to charge a 12 Volt 7 Ah battery from a 20 Watt Solar Panel. It works quite fine.

In another posting I had Shown C1 and someone pointed out that it was not needed. Originally I had put it into the circuit with the idea of smoothing the incoming supply. If its not needed I will take it out to reduce the size but just want to confirm it is in fact not needed.

 

What are the V/I of solar panel and how is the range of V, and how much voltage do you setup for the charger?

 

It doesn't need to be that large but you should have at least 0.1μF close to the LM317 input (<6") to insure stability of the device (see the data sheet, page 17.).

 

What are the V/I of solar panel and how is the range of V, and how much voltage do you setup for the charger?

The solar panel is

20 Watts
Voc 22volts
Vmp 17volts

I think the current is 1.5 Amps.

I set the regulator to 13.8 Volts

 

The solar panel is

20 Watts
Voc 22volts
Vmp 17volts

I think the current is 1.5 Amps.

I set the regulator to 13.8 Volts

If 17V to 22V are the range then the current should not be that much, check I=W/V.
If the output of LM317 is 13.8V the input Vin_LM317 = 13.8V+3V = 16.8V, Vin=16.8V + 0.9V(diode)=17.7V, so you may try 220uF/35V or change the diode to Schottky barrier diodes and more current and using 100uF/35V, you may do some test to make sure.

 

The diode in series with the battery protects the 317 from reverse current when the solar panel is below the battery voltage, so I think you can eliminate D1 for a little better efficiency.

ak

 

You can also eliminate C2.

 

According to the battery charger example of Figure 10 of the ST Microelectronics November, 199 LM317 datasheet, no capacitors are needed.

 

I set the regulator to 13.8 Volts

With a diode dropping voltage between the regulator and the battery, the regulator should probably be set higher. Depends on the battery and the temperature, but many should see 13.8V at the battery terminals.

I believe that it would be better to place D2 in reverse bias across the regulator. This would take it out of the power path and eliminate that loss, but still protect the regulator from a battery voltage in excess of the regulator's input voltage, which could damage it. D1 would still be needed as the blocking diode.

 

But that would try ti send charge back into the solar panel array, which probably is less tolerant of that than the alternator in a car (the originator of the "load dump" protection diode). Also, C2 is a tradeoff. It lowers noise on the output, but decreases transient response.

Let me stick my neck *way* out here, and slightly disagree with DC. The ST device might be slightly different inside, allowing it to operate without any caps. I vote for the original National Semi datasheet as the reference for the 317, and it has input and output caps. Low cost, can't hurt, cheap insurance, and they do improve performance even if not needed for stability.

ak

 

A brief check shows the same circuit are also in both of the versions of the LM317 datasjeets on the TI website (National's URL refers to TI now). I am not recommending following that advice, only taking note of what's in the datasheets.

 

I checked the TI sheet and came to the same conclusion. For this application, with a steady supply and a steady load, the caps are not needed.

 

But that would try ti send charge back into the solar panel array, which probably is less tolerant of that than the alternator in a car ....

As I said in #9, D1 would still be needed as the blocking diode.