Buck Controller not working.

[abbyrox]

I am building a solar charger which can charge Li-Ion batteries upto 9V. I am also using a buck converter using LM2596 for a 5V ouput. There is a switch separating the charge circuit and the buck converter. The battery charges perfectly but when I turn on the buck, without any load it brings the battery voltage down to 1.89V. Please help me guys!! I am using 4 solar panels in parallel(6V, 178mA)

 

[FvM]

The schematic looks strange. As shown, the panel terminal SPV- is more or less floating. I don't believe that it corresponds to your actual circuit.

 

[abbyrox]

Yes you're right about the error there, but i have corrected that in my circuit. R4 is in with series with C2 to ground so the panel ground is not floating. I'm sorry i forgot to mention that.

 

[FvM]

Apart from circuit details - there are more dubious points - the exact behaviour of the boost converter part can be only guessed without a simulation.

For the time beeing I'll assume that it's working somehow. But how's the efficiency, does it use a reasonable operation point of the solar panel? How about in- and output voltages and currents without and with buck converter connected?

 

[abbyrox]

The panels don't work at maximum power though. Using two of them wasn't enough so I connected 4 in parallel to get more current. While the battery is charging, the panel voltage is only 0.7V but the battery does charge quickly upto 8.4V (because of two 4.2V zener diodes). The problem earlier was that the schottky diode was connected opposite. Even after correcting that the output was only 1.25V. After some research on the net, I found out that the LM2596 might be an 'adjustable' one. i.e. the output can be adjusted using an external potential divider with Vref as 1.25V(from the IC). After doing that, the buck also worked. Now the whole circuit is working fine. :smile: Thanks for your input anyway FvM!!

 

[FvM]

Good that it works, but what you say about panel voltage suggests that you don't manage to operate the panel at least somewhere around the maximum power point. That's probably a principle problem of your design, so if you are targetting to efficiency, you might be interested to overthink the point.

Determining the I/V characteristic of the panel for the given irradiance by connecting a variable resistor can be a first step.

 

[abbyrox]

Thanks FvM, I'll work on that part.