Low Power Switching Regulator

[Opera House]

I've been working on  couple of switching regulators based on

the MC34063 chip. This is pretty much a self contained unit

except for the switching inductor and diode.  Realistically

these are good for about 3/4 amp.  This project was to build

a small regulator to maintain the batteries when I take the

larger panels in for a couple of weeks.

This was a 5V regulator cut out of a larger circuit board.  

Just love those narrow cutoff wheels for a grinder.  They

make pulling a chunk of circuit easy.  The inductor was

changed to get a little increase in efficiency and timing

capacitor increased to .001uF.  Large Schotkey diode on right

blocks reverse current from battery.  At about 25V input the

efficiency measured about 89% and would run cool at 6W.  The

circuitry on the left turns off the regulator below 24V allowing

the panels to operate around the power point. The MC34063 is a

single input and the zeners and transistor are used to fool the

chip into thinking the output voltage is too high.

There is no current gain with a single panel, input voltage needs

to be above 20V.  In photo I am using a new Solartech 5W 12V and

a used Unisolar 3W 6V in series. These put out .34 and .40 shorted

respectively in the sun.  The solartech I believe is a factory reject.

The voltage really drops with current and only puts out .21A going

directly into a battery without a regulator. Maximum output of the

regulator occurs at

about 23.7V and .17A input to the regulator.  That gives about .31A

going into the battery.  This is exactly what you get with the two panels

in series connected directly to the battery. So, in this case the regulator

only adds over voltage protection.  Defective 12V panel, but it was cheap.

If the 12V panel was good I would be able to get almost 6W out of

this setup instead of the 3.9W I am getting.  Having a problem loading

pictures.  These will follow later.

[Opera House]

The zener from the input voltage basically feeds the base of the NPN transistor.  Collector feeds voltage to chip sense input through a blocking diode from the 300K pull up resistor. When input voltage is greater than zener voltage the chip will regulate the output voltage.  Under zener voltage turns off the switcher.

[independent]

This is similar to what I've done with these chips. I get them out of car cellphone chargers (like the Nokia ones, but other brands work too). Actually, don't get them out, just modify them. The circuit is usually the same, the brand name (not "no name" chargers) have better components and a more importantly a more generous circuit board size. Two resistors on the board are replaced with a 10k pot, et viola a variable voltage adapter. Use it to provide regulated step-down power (from 12v) or make battery chargers. Made chargers for 2x 1.2v Nicads to 8.4v (9v rechargeables). Easy, cheap and available at most 2nd hand stores (junk shops). Stole the idea from a Silicon Chip magazine.

[independent]

More properly quasi-regulated. Not really regulated but if the current draw is not too much it holds within .1v

[Nothing40]

I've found some of the cellphone adapters have a third wire,for sensing the battery voltage I guess? I've put a pot between the + and - outputs,with the wiper to the third wire,and found I could vary the output voltage by about 2V. Pretty simple.

I haven't yet modified one for a larger range like you guys are doing,but I might,someday.

[wooferhound]

We do have some picture posting guidelines you know . . .

[Opera House]

A test of the 12V panel with a watt meter and electronic load gave maximum power

of 2.86W at 14.88V and 197mA.  This is far below the manufacturers 5W @ 17.1V

and .31A as listed on the label. This is clearly a manufacturers reject or

customer return. Sold as a new 3W panel but labeled as a 5W. Visually a couple cells are discolored.

The 6V panel gave maximum power of 2.17W at 7.7V and 292mA. This is about what

I expected since the panel is old and marked SECOND.  Even has a big dent in the

back  that should have cracked part of a cell.

Both panels in series gave an optimal power output of 4.98W at 22.1V @ 230mA.