I have a solar panel rated for 3v 150mA which gives out around 3.3v 90mA normally and I'm planning to charge 2x1.2v 1100 mAH NiMH batteries in series during daytime and run 2x 0.5W white leds at night for 3-4 hrs. I realize the batteries wont fully charge and so even 3 hrs is an over estimated value, but I dont know how to run 4v white leds on 2.4v. I have seen circuits for 1.5v, but here I have 2.4v. Any idea what to do?

 

Read up on "Joule thief".

 

2 x 1.2 x 1100mAh = 2.64Wh of energy. Your LEDs require 1W. Therefore the maximum time the LEDs could run in theory is 2.64 hours. In practice, allowing for inefficiencies in the charging and voltage conversion, you'd be lucky to get half that run time .

 

... I have seen circuits for 1.5v, but here I have 2.4v. Any idea what to do?

Those solar light circuits are not terribly sensitive to voltage and should work fine at 2.4V. By "fine" I mean they will boost the voltage enough to light your LED. You will still have to ensure that the LED does not receives too much current.

 

So should I continue with this or add another 3v panel in series and change the battery to a 4v 1200mah to avoid the 'voltage boosting' inefficiency involved and reduce the circuitry which may also allow me to add an additional led? Also is 1 single 1W led brighter than 2x0.5W leds?

 

2 x 1.2 x 1100mAh = 2.64Wh of energy. Your LEDs require 1W. Therefore the maximum time the LEDs could run in theory is 2.64 hours. In practice, allowing for inefficiencies in the charging and voltage conversion, you'd be lucky to get half that run time .

But a normal NiMH cell supplies 1.42v when fully charged and it drops to 1.25v at the minimum usable level and also I believe only 1000mah would be usable from the total cell capacity? Would it affect the output when the voltage is 2.8?

 

Also is 1 single 1W led brighter than 2x0.5W leds?

No. For a given type of LED, light output will be roughly proportional to power consumption.

Boost circuits can be quite efficient; I'd say 80-90% is typical of the commercially available devices. Less for a simple joule-thief I suppose.

 

Would it affect the output when the voltage is 2.8?

Using your new figures for voltage and Ah rating calculate the Wh energy of the cells again and tell us.

 

I decided to go with the present setup after testing out a joule thief circuit. The circuit I tried was this.

The toroid was from a 15W cfl lamp with 7 turns of parallel wire and the transistor used was BC547 and a 0.5W led on the output. The led looks bright but I'm not sure the led current draw measured is correct as it reads only 8mA but the led looks bright, maybe due to the pulsed output. The input to the circuit draws 100mA at 2.55v and the transistor becomes warm which also I do not understand as it is rated for 800mA. How do I control the current to the led in this circuit? Also how could I fix or prevent the transistor from heating up?

 

Test the LED with another source and the appropriate resistor to give you 8mA. I think you'll see it's plenty bright just as you are seeing.

You might try increasing the 1K resistor value until you see a dimming. Not sure, but it might be passing more current than you need. I also think the more efficient joule-theif circuits have capacitors in them.

 

Increasing the number of secondary turns on the core may also help.

 

Yes, just rectify the output and filter with a capacitor. The white LED can go across the capacitor. If you use a large capacitor, the circuit might fail to oscillate.