Joule Thief

[Pirate88179]

The NCP1402 is a good choice provided that you have a current limiting mechanism on the load side.  If you do not then the LED current can be highly variable and other problems can occur.  One way around the problem would be to steal the circuitry out of one of those garden stick lights.  They are limited int he total power that they can deliver, but they can generally operate to pretty low input voltages, are efficient and drive the LEDs with a current instead of a voltage.

Mark:

Thanks.  In my earlier videos of my attempts, I used the garden light circuitry.  It was fine for lighting one or two leds but that was about all it could handle.  I still have about 40 of those chips here.  Also, they only operated down to about .85 volts, which is not bad, but this chip is much better in my opinion.

I left the resistors installed on the lights circuit board to keep the amp draw below my 200 mA limit with this board.  One is for the 24 led circuit and the other is to limit the power from the 3 AAA's when only lighting the three leds.  So far so good.  I may need to increase the resistor for the 24 leds if I run into problems during testing but, so far, so good.

In the first puck light that I used this circuit board in, I removed the resistor as there were only 3 leds and it was fine. (The 3 AAAs at, 4.5 volts is higher than my 3.3 volt output from the board) Lighting 24 leds is another story which is why I will be testing the heck out of this light.

Thanks,

Bill

[dvy1214]

Might be one of the most efficient ways to run LED lights from an on site power supply. Like it!

[Pirate88179]

Might be one of the most efficient ways to run LED lights from an on site power supply. Like it!

Thank you very much.  This chip is rated at 90% efficiency, which is much higher than your average JT circuit.

Bill

[MarkE]

Mark:

Thanks.  In my earlier videos of my attempts, I used the garden light circuitry.  It was fine for lighting one or two leds but that was about all it could handle.  I still have about 40 of those chips here.  Also, they only operated down to about .85 volts, which is not bad, but this chip is much better in my opinion.

I left the resistors installed on the lights circuit board to keep the amp draw below my 200 mA limit with this board.  One is for the 24 led circuit and the other is to limit the power from the 3 AAA's when only lighting the three leds.  So far so good.  I may need to increase the resistor for the 24 leds if I run into problems during testing but, so far, so good.

In the first puck light that I used this circuit board in, I removed the resistor as there were only 3 leds and it was fine. (The 3 AAAs at, 4.5 volts is higher than my 3.3 volt output from the board) Lighting 24 leds is another story which is why I will be testing the heck out of this light.

Thanks,

Bill

Bill, that will do at the cost of ultimate efficiency.  I think that there is an answer with the NCP1402 but have not modeled or tried it out.  The one piece of advice that I can give you is to always check to see if the regulator is outputting its intended voltage.  If it outputs less, such as 3V instead of 3.3V then you will find that the efficiency is way off.

[Pirate88179]

Well, in the reviews of this board, I did read that as the input voltage drops down below .8 volts, the mA's drop as well.  So, it is not reality to expect 200 mA's down at .3 volts.  The cool thing about this type of light is that we now have 3 AAA batteries to supply the power.  Unless I am wrong, since this now will run on a single AAA, adding a second one should double the run time, and a third should triple it...give or take.

So, I agree with you that as the input depletes, the efficiency will also become non-linear as does the power output.  This is probably why the Sparkfun folks rate this board down to .8 volts.  Below that...I am sure that all of the numbers go to hell.  The NCP 1402 chip will switch down to .3 volts but, depending on the application, you might never get there. (I suspect flashing as we get to a lower v. input below say .5 v. These 24 leds are a large load.  I wish there was a commercial light I could modify that had like 10 leds.

As I said, I will do my testing.  I believe no matter what happens, this light is now better for giving off light for a longer period than it was designed for.  How much longer remains to be seen.

Bill