Joule Thief

[synchro1]

But, I use very little current.  Only high voltage at a high freq.  This is why my lights last so many hours on "dead" AA batteries.

Now, I know this is a little bit of a parlor trick in that the leds are flashing on/off at about 30,000 times/sec.  So, if the duty cycle is close to 50%, then they are off as much as they are on.  But, our lights in our homes flash on/off at 60 times/sec and no one notices this so, I do not think of this as a real trick.

As far as calculations go...I suck at math and I only observe what my lights can do.....I do very little predicting.  If/when any of my leds actually burn out, I will let you know.  I did fry my share of leds back in my early days, the smell was terrible.  But, that was long before I discovered (not invented by me) these high voltage/high freq. circuits.

Bill

@Pirate88179,

Here's how it really Works:

lumen/Watt efficiency improves with decreasing current. I have more videos, but if you doubled your number of bulbs and lowered your voltage and frequency by half, you would generate more lumens per watt. Et cetera! The relationshhip is not directly proportional. Cutting power in half and doubling the bulbs would result in less overall light, but much higher efficiency. You could probably get the same light with twice the bulbs for 10-20% less input.

[MarkE]

Output (lumen) per Watt is even worse than lumen per mA. As mA increase Vf also increases so the Vf x I product increases at a faster rate per lumen than just I does. So, again, maximum lumen/Watt is achieved at low mA compared to rated mA and lumen/Watt efficiency improves with decreasing current.

I²R loss is no secret.  Minimizing rms current in an LED does yield maximum lumens/Watt.  That means high duty-cycles, or better yet DC drive is more efficient than the same average current at a lower duty-cycle.  However, from a color rendition standpoint, it is better to use the same peak current and modulate the duty-cycle than it is to drive with different DC current levels.  Then there is the problem of the driving circuit efficiency.  Those cheap garden stick lights use circuits that tend to be very efficient delivering power to the LEDs but do so in pulses with very poor form factors into cheap LEDs, so that the overall input power to light output is nothing to write home about.

[MarkE]

There's two ways to cause problems increasing lumens in LEDS, both involve running too much current through them: Firstly; By raising the voltage which raises current draw, and Secondly; Raising the frequency with the same result. Running too much current through the LED wastes power and shortens the bulb's lifespan. Most of the apparent savings from increasing frequency and cutting duty cycle come from miscalculations. There's no way to beat the peak efficiency watts per Lumen ratio that the bulbs are rated for. The bottom line is, you're better off increasing luminosity by adding extra bulbs to the array and lowering the current input. I demonstrated that in my multiple LED video:

https://www.youtube.com/watch?v=IP39OoaSQb0

Raising frequency increasing switching loss in LEDs that are pulsed, and in the switching circuitry that pulses them.  But running LEDs from rectified switching supplies, the storage components get smaller and cheaper with high switching frequencies.  Hard switching converters are available that are very efficient:  upwards of 95% while switching at 1MHz to 3MHz.  It is true that more LEDs in parallel sharing the same current reduces I²R loss.  However it is also true that unless the LEDs are bin sorted, that brightness variation can be a problem.  There are lots of ways to go wrong with LED efficiency.  The: design of the luminaire, choice of the LED(s), and amount of heat sinking all play big roles before one ever gets to the driver electronics.

[MarkE]

@Pirate88179,

Here's how it really Works:

lumen/Watt efficiency improves with decreasing current. I have more videos, but if you doubled your number of bulbs and lowered your voltage and frequency by half, you would generate more lumens per watt. Et cetera! The relationshhip is not directly proportional. Cutting power in half and doubling the bulbs would result in less overall light, but much higher efficiency. You could probably get the same light with twice the bulbs for 10-20% less input.

It really depends on where one is on the V-I curve for the particular LEDs that one is using.  When 1W capable LEDs were $5. each a few years ago, using multiple LEDs was cost prohibitive.  These days there are many applications where it is actually cheaper to use multiple smaller LEDs and one gets the benefit of higher efficiency.  The catch is matching the LEDs for an acceptable appearance.

[SeaMonkey]

But, I use very little current.  Only high voltage at a high freq.  This is why my lights last so many hours on "dead" AA batteries.

Do you see the 'problem' or apparent contradiction
in your statement above?

...
As far as calculations go...I suck at math and I only observe what my lights can do.....I do very little predicting.  If/when any of my leds actually burn out, I will let you know.  I did fry my share of leds back in my early days, the smell was terrible.

The above statements partially explain the
apparent contradiction.

But, that was long before I discovered (not invented by me) these high voltage/high freq. circuits.

Those of us who are experienced in the 'art'
understand what you're driving at but others
may wonder what you mean by "high voltage"
in the above quote since you've already explained
that the source of energy is "dead" AA cells.

Where is the high voltage?