Pulse Width Modulation regulated flashlights

[shankus]

Since my eyes can't tell the difference between a constant on light, and a rapily pulsing one, isn't this the next step to take for regulated lights to achieve maximum runtime and efficiency? I mean to say, some lights are using PWM, but widespread use is what I'm referring to.

What lights, other than the eternaLights use PWM?

 

[IsaacHayes]

what freq do these things run on? it's gotta be at least 85hz or I'd notice the flickering. Also wouldn't that cause adverse effects if shined at something moving, especially rotating...

 

[Ray_of_Light]

PWM modulation turns the lamp or LED on and off continously, and is a very efficient way of controlling brightness. Among the PWM controlled light, I know the PT Aurora and Matrix. The switching frequency is around 250 Hz.

Anthony

 

[Doug Owen]

My new PT Eclipse dose as well. I suspect all the other 'key chain lights' that dim do.

It's easy enough to spot, wave the dimmed light around in the dark. A couple foot long piece of string can be kinda fun....for a while.

Doug Owen

 

[IsaacHayes]

ah, that high won't be perceveable. Do they use this PWM to overdrive the led more than they could with constant on?

 

[Doug Owen]

[ QUOTE ]
IsaacHayes said:
ah, that high won't be perceveable. Do they use this PWM to overdrive the led more than they could with constant on?

[/ QUOTE ]

Hey, please don't try to tell me what I didn't see.

And I just checked my Derringer and Eclipse in a dim room without the string trick. Both are clearly visible.

I assume you're guessing? That is have never tried it?

And no, I know in the case of the Derringer at least bright is 100% duty cycle, mid about 30% and Dim a bit over 5% 'cuz I measured them. FWIW there's no energy storage element (usually an inductor) in the Derringer, just the IC, switch and a single cap across the battery.

To be technically correct, we probably shouldn't call it a PWM controller, but a duty cycle one (it's open loop, right?).

Anyway dimming is in fact done by on and off switching, at rates you can observe (at least for these two lights). At least from what I've *seen*.

Doug Owen

 

[shankus]

[ QUOTE ]
IsaacHayes said:
Do they use this PWM to overdrive the led more than they could with constant on?

[/ QUOTE ]
This is what I'm thinking. The brilliance could be maintained (or overdiven), while saving capacity with off-time.

I can see the effects in my eternaLight, as well. I measured the rate again, and it is 100-166 Hz, depending on the dim level, with the duty cycle varying.
It is very noticeable. I notice it when I go into dim mode, and the ceiling fan shows the effects of the "strobing".

 

[Doug Owen]

[ QUOTE ]
shankus said:
[ QUOTE ]
IsaacHayes said:
Do they use this PWM to overdrive the led more than they could with constant on?

[/ QUOTE ]
This is what I'm thinking. The brilliance could be maintained (or overdiven), while saving capacity with off-time.
.

[/ QUOTE ]

Again, I don't think this is the case. At 'full blast' the duty cycle goes to 100%, right?

There's a solid reason not to do what you're suggesting (seriously overdrive for part of the time, keeping the average lower) if you're concerned about battery life. Check the battery specs, total energy available goes down very fast at high current levels. LEDs also have some failure modes that respond to peak (not average) currents.

But, to get back to the point, none of the lights in question seems to use this technique, they seem to have 100% duty cycle at 'full blast'.

Doug Owen

 

[shankus]

What I'm thinking, is a smaller duty cycle, say for instance, 50%. But the current during the ON time to be above spec, so that the light appears as bright as the same luxeon at direct drive, but it is consuming only 50% of the power that it would at the same current directly driven.

I think we're talking about two different things. I'm considering battery efficiency from cycling the luxeon off, not using the circuit as a dimmer.

 

[highlandsun]

What do you expect to gain by doing that? P=VI, E=PT. Twice the current for half the time is still the same amount of energy. And since Luxeons are most efficient at *low* power, you will actually get a dimmer result than if you had just used the rated power level for 100% of the time.

PWM is good *for dimming*. When you want to run at full brightness, you don't need to (and shouldn't) play any pulse-shaping tricks.

 

[Empath]

[ QUOTE ]
highlandsun said:
What do you expect to gain by doing that? P=VI, E=PT. Twice the current for half the time is still the same amount of energy. And since Luxeons are most efficient at *low* power, you will actually get a dimmer result than if you had just used the rated power level for 100% of the time.

PWM is good *for dimming*. When you want to run at full brightness, you don't need to (and shouldn't) play any pulse-shaping tricks.



[/ QUOTE ]

If you drive it at twice the power but only half the time you still get the same number of photons as driving it direct and constant. There are no free rides.

 

[Doug Owen]

[ QUOTE ]
Empath said:

If you drive it at twice the power but only half the time you still get the same number of photons as driving it direct and constant. There are no free rides.


[/ QUOTE ]

While this is generally true, I think that most LEDs are more efficient (that is more photons per Watt) at lower current levels. That is at a 50% duty cycle you might need to drive at 70 or more mA to get the same level of output over time as a constant 30 mA.

More importantly, the batteries we use don't do as well at high currents. Take the Alkaline AA, the Eveready graphs show about 5 hous to .8 Volts for 350 mA drain (about half the total energy you get at 25 mA), but only 1.5 hours at twice that (meaning 3 hours by 'double driving' for half the time). Even if the LED output was a wash, it's clear this is not a good way to go if the goal is long battery life. The key to long battery life (that is getting the maximum usable energy from the battery) calls for current rates as low as possible. This means continuous rather than pulsed drive.

Not only no free lunches, this one's more expensive.......

Doug Owen

 

[highlandsun]

[ QUOTE ]
Doug Owen said:
Not only no free lunches, this one's more expensive.......
Doug Owen

[/ QUOTE ]

Thanks, that was exactly my point.

 

[Entropy]

[ QUOTE ]
Doug Owen said:
[ QUOTE ]
IsaacHayes said:
ah, that high won't be perceveable. Do they use this PWM to overdrive the led more than they could with constant on?

[/ QUOTE ]

Hey, please don't try to tell me what I didn't see.

And I just checked my Derringer and Eclipse in a dim room without the string trick. Both are clearly visible.

I assume you're guessing? That is have never tried it?

And no, I know in the case of the Derringer at least bright is 100% duty cycle, mid about 30% and Dim a bit over 5% 'cuz I measured them. FWIW there's no energy storage element (usually an inductor) in the Derringer, just the IC, switch and a single cap across the battery.

To be technically correct, we probably shouldn't call it a PWM controller, but a duty cycle one (it's open loop, right?).

Anyway dimming is in fact done by on and off switching, at rates you can observe (at least for these two lights). At least from what I've *seen*.

Doug Owen

[/ QUOTE ]
PWM is PWM, whether closed loop or open. PWM and duty cycle variation are essentially the same thing.

There is *one* advantage of increasing the current and reducing the duty cycle - While white LEDs are much less susceptible to color changes when dimmed by current reduction as compared to incandescents, some LEDs (esp. Luxeons) will change color somewhat as the current is increased/decreased. Underdriven Luxeons are more likely to be a "greenie", overdriven ones more likely to have a bluish tint. (I'm guessing due to nonlinear responses of the phosphor.)

As such, if you have a LED with slight green tendencies, you may be able to fix it by increasing current and decreasing duty cycle. But doing it for the sake of efficiency won't work.

 

[shankus]

I didn't say twice the current for half of the time.

[ QUOTE ]
shankus said:
duty cycle...say for instance, 50%. ...the current during the ON time to be above spec, so that the light appears as bright as the same luxeon at direct drive, but it is consuming only 50% of the power that it would at the same current directly driven.


[/ QUOTE ]

I think it would be an interesting circuit to play with, and take measurements of.
Is it asking for a free ride to turn off the light when you aren't seeing it?