effect of ON time on lumen
- Thread starter sagar
- Start date
Help Support Candle Power Flashlight Forum
AnAppleSnail
Flashlight Enthusiast
If you PWM (30% duty cycle, etc) the LED's current supply, you get a nearly linear response.
If you change the current reaching the LED at 100% duty cycle, you get a bit higher efficiency at low current, but still nearly linear.
The long-term life span is not much affected, as long as the thermal needs of the LED are met.
If you change the current reaching the LED at 100% duty cycle, you get a bit higher efficiency at low current, but still nearly linear.
The long-term life span is not much affected, as long as the thermal needs of the LED are met.
DIWdiver
Flashlight Enthusiast
Absolutely agree. That is, at constant peak current, light output (lumens) is proportional to duty cycle. At reduced current, light output is a bit better than proportional to current.
I disagree with this. At 30% the operating temperature is likely to be substantially reduced, thus increasing the lifespan of the LED considerably. This will be very noticeable if you buy cheap LEDs and run them a lot. If you buy good LEDs and/or don't run them much, you may well scrap the lights before noticing a big drop in output (gradual loss of output is the thing that usually determines the 'life' of an LED).
Last edited:
ShineOnYouCrazyDiamond
Flashlight Enthusiast
I don't think that is has been clarified why to use PWM versus constant current.
PWM is pulsing the LED at 100% rated current (or around there) at a very fast rate (say 5000+ Hz). The benefit is a more consistent tint across the power range as you dim the LED.
If I am interpreting your question your are talking about running the LED at 30% PWM and then lowering the current? So you want to run PWM with a less than rated current? That's what I am reading. I think you would start to get tint shift as you lower the current regardless of PWM.
PWM is pulsing the LED at 100% rated current (or around there) at a very fast rate (say 5000+ Hz). The benefit is a more consistent tint across the power range as you dim the LED.
If I am interpreting your question your are talking about running the LED at 30% PWM and then lowering the current? So you want to run PWM with a less than rated current? That's what I am reading. I think you would start to get tint shift as you lower the current regardless of PWM.
DIWdiver
Flashlight Enthusiast
OP didn't say anything about lowering current. That was AnAppleSnail.
Or maybe I am misinterpreting...
Sagar, could you clarify the question for us?
ShineOnYouCrazyDiamond
Flashlight Enthusiast
This was the OP
Maybe I mis-interpreted, but that's how I read it. Would like to know how the OP meant the question because I always find PWM an interesting topic and I tend to learn something new every time.
RoGuE_StreaK
Enlightened
I also interpret this as meaning reducing current; if using PWM, you aren't providing lower power, you are providing full power for a short period?
Something to also bear in mind is that running higher currents also means a higher forward voltage; say for example you are using an XM-L2, you could 30% PWM it with 3Amps, or run it 100% with 900mA. @3Amps, theoretically you have a Vf of ~3.35V, whereas @900mA you have a theoretical Vf of ~2.9V
output wise, let's say a U2 bin
@3Amp = 325% x 300lm x 30% = 292.5lm
@900mA = 125% x 300lm = 375lm
The spec sheet is your friend, at least for getting a rough idea of what the outputs and requirements should be
DIWdiver
Flashlight Enthusiast
I see your point, but that's not how I read it. Perhaps we should wait for the OP to clarify.
If i give 30% duty cycle then for 30% it will take full power that is same voltage and current.
but is off for 70%.So power delivered to LED is averagely 30% right???
Then is there any affect on current???
As i think i am delivering same current for LED for 30%.
And again i have doubt that will it change color temperature.
but is off for 70%.So power delivered to LED is averagely 30% right???
Then is there any affect on current???
As i think i am delivering same current for LED for 30%.
And again i have doubt that will it change color temperature.
AnAppleSnail
Flashlight Enthusiast
30% duty cycle averages out to 30% current.
Strangely, PWM affects the LED's output less than Current Control, especially at very low currents. In short, the photons leaving the LED tend to become more tinted as current decreases (At any duty cycle). So a current-controlled LED below 1% of its rated power will generally become tinted enough to notice. A PWM-driven LED will not tint-shift so much. This tint-shift can be reduced by using PWM, optically combining several LEDs, or picking very nice ones.
DIW: I agree that I was incorrect about the effects of underdriving LEDs. Who knows why I wrote the sentence so wrongly?
ShineOnYouCrazyDiamond
Flashlight Enthusiast
From the readings I have done on other threads about PWM it seems the general consensus is that PWM avg. current doesn't affect tint. Now will there be a subtle difference that can be measured with equipment - probably. But can it be seen with the eye, maybe by some. I know in the PWM lights I have I can't tell the difference - the tint is pretty consistent to my eyes.
The theory behind this would suggest it to be true as well. With PWM the LED is driven in a binary sense. A 0 or 1. Either 0% or 100%. Regardless of level, when the LED is driven, it is always driven at 100%. The but here is that every electronic component has a ramp up and decay time. No matter how small the duration it still takes a certain period of time to ramp up fro 0 to 100% and then decay from 100% to 0%. I would imagine during that time there is a bit of tint shift. As the level of the light get lower the duration of the 100% on time shrinks and shrinks - there has to be a point where that ramp up and decay time becomes a significant part of the the actual time on - and that is where I think you'd see tint shift.
This doesn't take into account the frequency of the PWM though. Some awful lights use 100Hz PWM at levels and it just looks like a horrible strobe show if you move the light around. But, with a PWM this low I don't think the ramp up and decay times would matter. For higher PWM frequencies (and IMO 4,000Hz is really a minimum for quality PWM - my eyes are sensitive enough to see the PWM on anything less) with shorter cycles lengths there may be more impact at lower levels.
But - boy have we digressed.
The OP asked about lumens output - but this is still an interesting discussion.
The theory behind this would suggest it to be true as well. With PWM the LED is driven in a binary sense. A 0 or 1. Either 0% or 100%. Regardless of level, when the LED is driven, it is always driven at 100%. The but here is that every electronic component has a ramp up and decay time. No matter how small the duration it still takes a certain period of time to ramp up fro 0 to 100% and then decay from 100% to 0%. I would imagine during that time there is a bit of tint shift. As the level of the light get lower the duration of the 100% on time shrinks and shrinks - there has to be a point where that ramp up and decay time becomes a significant part of the the actual time on - and that is where I think you'd see tint shift.
This doesn't take into account the frequency of the PWM though. Some awful lights use 100Hz PWM at levels and it just looks like a horrible strobe show if you move the light around. But, with a PWM this low I don't think the ramp up and decay times would matter. For higher PWM frequencies (and IMO 4,000Hz is really a minimum for quality PWM - my eyes are sensitive enough to see the PWM on anything less) with shorter cycles lengths there may be more impact at lower levels.
But - boy have we digressed.
The OP asked about lumens output - but this is still an interesting discussion.
DIWdiver
Flashlight Enthusiast
Yes, the average power over time is 30% of what it would be at 100% duty cycle. The instantaneous power is either 100% or 0, depending on when you look.
Which current are you asking about? The input current to the driver will be on and off, just like the output current. During the on time, the input and LED currents should be the same as they would be at 100% DC. During the off time, the input current should be very small (essentially zero) and the LED current should be zero. For estimating forward voltage and tint shift, you should use the instantaneous current. For calculating heating and battery loading, you should use the average current.
This is how PWM differs from 'analog' dimming, where the current is on all the time but the current value is changed. When using analog dimming, you use the same current value for all estimates and calculations. This is why the tint doesn't change when using PWM dimming, but it does when using analog dimming.
PWM dimming should not change the color spectral output of the LED. Analog dimming WILL change it, but the effect is pretty small until you get to significantly lower output. Depending on the LED and the application, you may get well below 10% before noticing much change.
You should be careful to select a driver built for PWM dimming, and select the proper frequency. Some drivers don't work well with PWM at all, and some don't like high frequencies.
