I know PWM, varying the voltage and current.... right? I am wondering, I have at least 1 light that always seems to have the same tint, regardless of the brightness, that being my V10R .... Actually the HDS Rotary seems to be the same way... but I have other lights that the tint shifts..... Which I assume is a change in the voltage..... does keeping the voltage the same, and changing the current allow a light to dim/brighten without changing the tint?
What techniques are used to adjust brightness?
- Thread starter Doug
- Start date
Help Support Candle Power Flashlight Forum
Voltage and current are interdependent upon each other, you cannot change one without the other also changing.
For example on this led 'Vf vs I' graph the operating characteristics can be seen.
If you adjust Vf or I, the non controlled variable will change so as to follow the curve of this graph, or whichever graph corresponds to your led.
For example on this led 'Vf vs I' graph the operating characteristics can be seen.
If you adjust Vf or I, the non controlled variable will change so as to follow the curve of this graph, or whichever graph corresponds to your led.
kaichu dento
Flashaholic
I'm really interested in following this post so I can understand a bit more about this phenomenon too. I hope that as progress is made that I can eventually have a light that will go cooler as it gets lower and warmer on higher settings.
Unfortunately that is the opposite of what LEDs currently do. They get cooler at higher output levels, and warmer at lower.
See the updated xm-l datasheet. (see the added charts [page 5-7] of the change of x&y chromatic coordinates versus output & temperature)
http://www.cree.com/products/pdf/XLampXM-L.pdf
Although that is for strictly current controlled lights. HKJ can help out as to what circuits what lights have.
Tint shift = variable current
no tint shift, just brightness = PWM, I.E. an XM-L might get 3A when its on, but is only 1% duty cycle(on for 1% of the time), so its 30mA average, but with the same tint as full power.
Voltage has nothing to do with it, as the LED should be driven in a constant current manner
no tint shift, just brightness = PWM, I.E. an XM-L might get 3A when its on, but is only 1% duty cycle(on for 1% of the time), so its 30mA average, but with the same tint as full power.
Voltage has nothing to do with it, as the LED should be driven in a constant current manner
kaichu dento
Flashaholic
Yeah, I was really disappointed when I first found that out, but hope dies hard!
Spent a bit of time on the phone today with DaFab and when the subject came up he suggested that eventually all these things will be possible - but when is anybody's guess...
Bimmerboy
Flashlight Enthusiast
Doug - Your assumption that a constant current driver (the type that allows tint shift) changes voltage across the LED is correct. It's by very small amounts, but that's what a constant current driver does. It's essentially a voltage regulator, but with a sense resistor to determine proper Vout. It 'senses' the amount of current passing through the circuit, and adjusts the voltage output accordingly, which in turn allows the desired amount of current to pass through the LED... which leads to your question.
Given the nature of LED's, it's impossible to change current while keeping voltage the same, as shown by 2filthy3's graph. They are interdependent, hence the design of constant current drivers.
CKOD, while incorrect about the voltage thing, is correct in his explanation of duty cycle with a PWM driver. However, the other important variable with pulse width modulation is frequency. So, there's the pulse width, the percentage of on-time per pulse, and frequency, the number of pulses per second. Manipulating these variables are basically just functions of power delivered to the LED over time, and "tricks" the eye into perceiving different brightness levels.
And finally, the reason for the tint consistency with PWM is the driver pulses the LED with 100% of the desired current during the on-time, therefore tint remains the same at all duty cycles. Any perceived change in tint is an eye/brain thing.
KD - I think for that to happen won't be possible with LED's as we know them, or at least the phospor compounds presently used. Not to say it can't be done with a different material though... it probably can! We just have to figure out how to make it... *insert mad scientist smiley*
Given the nature of LED's, it's impossible to change current while keeping voltage the same, as shown by 2filthy3's graph. They are interdependent, hence the design of constant current drivers.
CKOD, while incorrect about the voltage thing, is correct in his explanation of duty cycle with a PWM driver. However, the other important variable with pulse width modulation is frequency. So, there's the pulse width, the percentage of on-time per pulse, and frequency, the number of pulses per second. Manipulating these variables are basically just functions of power delivered to the LED over time, and "tricks" the eye into perceiving different brightness levels.
And finally, the reason for the tint consistency with PWM is the driver pulses the LED with 100% of the desired current during the on-time, therefore tint remains the same at all duty cycles. Any perceived change in tint is an eye/brain thing.
KD - I think for that to happen won't be possible with LED's as we know them, or at least the phospor compounds presently used. Not to say it can't be done with a different material though... it probably can! We just have to figure out how to make it... *insert mad scientist smiley*
So, the only way they are adjust brightness is with PWM? My HDS rotary must be using a pretty high frequency.... But it does look like it's pulsing at the lower settings, the IFE1 I just got appears to have the same tint through it's adjustment range as well....
kaichu dento
Flashaholic
Understanding the process that determines output tint makes me wonder if one possibility for making the tint warmer on high and cooler on low would be to combine PWM and voltage regulation. I know it sounds complex, but the level of complexity in our present crop of lights already surpasses in many ways the room-filling computers of yesteryear.
In any case, I love cool tints at lower levels and warmer tints as output goes up and hope I'm not the only one.
Bimmerboy
Flashlight Enthusiast
Doug - Not sure I understand your last question. Could you re-phrase that? Just wanna' make sure I answer properly.
KD - It's extremely cool on MANY levels that the computing power of, say, the diminutive LF2XT is probably greater than the UNIVAC! In the grand scheme of things, that happened very quickly, and has produced wondrous things, which makes it easy sometimes to hope that computational speed could over-ride physical law. But alas, as always, there are the natural properties of materials. "Nature, to be commanded, must be obeyed" as the famous quote goes (damned if I remember who said it!). So goes the presently designed LED.
In order to do what you're looking for (which actually sounds like a pretty cool idea BTW), we have to turn to different emitter materials and designs. Combining PWM with voltage regulation is exactly what a PWM driver already does. It allows a timed pulse of a pre-determined current to happen.
That certainly doesn't provide an easy answer, but hopefully it'll give the peace of mind of not wanting to force something against it's nature, and help with a perspective of looking toward different technologies to achieve your desired goal... one which some researcher may have already thought of, and is working towards!
KD - It's extremely cool on MANY levels that the computing power of, say, the diminutive LF2XT is probably greater than the UNIVAC! In the grand scheme of things, that happened very quickly, and has produced wondrous things, which makes it easy sometimes to hope that computational speed could over-ride physical law. But alas, as always, there are the natural properties of materials. "Nature, to be commanded, must be obeyed" as the famous quote goes (damned if I remember who said it!). So goes the presently designed LED.
In order to do what you're looking for (which actually sounds like a pretty cool idea BTW), we have to turn to different emitter materials and designs. Combining PWM with voltage regulation is exactly what a PWM driver already does. It allows a timed pulse of a pre-determined current to happen.
That certainly doesn't provide an easy answer, but hopefully it'll give the peace of mind of not wanting to force something against it's nature, and help with a perspective of looking toward different technologies to achieve your desired goal... one which some researcher may have already thought of, and is working towards!
kaichu dento
Flashaholic
I guess what I was imagining when I said that was higher voltage for cooler tint at the same time as the PWM instructed lower output level, while lower voltage giving warmer tint was actuated while PWM dictates high output. It may not work that way, but I'm sure that eventually someone smarter than me will work it out - maybe you? :thumbsup:KD - It's extremely cool on MANY levels that the computing power of, say, the diminutive LF2XT is probably greater than the UNIVAC! In the grand scheme of things, that happened very quickly, and has produced wondrous things, which makes it easy sometimes to hope that computational speed could over-ride physical law. But alas, as always, there are the natural properties of materials. "Nature, to be commanded, must be obeyed" as the famous quote goes (damned if I remember who said it!). So goes the presently designed LED.
In order to do what you're looking for (which actually sounds like a pretty cool idea BTW), we have to turn to different emitter materials and designs. Combining PWM with voltage regulation is exactly what a PWM driver already does. It allows a timed pulse of a pre-determined current to happen.
That certainly doesn't provide an easy answer, but hopefully it'll give the peace of mind of not wanting to force something against it's nature, and help with a perspective of looking toward different technologies to achieve your desired goal... one which some researcher may have already thought of, and is working towards!
