Help understanding the dynamics of LED Vf and drivers

Hello all, this is my first post, so please bear with me! I’ve done quite a bit of reading, but can’t seem to find the answer to my particular topic. I hope this post may eventually help out others with the same questions.

A few months ago, I was bitten by the bug and I’ve become very interested in flashlights and the technology they encompass, especially after purchasing an XP-L HI Emisar D4V2. Last week, I assembled a simple Convoy S6 build with a Samsung LH351D and QLITE driver, and I’ve been quite happy with it. And of course now I have a desire for more complicated builds, but I’m a bit out of my element when trying to understand electronics.

While I have a fairly decent understanding of electronics, I would still describe it as rudimentary, especially when trying to understand the influence of emitter Vf. I understand that LED’s are diodes and therefore have a non-linear voltage/current relationship, and I’m under the impression that the applied current controls Vf independent of battery voltage. But I’m not sure that I understand how this factor influences drivers (or vice versa) and overall efficiency.

For example with my chosen LED, Samsung’s data sheet shows a Vf of ~3.3v at ~3amps, the approximate peak current of my driver. Assuming a 4.2v vBatt and a 0.1v drop through the 7135’s, this would leave 0.8v * 3A = 2.4 watts of energy lost to Ohmic heating when ignoring intrinsic resistances? I suppose my primary question is if the difference between battery voltage and Vf functions in the way I just described with a current regulated driver?

Even more confusing to me, is how this relationship changes in a multi-LED light such as my D4V2. If I run the light at the maximum of the single 7135, then ~350mA is divided equally across all 4 emitters (and therefore each emitter would have a Vf corresponding to the 88mA current)? But if I then run the light in turbo through the FET channel, all 4 emitters would be subjected to 4.2v from the battery?

You right with linear driver you have voltage drop between battery and LED. In that mode driver work as classic linear power supply but not stabilized by voltage but by current. So all excess energy from that convertion is dissipate in form of heat. In your example the calculation must be 4.2-3.3=0.9V. So power dissipation will be P=0.9*3=2.7W. Max Voltage drop in 7135 will only dictate how long the stabilizer will remain in regulation. In parrallel configuration the current across any LED will be different because LEDs do not have exactly the same V/I characteristics and they are not ideal. So in higher voltage LED with high battery voltage the linear driver will be more efficient compared to low Vf LEDs because there’s lower voltage drop between battery and load. Also some week ago I studied some leds datasheet and extracted graphic V I data in Excel. From that I saw that most of modern high power LEDs have very linear characteristics especially in high currents.

You forget also the voltage drop inside the battery itself, low drain cells sag quite a lot at 3A

+1
At more than very low currents you always have to figure in your particular battery’s voltage sag, especially at higher currents.
Looking at HKJ’s battery graphs with tested currents really help in figuring out what all is going on at different currents with battery sag figured in.
https://lygte-info.dk/review/batteries2012/Common18650comparator.php
The amc7135 linear driver is pretty efficient with the led vf close to the battery volatage at a given current.
At 3 amps and a led with a 3.6 vf at 3 amps the battery sag should be around 3.8 or 3.9v at the start from fully charged. Does depend on the battery used.
Once the light has run for a few minutes battery sag falls to around the 3.6 or 3.7v right about the perfect voltage needed for the led at 3 amps.
After that the driver is direct drive, the led is taking what ever the battery will give.
Choosing the right battery does help, a 30Q doesn’t have alot of voltage sag so you would have more energy wasted than using a Sanyo GA plus it has more capacity.
The 30Q would stay in regulation longer but would waste more energy because of its higher output voltage with a load.
When you want more current with less sag its better to go with something like the 30Q.

Haha internal resistance of battery is also important yeah. I forgot that Lexel :-).
Also about 7135 and PWM. With 7135 we don’t have trully constant current. In reality we have PWM average limited 350mA current across LED.