LED direct drive current calculator

Flashlight Modding and DIY Parts
1

Or more like an estimation : Graphing Calculator - GeoGebra

I used this online graphing calculator to graph the LED current-voltage curve and the voltage drop in the circuit and cell/battery, the intersection is the resulting current. To obtain each LED functions I plotted the current-Vf data from community tests (maukka, Texas_Ace, koef3) in excel and obtained a polynomial function which matches reasonably well the ploted curve (3 degrees seems sufficient).

How to use :
Select the LED
Number of LEDs
Cell Voltage
Cell internal resistance
Flashlight resistance

The last one is not something that is usually known unless measured, my PL47G2 for example is around 24mΩ, (incl 4mΩ for each spring), but obviously that will vary between drivers, wires used, spring resistance, bypass…

Some example of cell’s DCIR :
GA/35E/MJ1~35mΩ, VTC6/30Q~20mΩ, P42A/40T~17mΩ, 30T~12mΩ

The intersection of the two curves is the current (clicking on it gives the value), it’s only valid at the current (per LED) tested. (e.g. for E21A it’s only valid between 0.2 and 2A per led)

Stuff that makes this quite approximate :
LED Vf varies with samples, heat.
Cell DCIR varies with samples, heat, state of charge, and it disharges very fast at high current.
Flashlight resistance is generally unknown, varies with heat.

Testing with my PL47G2 with a partially charged P42A, calculated : 12.64A, measured 11.3A (current clamp, tailcap wire bypass). Close enough :partying_face:

Probably not very useful in the end but I liked the visualisation.

Tests used : sst-20 , lh351d , xp-l hi , 219c , 219b , e21a sst-40

2

led ’s vf changes as it heats up. and so does the current it pulls. it may be ignored with cells that do not provide much current to begin with, such as alkaline cells, but for anything serious, the fluctuation of current may be pretty dramatic. same with spring\wire resistance as you mentioned,

too many variables to consider it is just not worth it, considering linear drivers are dirt cheap and very reliable, having said that i did build few DD lights, that i still use, a decade later, but they are built with high Vf p7 leds, that are unfortunately aren’t available anymore. they were perfect for DD builds.

3

Cool, thanks for sharing.

Using this principle of matching the cell voltage minus I*R with the LED Vf, along with becoming familiar with typical resistances present in a flashlight, is very useful for flashlight modding.

Not only for designing high power lights with odd combinations of LEDs and batteries, but also with optimizing regulated lights. For example using a linear driver with a high Vf LED will not get you much regulation. Using these principles you can estimate the regulation durations at different currents and different LEDs and cell combinations.