From what I gather, they are a low resistance method of direct driving a light. That’s all well and good, but how do we keep FET drivers from frying our LEDs? Some batteries claim 10+ amps, but I don’t know of an LED that can survive 6 or 7 amps.
What am I missing? I’ve tried searching the forum and the internet, but I can’t find a “FET basics” post.
For what it’s worth, I’d like to play with something in a UF-1504 and a yet-to-be-purchased light with a large reflector(considering a yelz Y3, Supfire L3, and the Convoy L2). I generally try to keep it to single cell lights since I have no reason to take on the extra risk of multi-cell lights.
Resistance in the switch, in the springs, between tailcap and body tube, between body tube and head, in the wires between driver and LED all contribute to a lower maximum current to the LED. Additionally, there is the factor of voltage sag by the battery. When under a high load, the voltage of the cell drops to the point where current in the LED is limited. Back when red 2600mAh Sanyo cells were popular, I’d measure over 30A the instant I shorted the battery with my DMM set to the 10A setting yet, if I put it in a FET driven light, current would be under 5A. AFter saying all this, I would NOT be willing to connect a RC LiPO to a FET driven LED. Hahahha
Eebowler has a good point. I think it also has a lot to do with the forward voltage rating (the Vf) of the led as well. Voltage drops across the led’s resistance along with amperage.
We are definitely over driving the led but that’s why people are so into heat sinking. Copper pills, stars, heat fins, letting the heat go into your hand etc. I.e. Don’t tail stand a light with d.d.
A FET driver like the BLF17DD is not different than directly connecting your led to the battery.
Every led has a voltage/current relation, at a certain voltage a certain current is flowing through the led, for every led there's current/voltage charts.
Something the other way around goes on with a battery, with more current, the voltage drops, there's current voltage charts for specific batteries as well. A '25A' battery only delivers 25A with a certain, low, voltage.
If you connect the led to the battery the voltage of the battery drops while the current increases, at the same time the led voltage increases with increasing current. At some current those voltages meet and that will be the current the led gets.
It is a bit of a councidence that the voltages that modern leds run at are close to what a single li-ion battery delivers, that coincidence makes direct drive a possibility.
Are you guys telling me that I can’t fry an LED via direct-drive, as long as heat is managed? Or is it possible to have a battery(single cell) that is capable of delivering too much power?
Not the Cree XM-L2 at least. I drove it directly from a battery with no issue to check the tint. But be careful of heat if it is not installed in a flashlight.
It depends on which led, which battery and how much cooling the led has. With the led mounted on a modern copper board with a direct thermal path from the center pad of the led to the core of the ledboard (like Noctigon and Sinkpad boards), the most used leds (XP-G2, XM-L2 amd XP-L) can survive direct drive from a high current capable 18650 battery. But you will blow a XP-E led because it can not stand too high current, and you will blow a Nichia 219C led because the Vf is so low that it will draw an enormous current.
There’s a long history of magic smoke associated with various forms of direct drive and efforts to max out LEDs. Sometimes it’s bad for the led and sometimes it’s bad for the cell. Multiple parallel cells will sag less and deliver more current to an led. Multiple parallel LEDs will have a lower combined Forward voltage (Vf) and draw more current from a single cell. Heat kills so we use the now common direct thermal path copper mcpcb’s to draw the heat from the led junction quickly away from the source. Having done that the XML/XM-L2 seems to handle current from a single cell quite well allowing one to tune the output with cell choice and whether or not to impliment various resistance mods. XP sized LEDs run a bit closer to the edge of failure since they start with a lower max rated current and depending on the cell that rating might get exceeded by too much. Bottom line the LEDs can still be killed but DTP copper boards lets us push them much harder.
And there is no where for heat to go, do not throw a FET into one of those. I recently got my first and last SK68. I installed a driver that PilotDog68 developed for that light. I think I would like a quality zoomie, but these are not it. I will likely pull the driver and the led and o-rings and part ways with it in the circular file marked never again.
