Ok, possibly a pretty simple question here but I can't seem to get my head quite round it. I know about setting the current for an LED with a series resistor etc. but...
what if you have an LED requiring roughly 3V forward voltage, and you only have a 3V (2xAAA batteries for example) supply to power it? There is no excess voltage for you to use for calculating the required resistor value. Is it a case of looking (through testing) at the absolute minimum voltage required to switch the LED on, and working with that. So if through trial and error you found the LED worked at 2.9V, you'd use R = 0.1 / 0.02 (for 20mA), = 5 Ohms?
I've been playing around in LTSpice and i'm not getting the results I expected to get.
Note that if these are alkaline batteries end-of-life is sometimes defined as 0.8 volts. So, you might want to design your circuit to work with a total voltage of 1.6 volts.
How do you run a 3 volt LED with 1.6 volts? There are many "white LED" driver ICs available or you can roll your own by building a charge pump voltage tripler. A voltage double is suitable if your batteries are always fresh.
Generally the specs for LED forward voltage are close enough that you should follow them. If your system voltage and the LED Vf are the same, then you need to explore other options. You might get away with a very small series resistor, but you have to be able to limit the current so that you don't destroy the LED. As for using Spice, I would just use a breadboard instead. The trouble of getting Spice to perform with results you want are probably more than simply trying it out. After all, it's only a few parts.
You have a couple of options, however: 1. *Switch to a different source voltage*. You may already be limited here, but if not, try using a Lithium cell(s) instead with a higher voltage. 2. *Change the LED to one with a lower Vf*. Again, you may be limited but you could easily find another color (such as green, red, etc.) with a Vf of 2.0V. 3. *Boost the system voltage*. Using a boost converter, you could bring 3.0V to 3.3V or 5V, for example. This may not be an option for the rest of your circuitry, but if you're using a micro to drive it, you could just as easily run it at a higher voltage, most likely. 4. *Have a supply just for the LED*. You could have a small (50mA, etc.) supply for your LED, boosted from the battery voltage, and then just use a NPN transistor to switch the current through the LED on/off with the lower logic level. The supply could be generated with a voltage doubler, linear regulator, op amp, voltage reference, etc. 5. *Use an LED driver IC*. Several manufacturers have chips dedicated to driving strings of LEDs and have the ability to dim them with a data bus. You could use one of these, but it would require you to control the chip from a micro. It's more complicated but you have some more flexibility. This would be an example of such a device: http://www.ti.com/product/tlc5917
Thanks for the insight! Yeah this does help. Without any spare voltage it's difficult to limit the current as there's nothing for the resistor to work with. The most simple solution at the moment is to just use 3 x AAA's instead of 2. There are space constraints but this shouldn't cause much of a problem.
Either that or use a boost converter with 1 battery to up the voltage from 1.5V to around 5V. Or even make a simple joule thief circuit, but I'm not sure how consistent the current is with that kind of circuit. Thanks for your help!
First of all, LEDs are current devices. The forward voltage drop is not consistent. It varies between devices and it can vary with time and temperature. The only constant is the current. The forward voltage comes into play when determining how much supply voltage is needed in the circuit to even achieve the required current--i.e. the supply voltage must be equal to or higher than the highest possible forward voltage (for desired LED current level).
If this is a "one off" design, then you can cherry pick your LED (i.e. test LEDs until you find one with a forward voltage low enough to function in your design). But for a system that will go into production, you need to design for extremes.
So, when designing an LED driver circuit, first determine the driving current, so consider these points: * Must be at or below the LED MAX forward current spec--and I would run it at about 67% of that spec if longevity is at all desired -- e.g. if 30ma max, run it at 20ma. * Consider if continuous operation or intermittent. And if intermittent, check the Max pulsed current for max pulse duration. * Consider the brightness required. I have often run a super bright LED at well below their max current to achieve low power (i.e. the LED glows at "indicator brightness" rather than at "illumination brightness", but at much less current than a standard brightness LED). This, will, also, reduce the forward voltage, sightly.
The LED driver circuit can be as simple as a resistor or as elaborate as a constant current regulator. When using a resistor, consider that what you want to acheive is as constant a current as possible. The higher the resistance the more stable the current over the range of possible forward voltages. But, to achieve that, you need a voltage high enough to develop the specified current though the resistor (again, the higher the better [in terms of current stability]). After all, the dynamic resistance of an ideal constant current source/sink is infinite!
Another trick is to select a battery that has enough internal resistance to limit the current in the LED. What you want is a balancing act between the current delivered and the LEDs forward voltage. Done right, this becomes a self regulating system. Because of the exponential rise in current vs forward drop (an LED behaves just like a semiconductor diode, only with a larger forward voltage), the current drawn by the LED will quickly swamp the batteries internal resistance causing the voltage delivered by the battery to drop, thus causing current regulation.
Another thing you can try is a "Joule Thief" circuit. With it you can drive a standard LED with as little as 0.5V, thus the name "Joule Thief"--it steels most of the joules from a battery (.i.e will run it down to about 0.5V). Here are some links that describe this cute little "gem":
The Linear Technology LTC3216 is designed to drive LEDs from a low voltage supply. This part is a charge pump that automatically switches between 1X, 1.5X or 2X step up and regulates the LED current.