I assume you’re looking at using super/ultracapacitors to light an LED, to make some form of super-fast-charging light duty flashlght, something that every electrical engineer has at one point or other thought about upon first encountering supercapacitors.
So, we make these assumptions:
- 3.8V 40F ultracapacitor, such as a Taiyo Yuden LIC1235R3R8406 (awesome little ultracaps, the higher voltage than a Maxwell boostcap means higher energy density).
- A boost converter to allow the LED to run off a low voltage, which can handle supply as low as 0.7V (converters designed for operation on single AA cells can go down this low, such as the Maxim MAX757) at 87% efficiency average, and driving the LED in constant current configuration (maybe self-regulating, no resistor).
- Assume a reasonably bright white LED running at 20mA at 3.3V, which is the type used in cheap keychain LED lights.
The energy stored in a capacitor is 1/2 CV^2. So the total energy we’re going to harvest from a 40F cap dropping from 3.8V to 0.7V is 1/2 * 40 * (3.8^2 – 0.85^2) = 279 Joules.
Power requirement is 20mA at 3.3V = 66mW, at 87% efficiency, meaning we need 75.86mW to be supplied.
279 Joules at 75.86mW is 3677 seconds = about one hour.
One hour on a small cylindrical supercap that you can charge in seconds is pretty good going.
What happens if you just wanted an indicator light that you can see in the dark, like a red LED running at 2V at 5mA? Power draw is just 10mW, you’d be able to run your capacitor for 6 or 7 hours.
What you’ll be probably feeling right about now is a sense of disappointment, given that a battery of an equivalent dimension to the supercap can run the LED for significantly longer, which reinforces the overall lesson that super/ultracapactiors are awesome for power-duty, where high power is required, but still pretty poor at being an energy storage, which lighting LEDs is a typical application for.
article from: https://qr.ae/pC0Rfg
Post time: Mar-11-2026