Wall transformer stays charged

Thread Starter


Joined Mar 3, 2011
I have a little fish tank with LED lighting. It has a red green and blue but it's set to only have one blue LED on. I hooked it up to a remote switch to turn on and off at the wall supply, but the LED uses such little power that not only does it slowly fade off but stay very dim for quite a while. And fir some reason doesn't fully discharge until I switch to the other colors. Or I assume wait a few minutes but I havn't actually even been patient enough to sit in the dark and see how long it takes. Anyway the wall transformer is 12v, 150ma. I was wondering if there is an easy way to make it fully discharge when switched off. There is a barrel connector at one end so it would be easy to wire to it, I just don't know what sort of a circuit would do the trick. I would imagine I might need an IC that detects when the voltage drops to switch a bridge between the anode and cathode of the power supply. If that's the case I may not want to bother. But if there's something more simple I could give it a shot.



Joined Jul 13, 2008
The wall transformer has capacitors, as filters for ripple or other apps, depending on how fancy it is.

When you switch the unit off, your led is acting like bleeder resistors until the capacitors level, or lose their operating charge.

This is a non-negotiable feature of the "wall wart" so just rest with the knowledge that it does indeed go out after awhile.


Joined Jul 17, 2007
If you'd like for the LED(s) to go out faster, you could connect a 10k resistor across the output of the supply. The resistor will drain only ~1.2mA when the supply is on, but it should help turn the LED(s) off more quickly than they do now.

It is surprising how long that an LED will remain dimly lit without such a "bleeder" scheme.


Joined Jul 3, 2008
It is surprising how long that an LED will remain dimly lit without such a "bleeder" scheme.
Yes, and it's also amazing how sensitive the eyes are (particularly to green LEDs) in the dark!

Even a very low voltage can put microAmps into the LED and produce enough light to see. Years ago I used a very sensitive optical spectrum analyzer to measure the optical power of some unexpected green light that was visibly showing up from an optical experiment. It turned out that the level was barely above the noise floor of the instrument, at about -100 dBm, which is less than 1 picoWatt! It looked fairly bright to me when the lights were out.