Hi guys, this is my first post here.

I need a circuit that slowly fades in a LED driver when turned on, stays on, then, when turned off, slowly fades out.
I was thinking on a 555 oscilator, but I can't find any circuit working like this.

For LED dimming and driving, I've found PT4115, which is very resourceful and I think will help me accomplish my needs.

Can someone help me?

 

Define "slowly fades". Is that 0.1 seconds, 1 second, 10 seconds or a minute.

Also, from what you said, it sounds like the LED will be turned on/off with a switch by a user. Is that correct?

 

Define "slowly fades". Is that 0.1 seconds, 1 second, 10 seconds or a minute.

Also, from what you said, it sounds like the LED will be turned on/off with a switch by a user. Is that correct?

Hi GopherT.

By "slowly fades" I mean about an hour.
Actually, I plan to control this LED light with a simple mechanical timer, like this one:

Maybe this circuit may be split in two parts. One power part to handle LEDs and other to handle on/off times, and use a relay to integrate both.

Maybe (yes, another "maybe"), a 555 PWM could do the trick controlling PT4115 IC.

Edit: I'll use this LED circuit with an aquarium light.

 

It will be difficult to come up with a slow increase/decrease over an hour unless you use a microcontroller with PWM output and 10 bit resolution. Maybe someone else has a better idea.

 

Thanks for your help GopherT.

PT4115 can be controlled by PWM and DC from 0.5V to 2.5V.
Is there a circuit that takes an hour to go from 0.5V to 2.5V? It will help.
I've found this video on Youtube that shows how slowly fade LEDs, but using capacitors. I don't think capacitors could keep 2 10W LEDs for an hour.
But maybe this fade circuit could be adapted to slowly increase DIM pin on PT4115.

I've found this blog talking about PT4115.
Here's a sample circuit:

 

You will need two 'modules'. One to generate a PWM frequency with variable duty cycle. Another to ramp a control voltage up or down (ONE TIME) to ramp the PWM duty cycle. Each could be done with a 555 timer so you will need two.

You will also need a low-leakage capacitor that charges very slowly (100uF to 1000uF) and controlled by a 10M to 1M ohm resistor (respectively). I have t seen a 555 timer work well more than 10 minutes or so but I haven't played with long curation (slow) signals often.

I'm off to start my day so I'll let you track down the rest. Good luck.

 

If small steps are acceptable then a reasonably simple solution is to use a 555 for clock, up-down binary counter & R-2R digital to analog converter + some logic for control.

 

Another posibility would use one line V timer, PT4115 rated at 95 uA in dim state which is much less than another timer would draw. An AC relay would control charging and discharging of timing cap, C1. On breadboard C1 is a verry low leakage 10,000 uF , 10V cap. When pulled out of storage it still had 3 V on it, last used ?? C op. V shoul be higher than supply V, or clamped with a Z diode. A 5.6 V Z diode showed little leakage below 3.5 V. MPSA14 seems to have no effect on charging time. Input Z of dimming pin not stated but prob. high. R2 presently 2.2 M. After one hour, C is up to 1.86 V and climbing, but supply V is only 12V. 1 hr & 12 m C =2V, shutting down.