How can I generate a logic pulse from a broken IR beam?

I would like to trigger a [7]555 timer when a phototransistor stops receiving light from an IR led, more precisely when the (expected to be smallish) signal from the light-sensitive voltage divider, consisting of a phototransistor and a resistor, changes quickly. How should I wire the timer? I expect if I decouple the light sensor from the timer with a capacitor I might be able to get a sufficient voltage swing to trigger the timer from a small but quick change in the signal?

I would like to implement a version of the LED staircase. The project idea is to build a number of as-cheap-as-possible photodetector-triggered lamps on one side of a 3-foot gap opposite IR LED throwies on the other. When someone passes by they break the beam and trigger an entertaining trail of lights that slowly turn off, one by one, a couple of seconds later.

I was overthinking the problem. After some experimentation it appears all I need is a photodarlington (a phototransistor by itself is not sensitive enough) in series with a resistor. This forms a voltage divider with a usable signal. If a potentially long low pulse is a problem then merely decouple the photodarlington from the logic with a capacitor differentiator.

-

I would start it using a LM567 to detect the presence of someone on the stair. I recommend this circuit: http://www.mondotronics.com/PDFs/3-337_Mod_IR_v22.pdf

Using the output of this circuit, you should trigger your LM555 to keep the light on as much time as you want, but not using a large capacitor as you mentioned, but just a LM555 in monostable operation.

The output from LM555 can drive a transistor to control as many LEDs as you want!

Some level or signal inversion maybe required.

-
Everyone loves signal inversion. –  Kortuk Sep 29 '10 at 14:54

This may seem like a challenge, but I can give you a link to a relatively common circuit. You can then use the photo-transistor as a discharge from the capacitor. when the transistor has light, it will keep the capacitor discharged, you just need to use the potentiometer to tune the circuit to have a low enough resistance to charge the capacitor when the transistor has no signal.

There is another easy way to do this with two 555 timers, but I need to draw it. Let me know if you need it, another answer popped up. I am still posting this as another option. I was going to use a 555 to detect the change, and then have it trigger a 555 to stay on as long as you want, as andre says.

-

I'm sure there's some way to create an analog solution to this but nowadays I wouldn't even bother. If a small microcontroller can handle it, use that. I'm not sure if I'll ever use 555 timers anymore because a microcontroller can do it and so much more in the same package. True, the ATTiny45 costs about 4x the LM555, but if your application requires a microcontroller it might as well do one more job. Using the 555 has always been a lot more difficult for me than figuring out some C code to do the same thing.

-
For this application, the (7)555 timer has the important feature of costing 35 cents in quantities less than 10,000. –  joeforker Sep 29 '10 at 15:37
Learn the 555 timer. One of the best chips I have ever used. It is very low cost and it can be used for crazy crazy things. You should really take a look at some of the application circuits. –  Kortuk Sep 29 '10 at 15:54
They are also quite noise tolerant because of their reference being relative. –  Kortuk Sep 29 '10 at 15:54
Another problem, there's a learn curve related to MCU products that "maybe" more complex than learn how to use a LM555! –  Andre Oliveira Sep 29 '10 at 16:18
Curves depend entirely on where you come from. Neither is exactly complicated. Odds are, though, that the MCU will require fewer (if any) external components and can put itself to various sleep modes etc. It will almost certainly require a computer. AVR can be programmed with a few resistors and a real parallel port, though. –  XTL Sep 29 '10 at 16:36