Turn Short Pulse Into Longer Pulse

Hi, I have an application where I expect a certain amount of environmental noise on my input, I have my hardware set up so that an indicator LED lights in the presence of this noise however the duration of some noise is too short to visibly light the LED. I would like to use hardware to generate a longer visible pulse, I know that I can easily do this in software, but I like the current arrangement where the hardware is reporting its own state - its a sanity check on my software.

Can anyone point me in the direction general direction of a hardware approach - I am thinking something along the lines of a transistor to switch a higher current that would charge a series capacitor. I know I can use a 555 Timer but would be interested to learn of and try out an alternative.


Duane B


If the pulse is strong enough then let it charge a resistor-capacitor(-ground) to light the led or hold the pin HIGH. If it isn't then use a transistor to make it strong enough.

Look for a 555 "monostable" circuit, should do what you want?

do it in software, add 2 potmeters to make the delay for the rising and falling flank both adjustable, - if both are 0 there is only a propagation delay - if A = 6 ms and B = 5 ms the pulse will be shorter - if A = 5 ms and B = 6 ms the pulse will be longer

What is the length of your pulse?

Hi, Thanks for the replies.

Its for an IR Decoder, I am generating a 40Khz IR signal from a 555 Timer which activates the decoder, this 40Khz Signal is being switched at 100Hz by another 555 Timer which is the signal I am interested in at the IR Decoder end. In daylight there is sufficient light to cause momentary pulses on the IR Decoder, its very simple to filter these out in software, however I have an LED attached to the decode which lights whenever it reads a signal, its very simple an LED connected between Vcc and the decoder output, when the decoder output is pulled low, the LED Lights. This works very well for testing my hardware with or without software running, but the pulses caused by environmental light are generally too short to visibly light the LED. What I would really like is a low component count solution to extend the pulse length in hardware so that the pulse is visible on the LED. I was hoping for something with fewer components that a 555 monostable, but given the difference in the tigger pulse and visible pulse durations it might be my only options unless anyone can expand on the suggestion from goforsmoke -

"If the pulse is strong enough then let it charge a resistor-capacitor(-ground) to light the led or hold the pin HIGH. If it isn't then use a transistor to make it strong enough."

Duane B


Thanks Duane B.


You could wire a led with one leg (- leg) to ground and the other to a resistor. At the other end of the resistor, a capacitor and the other side of the cap to ground. Then connect a diode between the resistor and cap. A short current that flows in through the diode will charge the cap even as some flows through the resistor to light the led and when the current flow through the diode ends the cap will still empty through the resistor. You can try and forgo the diode and connect between the led and resistor as well, I have done similar to that but there was other circuit that won't suit your ends. Be sure about this, the values for the cap and resistor can vary the delay greatly.

Some time check out how a cheap digitizer circuit works. They time the bleed of a loaded R-C circuit. It's the same basic way that capacitive sensing is digitized though that circuit is a bit different, and that leds as light sensors are digitized (except my way is different) in either the learning or playground example, I forget which.

A 74HC123. It's RC-based, so it's the IC and an R and a C, and a couple of wires to set it for rising or trailing edge trigger. It's a "pulse stretcher".

Hi, Thanks for the replies, I have found a few sample circuits by searching for the term 'pulse stretcher'.

I was planning to try them out this evening, but I picked up some 556 Timers earlier today and have been migrating my transmitter from cascaded 555s to a single 556 - its so much easier to work with on strip board :-)

I will get back to the pulse stretcher tomorrow.

In the meantime, thanks for all your help

Duane B