Analog pulse stretching

Thread Starter


Joined Jul 21, 2022
I have trying to stretch a 15ns pulse while keeping the same amplitude. I have managed to do the stretching via a monostable multvibrator but my problem is with the voltage. My output voltage would follow the rail to rail voltage of the op-amps. I would appreciate any help I can get on this. below are the requirements of the input and output.


1. Input requirements:

Minimum input pulse peak amplitude: 1mV.

Maximum input pulse peak amplitude: 3.5V.

Rise time: 12ns.

Fall time: 12ns.

Pulse width: 15ns.

Pulses frequency: 20Hz > Frequency > 8 Hz.

1.2 Output requirements:

The pulse output peak amplitude = Pulse input peak amplitude.

The pulse output pulse width: 1us < PW < 0.09ms.

The output stretched pulse’s maximum allowed overshoot is 100nV.

Propagation delay < 10us.


Joined Nov 6, 2012
You left-out the reason why You need a "stretched-Pulse".
This problem "sounds like" you're trying to fix a problem, or incompatibility,
with what ever Circuit is interpreting the pulses.

I'd really like to see if someone can actually come up with a solution to this.
It looks like you're trying to fix the wrong component.
Not an unusual type of request in these Forums ................


Joined Jun 5, 2013
Not a bad description of what you want, but it is my some things.

What happens if a second pulse comes before the stretched one finishes?

What is the level when there is no pulse?

What if a shorter or longer pulse occurs?

This is why we ask you to describe the problem instead of your proposed solution.


Joined Mar 14, 2008
That could conceivable be done with a very fast op amp configured as a peak detector (with the peak value dumped at the end of 1µs), but it's problematic at those short pulse widths going down to 1mV and with those accuracy requirements.


Joined Jan 23, 2018
The function can be provided with a sample and hold circuit triggered by a pulse detector, and reset after some preset time interval. OR use a very fast A/D converter to catch the amplitude after the trigger, followed by a digital latch circuit, feeding a D/A converter, with a preset hold time o the latch. More complex but possibly more easily implemented.

For the pulses described, a fast sample and hold makes the most sense.