Spectral width of an optical pulse

[narra]

If we take a CW laser,say, and operate a pulsed regime, will this directly effect the spectral content of the emitted light?

Is there anyway that we can pulse this laser without introducing linear chirp?

Thank you

 

[sambristol]

If we take a CW laser,say, and operate a pulsed regime, will this directly effect the spectral content of the emitted light?

Is there anyway that we can pulse this laser without introducing linear chirp?

Thank you

It's a bit of a pain to retrofit pulsing to a cw laser better to design it in. One option has been used for a long time - Q switching. You put a fast optical shutter (Kerr cells are often used) in the beam path between the mirrors. You pump the laser medium to get a population inverse and then 'open' and 'close' the shutter to generate a pulse of the required duration. Pitfalls to avoid - the power ofthe pulse will be many times the cw power of the laser so cooling may be required to stop the laser medium cooking. Also keep the pulse short enough that you still have a population inverse at the end of the pulse (just) or you will mess up the wave the wave train and hence the spectrum.

Speaking of spectral width if you Q switch the power / frequency curve will be a gaussian (bell curve). It will get get wider the SHORTER the pulse width (direct consequence of Fourier transforms) but in no way will it chirp. The rule is shorter the pulse the wider the spread of the spectrum.

Another way to pulse a laser is mode locking but this is more complex to do and the spectral signature is more complex.

Hope this helps

 

[narra]

Thanks Sam, this was a very interesting reply and I am tempted to ask further questions relating to it. But at the moment, at least, can I ask what if Q-switching wasn't a reasonable solution. Let's say we had I diode laser which we wanted to pulse. Pulsing the drive current seems like an easy first approach because all we need is a waveform generator (talking simplistically). We could then control our pulse length with ease by varying the drive current profile.

If this was the case, I am a little bit unsure of what's happening to the pulse: are we changing the pulse width and thus directly changing the spectral width (as in a Fourier relationship)? Or are we just making the pulse less transform limited?

Also, is a pulse that is emitted without chirp effectively the exact same thing as a transform-limited pulse, assuming this pulse has not yet undergone any broadening?

Thank you