I want to build up a single frequency q-swiched pulse laser.
Before trigger the laser cavity will be locked to an external CW-laser scource. I want to use PDH locking as this way very narrow bandwiths can be reached as it does not require modulation of cavity length.
After the pulse is extracted the laser will require a full restart of the regulation of cavity length. This locking will need to be stable at the time the next pulse is extracted. A repitition rate of 1 kHz limits the time for relocking to less than 1 ms.
Looking at the Error signal of a PDH locking I find a wide catching range compared to cavity length modulation (due to the much higher modulation frequency), a fact supporting fast relocking. But outside the sidebands the error signal leads to a repellant error signal.
How can I assure instant relock? Is phase modulation at 50% of the free spectral range a solution?
Hi Vincent,
Good question,
I am using a pockels cell in a q-swiched resonator. Before pulsing the resonator is adjusted to an outcoupling rate to keep it just below lasing threrehold, in ths situation injection locking is easy. During pulse build up the cavity is closed and the cavity internal power increases to the MW range, in this situation no locking is possible.
It is possible to hold the lock during the pulsing process, and it could be the case that the cavity after the pulse still matches the seed mode. In fact the error signal after demodulation chould turn zero during pulsing, in this case an integrating regulator should stop by itself during pulsing.
However I am not sure of this, so I would prefer to be on the save side. There are some effects changing the cavity length.
There are some effects changing the effective cavity length: the change of crystal inversion leads to a changed refractive index (Kramer Kronig equation), pulse light absorbtion causes sudden heat up of laser component and so on.
Hi Karsten,
This problem was solved by a sample-and-hold circuit in our old lamp-pumped Q-switching laser. That is to turn-off the feedback loop and hold the cavity control voltage while laser is outputting. This can be done be a single 8-pin chip circuit. I think there is no way to move the cavity during the very short (~100ns or shorter) duration of Q-switching. For a system without fast disturbance ( 10 should be fine (high frequency (10kHz?) acoustic vibration must be minimized and temperature fluctuation is at a much slow time constant). For a current pumped (semiconductor) laser, it could be a problem, because the current noise (changing the gain medium reflection index and cavity length) is usually at the high frequency end and a very fast feed-back is needed. I have no experience with that.
Hi Yi-Wei,
this is very promising. If no relocking is needed stabile pulsing should be no problem.I guess my main laser is save.
Do you have any experiences how this will work out on an OPO? Are there some funny effects changing the refractive index will I have to consider?
Hi Karsten
I have no experience in pulsed OPO (relies on nonlinear effect, on population inversion involved, the refractive index effect should be small?!). But, if the change of refractive index caused by pumping laser turns out to be a problem, I guess, it can also be compensated (by EOM, or something like that). If the pumping pulse is stable enough, the changing of refractive index is "predictable" (more and less) and a pre-engineered ramping curve of cavity length can be added after the cavity being "frozen".
- Badges
- Science topic