Could be may be due to heat dissipation of the Nd:Yag.

this type of problems comes commonly due to heat dissipation problem in the laser crystal or any partial damage in the crystal.  because these two problems can be the responsible for time dependent power loss if the cavity is perfect.

you may check also:

stability of pump power with time

checking the laser out put power as a function of time for different flash lamp voltage.

Is it the power of fundamental or one of the harmonics? Anyway, it's normal to lose some power during warm up (20 min is a typical warm up time), don't care how much is your power in the 1st minute. If this system was not turned on for a long time, you may wish to clean the (gold?) reflectors which collect the light of the flashlamp into the YAG rod, because calcit can deposit on the surface and reduce the reflection efficiency. 

Hi Martin,

Nevertheless it all looks like thermal effects. So, what do you mean writing "...We do not think that heating of the cavity is the problem..." You simply "...do not think..." or you have inspected this issue and you are quite SURE there is not a problem?. So,

1. Thermal misalignment of the cavity? Certainly, alignment of cold and hot cavity may be different.  But you wrote that "... then readjusted after the power has fallen to try and correct the problem.  No improvement is seen...." So it's not the reason

2. As I remember Minilite has a number of extra-cavity optics (folded mirrors, etc.) They can be misaligned by the frame thermal deformation too. You can check power generated by "bare" laser. But, again, I think it's not the reason too.

3. More likely it is thermal lensing in active medium. I have no detailed information on the pump illumination system, but Continuum was and is good company and I hope that they created rather homogenous pump flux in active medium. In this case thermal lens (TL) is about symmetric. For Nd:YAG TL strength  ~~  dn/dT (75%) + Stresses (20%) + dL/dT(5%) and has positive sign due to the main fraction (dn/dT) is positive for YAG. Then, positive lens in the cavity decrease the mode size. Smaller mode size extracts smaller pump power into laser emission. If the laser is multimode, it is valid for all transverse modes. So, you cannot increase output power of symmetrically-lensed laser by angular adjustments. You should operate by cavity length (more likely it is disabled).

How it can be examined? Certainly, changing pump power etc. gives no results due to different heat dissipation.  You need  1) powermeter, 2) beam analyzer, 3) Fresnel attenuator (a system of glass wedges), 4) any lens or mirror with  500 - 1000 mm EFL. Items 2+3+4 create monitoring channel. The man problem is to couple a laser power in this channel with full signal illumination of your camera (to protect from saturation  and keep camera within linearity range). To say in a brief, using this scheme you measure a beam divergence ~ mode size. Then switch the laser on and control power and spot size simultaneously. If the power drop appears simultaneously with the beam size change, you can surely say that it's thermal lensing (Then, moving BA along the beam and finding cold and hot focus you can estimate an optical power of thermal lens).

What to do?  As I mentioned, Continuum produces good lasers. But are you sure this laser worked good while it was alive 10...15...years ago?

- You wrote that "...cavity optics have been optimised..." You can try to return back. Maybe your optimisation was not too perfect? 

I have forgotten more fundamental reason as thermal dependence of gain. It is mostly caused with two processes: 1) Thermal decrease in emission cross-section (low overheat of the crystal) and 2) thermal overpopulation of the bottom level of 4-level medium (high overheat of the crystal). Please find someting on these items in literature. Our old experiments demonstrated about 25% decrease in pulse energy at 100K overheat. Generation failure was observed at 200 - 250K overheat. (Valid for Nd:YAG pulsed flash lamp laser, temperature have been measured in the mode area of the crystal).

So we returned back to heat phenomena again. And again we have no answer for the question how this laser operated that 10... years ago, because gain factor depends on temperature, not on years :) 

Thank you for the very helpful replies.  When I am back in the lab I will dis-assemble the cavity and give the reflectors and Nd:YAG crystal a thorough inspection once more and I will also start running through thermal/cooling checkpoints aswell.

Thanks again,

Martin

One quick thing you can try to change the chiller temperature while the laser power is stable after 20min (power does not drop anymore). Sometimes the chiller flow rate becomes low due to some unforeseen reasons and the heat from laser crystal can not be removing efficiently.

Did you clean the system, use some Nalco Fluid to remove dirt and may be some algae, sometime also on the YAG crystal. If you working with SHG crystal, there could also be some heat dissipation in that crystal and you have to adapt the phasematching.

When a light beam is incident upon a material some of the energy is absorbed in the form of heat. The interaction depends significantly on both the relative sizes of the beam and the component under test, on the ambient conditions, on the mounting conditions and on the optical, thermal and mechanical properties of the irradiated component (optics).Dear Martin Ward I suggest you to replace the old Nd:YAG laser rod with a new one (one of the more costly elements in the laser).