Intracavity beam skewing due to thermal lensing?

[brianpe]

I've been working on a large DPSS green for a couple of years and am trying to implement a design that uses passive Q-switching to increase the intracavity intensity for green conversion. The design I'm trying is a z-fold configuration where two concave mirrors form a focal point to provide the right intensity to the saturable absorber, a fairly large pump mode volume, and about a 2x reduction in area going into the KTP. I've been using the free software Rezonator to work out the cavity and the software predicts its stable down to about 30cm of thermal lensing in the vanadate crystal. Here is a diagram of the resonator design:



Attachment 60238
I have this lasing on my workbench in CW mode -- all the mirrors are in place but there is no Cr:YAG or KTP crystal in the cavity. The problem I'm having is there is a very narrow range of pump powers where it lases. If I turn up the pump power past this range lasing stops. I think this is thermal: If I start with the pump up I'll get lasing and then the beam will fade, flicker and finally go out within a few seconds. I don't have a TEC on the vanadate but I do have it in a copper block and the block doesn't get warm so I don't think adding a TEC will help. A thermal camera on the vanadate shows it's only about 80ºF. I think something about this cavity is magnifying the thermal lensing of the vanadate.

I've done some simple thermal lensing calculations and they show the vanadate should be nowhere close to the 30cm of lensing needed to lose stability. This is a pretty long cavity design (about 650mm) and there is some astigmatism caused by the beam hitting the concave mirrors at an angle (max about 10º). I don't know how much of a difference that makes (Rezonator shows the astigmatism actually decreases with more lensing).

I had a theory that maybe the lensing is causing the beam angle to change in the cavity and this is causing it to walk off the pump mode overlap. As a test I turned the pump up until the laser stopped, and then adjusted the X-axis of the pump and viola! I got the beam back. There is increased deflection as I increase the pump, and if I back the pump power off the deflection decreases so there is really only a narrow sweet spot of pump power where it works, but I can tune that spot for any power.

As lensing increases the focal point between the concave mirrors shifts to the left. I assume my alignment is not perfect enough in the vanadate and the beam is not completely parallel. But when I aligned this I used several 1mm pinholes and the beam seems to be going straight through the vanadate. I don't know how I could make this more precise.

Can anyone think of any other issues here that may be causing the beam angle to walk in the vanadate as the power is increased? Any idea how to eliminate this?

Thx,
Brian

[Farbe]

Take your Thermal cam and check the folding/focusing mirrors. If you see a hot one -> there is your problem.

[brianpe]

Take your Thermal cam and check the folding/focusing mirrors. If you see a hot one -> there is your problem.

Great idea. I tried this, but all the mirrors show as cold. I was also wrong in my assessment that this can be fixed by walking the pump beam. Yes, it can, but after re-aligning again I still have the problem but walking the beam no longer works. So it is still a lensing issue.

One thing I tried is to shorten the lower arm where the KTP goes. I shortened from 98mm to 73mm. This increases tolerance from lensing and allows lensing down to 25cm (from 32). When I do this I can increase the power more. It still fails far before where I predict but I think this shows some predictions hold.

So three things:

First I designed this cavity to have a largish gain mode area to reduce pump lensing. It's 430µm, but elliptical due to the concave mirrors -- the larger radius is 550µm. I've read that while increasing the area does reduce thermal lensing, it makes the cavity more susceptible to lensing. Is this just too big? The mode diameters don't change too drastically as the lensing increases.

Second is that maybe I'm estimating lensing wrong? Vanadate has different dn/dK values for different polarizations. My pump is a multimode fiber that causes random polarization. I'm estimating this by calculating lensing for 50% e polarization and 50% o polarization and taking the average. I've also tried adding these together and this is still double what I'm seeing on the bench.

Finally, does cavity length make a big difference here? I've built several cavities and never had such an extreme lensing problem, but they have all been shorter and have a smaller gain mode.

I'm only driving this with about 5 watts of pump before it stops....I find it surprising that this would yield 32cm of lensing so early on.

[Farbe]

If I am not mistaken, at 5W heat load and 200u beam radius you should expect a thermal lens with focal lengths of around 85mm or 30mm depending on thermal optical coefficient used.
The low thermal conductivity is worsening that by quite a bit.
Its only 5W/m/k.

[Vidal Wolf]

I have had lensing issues with my 5 watt green YAG in the past. it was an external aluminum mirror that had a protective coating on it that would cause the beam to bloom or diverge at 4 to 5 watts right off the head. I swapped to an uncoated mirror and the problem went away.

[kecked]

you might try skewing the crystal in the cavity. The focus of the gauss point in the beam might walk a tiny bit and that might fix it just by moving the point of focus in the crystal. It likely will drop output power a little too. I'm thinking this might select a slightly different phase and or mode. You could also try temp tuning and or rotating the crystal to see if there is a better phase match. There are usually four places where (n1+n2)/2 is met. KERR might be self focusing...is self focusing and could skew the acceptance angle where efficient phase matching occurs. You might put a rotator in the cavity and see if changing the phase of the pump beam might help a better phase match while reducing self focusing. You have to satisfy the collinearity of the photons or no SHG. Depending on how the birefringence is located in the crystal and the cut you might have to play a bit. I'd expect a bought crystal would be labeled as such.

This paper gets the points pretty well. https://authors.library.caltech.edu/2871/1/ANDao84a.pdf
Look at figure three. You can do the math but its all there in he text.

I missed the passive qSwitch element. Does the beam walk when the qswitch dumps? iS it stable in CW?