Diode pumped Ti:Sa Laser

I always got problems with the glue drifting over temperature / while hardening.

I assembled my module and it was “eh ok” aligned but moved out of alignment while changing in temperature or while the glue hardened further.
I then changed my design to include small 2 DOF mirror holders which added bulk and isn’t long term stable.
Glueing is the holy grail of optic placement.

But I might try that again in the future.
Really nice that you got that to work!!

1 Like

Hey guys,

I have now fiddled with the resonator(s) for probably about 20hrs, no luck yet. I have two main problems that I didn’t really anticipate, although I definitely could have.

1. Energy density in the crystal. One issue is the focussing inside the crystal, which of course is directly connected to the pump module. Either the glue drifted a bit or builder was right and my alignment was not as good as I thought. Either way, two spots (1 green and 1 blue) overlap very well in the crystal, but the third one is way off and totally blurry, so the PBS and collimation train of the 2nd blue diode are off. Is there any way to soften/remove cured UV glue? But just the overlapping spot has 4W blue + 1W green which should definitely work. The other issue related to this is defocussing by the pump mirror. Remember that the beam is focussed through a plano-concave mirror, which increases the spot size a lot.


It about doubles the diameter in both dimensions (so only 25% power density)… No paper ever mentioned this though they must all have gone through this issue! Makes me wonder if much below ~50x50um spot size is even possible in these circumstances.
I do happen to have a f 250mm achromatic lens that corrects this a bit, but still, its about 57% of the originally planned power density.

This setup did not work either.

2. Resonator alignment. The original plan was to use a x-Fold four mirror cavity, which is reported all over the literature. Aligning this resonator has not worked out for me yet since my alignment laser (1mW HeNe) fades out quickly after so many bounces. The Pr:YLF laser at 640nm (and tens of mW at the lowest pump power) works way better, but still, no luck. I suppose the difference in index of refraction between 640nm and ~800nm leads to problems since the crystal is brewster cut. Here’s a pic of the resonator by the way:

One issue is not only making sure that the resonator is aligned but that the pumped volume also overlaps with the mode volume. I can definitely do this in principal using the 2 pump fold mirrors. I align the resonator by coupling the Pr:YLF into the outcoupler (bottom left) and overlap the spots (simplified…). One could think that I should just get the pump beam to retrace the alignment beam back to the pinhole on the bottom left, but again, the index of refraction is a problem.
So, I thought about it and sought to simplify the resonator. Although I must say that I have never seen a different TiSa Resonator other that x-Fold or more complicated. Generally its for astimatism compensation of the brewster cut crystal I think, but is that even important for cw-lasing? I don’t know.

So I started by eliminating one plane HR mirror by folding one concave mirror directly back into the crystal, essentially creating a V-fold (see 2). Didn’t work. Then I used the simplest resonator I can do with my mirrors: a hemispheric one. This one was risky since the plane outcoupler reflects some pumplight back into the diodes but luckily I didn’t destroy anything. Alignment should be no problem essentially since I’ve done that tens of times with other lasers, but still, no luck. I tried resonator 3 at first, but then thought I could drastically reduce the spot size by using shorter focal lengths for the pump lens. With 4 this is possible (remember, the resonator length for 3 and 4 is 100mm). A dichroic should then separate the pump and IR. Neither setup worked, so maybe the pump is insuffient / OC has too much outcoupling. Although in the literature, this combination has worked fine! So I am torn whether just bad alignment, bad pump quality/power density or too much outcoupling is the bottleneck here. By the way all resonators are stable at least in theory (checked with ReZonator 2.0).

Although I can report that I have seen some IR fluorescence from the crystal, though very dim:


This is on resonator 3, but also worked with 4.
Here’s also a video . I recognize this spot behaviour from my other hemispheric resonators, but overlapping the spots produced no lasing. Especially the tighter spot has high divergence and fades quickly so this is no laser action. Maybe 1 and 2 produce this as well but since the distance from crystal to OC is so much larger, it may not be detectable by my camera anymore.
So here I am, wondering if laser action is shortly around the corner (just one more hour of alignment!) or if my setup can work at all.
I have also ordered a 808nm laser module for alignment, this should have no issues regarding different index of refraction and shouldn’t fade from the reflections due the mirrors being HR for that wavelength. Have to use my camera for alignment is really annoying though. But if it works, why not.

I might also make a post on photonlexicon for that issue since I’ve read some posts where peoply claim that they have experience with TiSa, which is really rare… I only have till January to get this thing aligned (semester abroad begins at 12.01.) so I am a bit stressed out

Anyways, as always, I’m always happy about comments!
Nik

1 Like

You should be able to dissolve the UV glue in acetone.
Keep applying drops of acetone to the bonding area. After some time, the glue will get like jelly and can be removed.

Heating the joint (>150C) also helps but as you got a single plate for all optics, this would kill all bonds with leads to you starting over…

Hi all,

sorry for the long silence - I’ve been busy with other stuff. The TiSa has never left my mind though…

First of all: Thanks builder for the acetone tip, it worked fine. It took about 3-6 hrs to dissolve the glue though, didn’t expect that. After realigning the lenses the beam is definitely cleaner - although it did shift a little bit again, just like you said it would. Optics, man… So sensitive.
Also, the 808nm module works great for alignment, at least for getting it in the ballpark, after which fluorescence overlapping is my preferred way of alignment.
I did actually have some correspondence with some researchers who published papers for DPSS TiSa’s, who gave me a lot of great tips.
Mainly, to actually measure the spot radius of the diodes (which is probably too big), use a photodiode to maximize fluorescence output and check the polarization. I used the knife edge method to measure the spots, my green beam is 35x40um and the blue beam is 30x70um, which is actually pretty much dead on what the simulation predicted.
I already posted all of the details of my current situation (including videos) in the photonlexicon thread, in the hopes that mixedgas aka steve sees it. But I, of course, invite you to read the post :wink: It is long though…

I am starting to get the feeling that I won’t get it to lase until next thursday… Would be shame, but, to cite the terminator: “I will be back!”

~Nik

1 Like

… And I am back.

With a big milestone! Though I have made no progress with diode pumping, I got it to work with a Verdi. I am now working at a laser research facility and some of my co-workers had one of those left over from a project so they let me bring my setup and test it! It took some re-designing but finally… finally! I got that first spark this evening. Here’s a video of the setup working:

I’ve also attached a few pics and a first power curve. I have used a simple three mirror (actually four, one for coupling in the pump) resonator, similar to one of the first designs. The mirrors are still 100mm ROC: ROC - 80mm - plane pump coupler - 20mm - TiSa - 50mm - ROC - 500mm - plane 1% OC

The plan now is to get more data and more experience, i.e. play with pump densities, arm lengths, OCs, pump focal lengths, etc. to get an intuition on how flexible this is. Then I can make informed decisions on how to make it diode pumped. The end goal is a small and compact benchtop box that contains everything: Electronics, cooling, pump, resonator.

Once I get more data, I’ll share it - along with my alignment strategy and resonator details :slight_smile:

~Nik



3 Likes

So, finally I got it to lase! First used the Verdi to optimize at that working point and then I got it above threshold with one of my 461nm FAC diodes. It is barely above threshold and only with a near HR mirror (prob 0.1% T) so there’s a lot to do still. But: It is done, it lases only with a blue diode!

Here’s two videos showing the results:

An infrared viewer makes the spot much more visible than my smartphone:

I cant use all the diodes since they are not perfectly aligned. But I will probably redesign the entire thing anyways. Those >3 W blue diodes are just a nightmare M²-wise.

Right now I am thinking two green NDG7475s pumping from both sides… We’ll see.
Btw: I was able to do a Findlay-Clay analysis and the resonator with 500mm collimated arm length and 40mm focus lens for the pump has roughly 3 % losses. Bit too high for my taste, but I’ll redesign it anyways.

Greetings from Germany
Nik

2 Likes

Nice work! :slight_smile:

2 Likes

Great, thank you for posting the info! Greetings from Qatar.