Without knowing all the details...why not move the table? This can be done easily, rapidly and with high precision. If not, then you will likely need something like a genetic algorithm (linked to a deformable mirror) to keep the wavefront clean and keep the focus stable (this is easy at a kHz) when moving either a steerable mirror, or lens to redirect the beam. The precision of this is then related to how stable your optics are, drivers, etc. However, I think this is much more complicated then moving a table/platform. IMHO. Perhaps you can supply more detail as to the application?

Thank you Kirk, and sorry for the absence of details.

I am afraid unfortunately moving the table is not very possible.

The laser beam shall interact wioth the human eye, so moving the patient bed is not what may help.

We are in the search for a focus shifter with large amplitude and response frequency, high dynamic range:

Focus shift system with ~+-25D range and response of "several 100-Hz" (1KHz ideal, 350Hz just acceptable in an optical pathway with pupils from 1 to 60mm diameter: actually 1mm, 2mm, 4mm, 8mm, 16mm or 60mm (I mean we have the possibility to integrate the focus shifter where the pupil is 1mm, or where the pupil is 2mm, and so not, not that the pupil shifter shall manage all pupil sizes "at once")

As further details we are using a laser beam 355nm, and

Combination of two lenses (of different focal lengths) with one of them moving should be the trick. This should be independent of the rep rate (KHz). High peak powersof your ns laser than a mJ can also change focus due to self -focussing. So, one needs to be careful for the average power.

Dear Samuel, probably we can do such a precise movement through coupling the laser output to fiber cable and then focusing optics. This combined assembly can be driven through Pizeo systems.

What about a solution like barcode scanners? you have a prism that rotates and deflect the laser beam.

The bladeless Lasik machines from Ziemer have a commercial system for doing this. There are many papers from IntraLase from work done at the University of Michigan's Center for Ultrafast Optical Science and the Kellog Eye Center by Ronald Kurtz and Tabor Juhasz, look up their articles, they have solved this problem very precisely.

http://www.abbottmedicaloptics.com/products/refractive/ilasik/intralase-fs-laser

These videos a cool, and a bit gross.

http://www.youtube.com/watch?v=Ved91TJ9g9Q&feature=related

http://www.youtube.com/watch?NR=1&feature=endscreen&v=36_4ri464u0

http://www.youtube.com/watch?v=eFqF-ycN8lM

If you can combine GRIN optics with the wavelenght shifting in close loop, as proposed by Mykhaylo, it would be a nice, clean and efficient way of doing it. Zone plates are not as fine in terms of final beam shape and quality as gradient index optics, but I'm not sure how easy it would be to find a comercial solution for your particular system. We usually get them custom from the GRIN research group of the University.

If, instead of fast focal lenght switching, you want to artificially extend the DoF far beyond the gaussian limit, then a nice trick is using Bessel beams, if you have very good power stability in your laser source. You can get Bessel beams with an optical axicon, and it produces an impressive depth of focus out of a collimated gaussian beam. It is a completelly different way of doing the job, it will not allow you to focus the beam in a point in the space, but if you work with UV you can very precisely machine on 3D surfaces without caring for the focal lenght.

I wouldn't reccomend SLM for this particular application. I imagine you don't want zero orders or higher orders in undesired locations in the patient's eye!. Instead, adaptive optics cound be very nice, if they are coupled with a HS-type wavefront sensor. Optotune optics similar, they are lovely, but I don't think they will work kHz... unless you want to pulse kHz but focus-tune slower.

Thank you Mykhaylo,

this seems a very promising approach and may perfectly fir my needs

great, I will keep you posted

but other ideas are still welcome (I prefer exploring more possibilities before making the wrgon decision ;-))