Solved

# Merit functions for light collimation

• 6 replies
• 3221 views

• Monochrome
• 2 replies

I would like to use lenses from Thorlabs to collimate the NA=0.22 beam to around 0.4mm size. What merit functions should I use to achieve this goal? Can I also use Stock lens matching and merit functions together to achieve this goal?

icon

Best answer by Mark.Nicholson 3 November 2020, 18:23

View original

### 6 replies

Userlevel 7
+3

Yes that's right Steve: the Angular criterion reduces RMS angle the same way that the spot criterion reduces RMS spot

To set collimation at surface i, and spot (say) at surface j, use the IMSF (Image Surface) operand in the merit function to redefine the image surface on the fly. So in the MFE:

```IMSF i #set surface i to be the image surface #add operands here using the Start At dropdown in the Optimization wizard IMSF j #set surface j to be the image surface #add operands here using the Start At dropdown in the Optimization wizard```

- Mark

Userlevel 7
+3

Use angular radius and drive it to zero to get a collimated output beam

+2

Are there specific operands that give angular radius?  Or are you saying to choose Angular as the Optimization Function Criteria in the Optimization Wizard?

If the collimator is the first lens in a system, how might the optimizer be set to optimize collimation at a surface just after the collimator and then some other criteria later in the system?

Yes that's right Steve: the Angular criterion reduces RMS angle the same way that the spot criterion reduces RMS spot

To set collimation at surface i, and spot (say) at surface j, use the IMSF (Image Surface) operand in the merit function to redefine the image surface on the fly. So in the MFE:

`IMSF i #set surface i to be the image surface #add operands here using the Start At dropdown in the Optimization wizard IMSF j #set surface j to be the image surface #add operands here using the Start At dropdown in the Optimization wizard`

- Mark

Hi Mark,

I tried that method and it didn’t keep the beam collimated at the intermediate, temporary image plane.  I have my own custom merit function that is before the suggested combo of default merit functions, which I would guess shouldn’t cause a problem, but maybe I’m wrong?  Any other suggestions for someone wanting to keep a beam collimated at one surface while also optimizing the beam’s focus at the final image plane?  Maybe I’m not using the wizard for the angular criterion correctly?  I followed your general guideline, including those terms after my custom merit function, and using the defaults for the two image criterions in the optimization wizard, except for the start at.

Thanks for any advice you can provide!

Stephanie

Optimizing for collimated beam in the system and “focus at the final image plane” only works if there is a simultaneos optimum for both.

Checking for collimated beam diameter without even looking at imaging quality can be done with the REAY operand. Use the Aperture definition Object Space NA. REAY with Hx=Hy=Px=0 and Py=1 gives the beam radius, assuming the edge of the aperture is a well suited definition for the “beam diameter”.

Operand RAID on the same ray can be used to check for the beam beeing collimated on any plano dummy surface, if imaging quality is asured elsewhere in the merit function or the design constraints of the system.

Axel

Yes that's right Steve: the Angular criterion reduces RMS angle the same way that the spot criterion reduces RMS spot

To set collimation at surface i, and spot (say) at surface j, use the IMSF (Image Surface) operand in the merit function to redefine the image surface on the fly. So in the MFE:

`IMSF i #set surface i to be the image surface #add operands here using the Start At dropdown in the Optimization wizard IMSF j #set surface j to be the image surface #add operands here using the Start At dropdown in the Optimization wizard`

- Mark

Hi Mark,

I tried that method and it didn’t keep the beam collimated at the intermediate, temporary image plane.  I have my own custom merit function that is before the suggested combo of default merit functions, which I would guess shouldn’t cause a problem, but maybe I’m wrong?  Any other suggestions for someone wanting to keep a beam collimated at one surface while also optimizing the beam’s focus at the final image plane?  Maybe I’m not using the wizard for the angular criterion correctly?  I followed your general guideline, including those terms after my custom merit function, and using the defaults for the two image criterions in the optimization wizard, except for the start at.

Thanks for any advice you can provide!

Stephanie

Hello stephanie,

You can also try the merit function operator RANG which optimize for the the ray angle with the Z-axis. you can choose the surface in which you want the rays to be collimated and make the target equal to zero

Khaled