[Raytrace] Ross Null

[Robbin Palmer]

More confusion over Ross Null test. Using an f/4.2, D= 22.4 mirror, f= 
450 mm, D = 145 mm, t = 15.5 mm lens.  I used the software program from 
Mel Bartel's website to find the lens-source and lens-mirror distances. 
Got source-lens = 255.657mm, lens-mirror = 4145.66mm, which is at least 
in the same ballpark as the numbers, 250.29mm and 4159.71, that I 
calculated from the Stoltzman and Ceravolo Mar/93 Vo 32 No.7 Applied 
Optics article using green light (entered the index for BK7 green light 
in the above software - 1.518722. So I entered the above numbers from 
the software into OSLO lite.  Here is  how I set it up:
                RAD            THK            AP RAD            Glass    
        Special
OBJ            0                255.657                  S            
    air                       
AST            0                15.5                        S           
    BK7 C                   
2          -238.719       4145.7                        S                
air                           
3        -4775.2          -4145.7 P                    S          Refl 
Hatch            CA
4           238.719 P        -15.5                      S               
BK7 C                   
IMS            0               2437.4  V                                 
                               

The thickness value for the image was calculated by OSLO.  I tried it 
using the source-lens
thickness as a variable, as well as the lens-mirror thickness as a 
variable. Results were no better.
With the above I got a P-V OPD of 9.44! Plus the image plane is far away 
from the source. I need it the same as the object plane for my slitless 
tester.   The software from Mel's site gave a max OPD of 0.0692.  I 
tried iterating the source-lens distance by hand to improve things, to 
no avail. What am I doing wrong?  I calculated lens radius of curvature 
from r = (n-1)(f + th/n).  I used the lens center thickness for th.  The 
catalog gives the principle plane location at 10.2mm from the planar
surface. Should I adjust for this? I varied the thickness a few mm with 
no improvement in results.
In fact, it is an enormously robust bad solution.