The Maksutov Approximation



A simple lens has inverse image distance:

       1/q = ( 1/p + (n-1) / r1 ) / ( 1 - (t/n) * ( 1/p + (n-1) / r1 ) ) - (n-1) / r2

where n is the refractive index, t is the lens thickness, r1 and r2 are the first and second surface radii, respectively, and p is the object distance.

Differentiating with respect to refractive index:

        d(1/q)/dn = [ 1 / r1 - ( t / n2 ) * ( 1/p + (n-1) / r1 )2 ] / [ 1 - ( t / n ) * ( 1/p + (n-1) / r1 ) ]2 - 1 / r2

Achromatism occurs when this derivative is zero:

         1 / r2 = [ 1 / r1 - ( t / n2 ) * ( 1/p + (n-1) / r1 )2 ] / [ 1 - ( t / n ) * ( 1/p + (n-1) / r1 ) ]2

Or:

         r1 - r2 = (t*r1/n2) * ( 1/p + (n-1) / r1 ) * ( 2 * n - ( r2 + t ) * ( 1/p + (n-1) / r1 ) )

The simple case with infinite object distance ( 1/p is zero ) is solved here:

         r1 - r2 = (n-1)*t/n2 * ( n+1 - (n-1)*( r2 + t - r1 ) / r1 )

The hypothesis here is that the term:

        (n-1)*( r2 + t - r1 ) / r1

may be ignored. If so, then the approximation is:

         r1 - r2 = (n-1)*t/n2 * ( n + 1 ) = ( (n2-1)/n2 ) * t

However, suppose the approximation is used. Then the ignored term is:

        (n-1)*(t - ( (n2-1)/n2 ) * t) / r1

Or:

       ( (n-1)/n2) * (t / r1)

As long as this term is small relative to n+1, the approximation is valid.

UP