A little bit of trigonometry...
Some theoretical elements can allow us to
better understand needs and constraints to consider for the
realization of a spectrohéliographe:
 Angular dispersion of a reflective grating :
sin(α) = k . λ / (2 . p)
 Linear dispersion in the spectrum plane :
D = p . cos(α) / (k . f)
 Resolving power :
R = k . N
 Spectral resolution :
Dl = λ / R
with :
α = diffraction angle (in rd)
f = focal length of the camera lens (in mm)
k = order of spectrum
λ = wavelength (en nm)
N = total number of grooves of the grating
p = groove spacing of the grating = 1000000/number of grooves per mm (in nm)
R = resolving power of the grating
 example:
For a 1200 gr/mm grating enlightened perpendicularly to the
surface, the wavelength of the Ha line is diffracted in the
first order according to an angle of :
a = arcsin( 656.3 /(2 x 1000000/1200) ) = 0.4047 rd = 23° 11',4
With a 500 mm f.l. camera, the linear dispersion is :
d = 1000000/1200 . cos(0.4047) / (1 . 500) = 1.5 nm/mm
A 30 mm aside grating has, in 1st order, a resolving power :
R = 1 . 1200 . 30 = 36000
At this wavelength, one can resolve, in theory, two lines separated by :
656.3 / 36000 = 0.018 nm
See this page for more informations about gratings
... geometry ...
Here is a layout of spectroheliograph :
 Lens of telescope : Aperture D1 and focal length F1
 Collimator : Aperture D2 and focal lengthF2
 Camera lens : ApertureD3 and focal lengthF3
 Width of the entrance slit : S
 Width of the grating : C
The telescope gives a Sun image diametre = F1 . tg(alpha) with alpha = apparent angle of the Sun, about 1/2 degree.
focus plane of telescope, focus plane of collimator and entrance slit are merged.
The spectrograph gives a magnification of this image = F3/F2 . It is possible
to magnify or to reduce the size of the solar image on the sensor. Width of the
entrance slit image varying in the same ratio, it may be interesting to use long
f.l. F1 and F2 with a rather large slit, and to reduce final image with F3 if
the sensor is small. In the case of Littrow mount (collimator and camera are
the same), magnification is obviously = 1.
The grating must be totally illuminated to obtain maximum resolving power. this
implies to use a collimateur with aperture D2 = 1,414 C or greater. If you have
only a 40 mm aperture collimator, a 1 pouce width grating is sufficient. This
lens must also be illuminated in the same ratio (1.414 C) by the beam comming
from the entrance slit of the spectro . We can see that, aroud the slit, the
incident and emerging angles are equals (in yellow) and the F2/D2 ratio must
be equal to F1/D1 (or smaller).
Camera lens must have the same aperture D3 than D2.
... and a spectroheliocomputer (use mm)
