Figures 1a,b show the continuum spectrum produced by the AB model
compared to that produced by the Pella model, for a Cu target at
5 and 30 kV. The takeoff angle for both of these
simulations is taken to be 
. There is agreement within
a factor of 
between the Pella and AB model over a wide range of
photon energies, with larger discrepancies for lower
(
keV) energies and also toward the target cutoff energy. Figure 2
shows a comparison of 
and 
line intensities, as a function
of target voltage. The AB model produces smaller line yields at lower
target voltages. However, the Pella model may also produce insufficient
line intensity for lower Z targets (see below).
Figures 3a-d show the comparison of the AB model for Mo, Ti,
Cu, and MgO targets with data taken with a Manson Model 5 multi-anode
source (takeoff angle of 
)[3,4].
A solid state detector (
eV FWHM) was used to collect the data, and the detector's
resolution and efficiency have been included in the model
spectra.
Figure 1 Caption: The x-ray continuum spectrum produced by the Pella model (solid squares) and
the AB model (open triangles) for a Cu target at (a) 5 kV and
(b) 30 kV. The
takeoff angle for both simulations is 
.
Figure 2 Caption: The 
and 
line intensities produced by the Pella (solid symbols)
and AB (open symbols) models for a Cu target, as a function of target voltage.
Figure 3 Caption: The x-ray spectra measured from a Manson source for (a) Mo, (b) Ti, (c) Cu, and (d) MgO targets, compared with the AB model simulations. Detector effects are included. Also shown (solid line) are best fits provided by a simple Kramers model for the continuum; line emission has been also included, but the line intensities have been arbitrarily adjusted to improve the fit to the data.
The plots show agreement of the model and data within a factor of
2--3 over a wide range of energies, with larger disagreement
at photon energies near the cutoff energy and below 
keV.
Both the model continuum and line intensities appear to be
lower than the data by similar factors, and all four target models
are all lower than the data by similar factors. Note that only the model
for the MgO target has more than one element (Mg and O) been included;
no attempt has been made to incorporate the contaminants seen in the
spectra (C, O, Si, W, etc.). Also included in these plots are simple
fits of the data using the Kramer's formula[10] for the observed continuum,
and line intensities calculated from Pella.
No attenuation effects by the target have been included.
In many cases ( i.e.,\
lower-Z elements) the Pella-estimated line intensities were substantially
smaller than that observed[3]. Therefore, the Pella-estimated line intensities
for the have been arbitrarily adjusted to provide a good fit to the data.
