In the course of analysis of the 21Na(p,gamma)22Mg data from Nov/Dec 
2002, difficulties were encountered in distinguishing recoil and leaky 
beam peaks. In an effort to distinguish the two peaks, a relationship 
between ion energy and peak position was determined.

Data was not only taken from runs involving 21Na, but from earlier 
21Ne, 20Ne, 24Mg runs and their recoils, (22Na, 21Na, 25Al, 
respectively), at different energies. 

Once plotted, it was found that five of the 21Ne runs (6898-6931) did 
not seem to follow the general trend (see Figure 1). In addition, the 
alpha source peak was not at the expected energy channel.

The solution: an error in the gain and offsets used in PAW++ for those 
runs. Looking at the odb file for run 6925, we found that the gains and 
offsets were different to the usual PAW++ files. Instead, the mid files 
were rerun in NOVA, with the original settings, to find the correct 
energy channel number of the peaks.