Slide 18 of 27
Things to check in SUMMARY
- Quickly check through all the graphs to make sure that there are no sharp discontinuities, and that all the graphs vary smoothly through the data processing run. Some parameters will not change at all (e.g. cell parameters should not be refined during the integration run), while others will drift slightly.
- Check carefully that YSCALE is not changing by more than 1 part in 1000. Likewise, DISTANCE should be constant for a properly centred crystal; a large random variation in either of these indicates that postrefinement was not performed correctly - the cell parameters may be wrong, and the machine parameters may be unreliable. A regular variation (following an approximate sine curve) in DISTANCE indicates that the crystal is precessing about the true axis of rotation.
- Check the Residual and Weighted Residual; both should remain roughly constant, but the weighted residual should be ~1.0 for a laboratory source; values higher than this indicate an incorrect GAIN for the detector, while smaller values are common for data collected at synchrotrons.
- Check the I/sig(I) graphs; is the crystal dying? Does it diffract at the highest resolution specified in processing?