During recent IDT time it was discovered that downstream collimating jaws are a significant source of background. The line-of-sight to detectors has to traverse 0.5" B4C plate, so this had in the past been overlooked. The addition of substantial amounts of borated poly and further B4C plates blocking this line-of-sight has halved background levels when a DAC is in place.
Figure shows the raw focused counts from a 10 minute dataset from a ~50ug sample of (our favourite calibrant) ice VII in a diamond anvil cell at 10 GPa. The clear Bragg peaks are from the sample. The signal to background level is actually better than that achieved with the much larger volume Paris-Edinburgh cells.