we are noticing we appear to see less implant events in detector 3 than we expect. This was first noticed in the xy spectra of the DSSD in the Go4 online.
As part of testing we have lowered the threshold in aida09 which from the xy appeared to have less events.

Following this we still seemed to see less events.

OH downloaded a selection of recent files to analyse. Looking at the total number of HEC events it appears that aida09, aida10 and aida12 are all seeing less HEC events than I would expect.
The ratio of HEC events in aida09 to aida11 was 0.36.

Looking at the energies of the HEC events there appeared to be less events everywhere and not just a threshold issue. We thought it could then be radiation damage and poor
charge collection. To test this we raised the bias to 120V on DSSD3.

The same test was again performed and the ratio again came out to be 0.36
We did not appear to recover any HEC events

Comparing the scaled energy distributions of aida09 and aida11 we see no evidence of a different energy response other than an increase in lower energy events in aida09 due to the lower threshold

I now think it may be a hardware issue. Will look at the last file before the most recent powercycle.

Using TD's analyser program can confirm the rate between aida09 and aida11 before 08:30 this morning was ~1:1.
The problem seems to have appeared during the power cycles following this.