In the morning we reinstalled FEE64s aida06, aida13, aida07. The rails of the FEE64 boards were found to be incorrect (that of the rails that were used in CARME).
After replacing the rails on the abovementioned FEE64s we powercycled AIDA and found the following results in the histograms and statistics.
As can be seen in attachement 4, aida16 has a noisy waveform and we are going to investigate this further.
When setting the threshold to 0x64 (1 MeV) on the slow comparator threshold we saw that the low energy branch of aida04 was still noisy. (Attachment 6)
- Setting the slow comparator threshold to 0x32 (0.5 MeV) introduced noise into aida08, aida16 and substantially more noise in aida04. (Attachment 7)
- Setting the slow comparator threshold to 0x14 (0.2 MeV) introduced noise into aida11 and substantially more noise in aida04, aida08, aida16 and aida06. (Attachment 8)
- Setting the slow comparator threshold to 0xf (0.15 MeV) introduced more noise into the same aforementioned channels aida04, aida06, aida08, aida11, aida16 (additionally aida02), barring hot strips. (Attachment 9)
We covered the end of the snout with a black cloth and this reduced the noise in some of the channels (Attachment 10 and 11).
- We covered the end of the snout with two layers of aluminium foil - this produced the same result as covering the snout with the black cloth (Attachment 12, 13, 14 & 15).
We resumed after lunch.
- Replaced the aida16 adaptor PCB with a spare, and this produced a much more sensible rate and hitpattern in aida16, which could lead us to believe that the aida adapter card was faulty or misaligned. (Attachment 16, 17, 18 & 19).
- Placed two LK1 back on the n+n side of both DSSSDs on aida06 and aida02. The noise and rate did not change at all. (Attachment 20 & 21)
Pulser tests:
- The test + pulser was connected to the p+n sides (top and bottom) at 25 Hz.
- After checking the ASIC control 2 times over, the pulser seems to be ok for all p+n channels excluding aida01, aida13 and aida14. (Attachment 23)
- The pulser peak width of aida09 1.8.L was found to be c. 55 channels FWHM on uncalibrated scale which corresponds to 35 - 40 keV FWHM in energy resolution. (Attachment 24)
- After reseating the adaptor board PCBs we found the same result as before reseating the adaptor boards (Attachment 25).
- Connecting the test - pulser to the n+n side we found that most of the ASICs behaved normally which ASIC 1 in aida04 was showing abnormal behaviour. (Attachment 26, 28). The test + pulser loop is still in place.
- The pulser peak width of aida02 1.8.L was found to be c. 118 channels FWHM on uncalibrated scale which corresponds to 95 - 100 keV FWHM in energy resolution. (Attachment 27)
- We set the thresholds on the n+n side to 0xf and the p+n side to 0xa. Subsequently, we observed improved noise conditions in many of the channels. The still noisy channels are aida01, aida04, aida08. (Attachment 29, 30). This might be because we have introduced an addtional ground loop with the pusler tests, but could also mean that the ground loop that is currently being used is not sufficient or poorly implemented.
Summary:
- From the pulser test it is clear that aida14 ASICs 1&2 are not in good shape and appear to be faulty.
- We managed to improve from 1 aida module in the sub 20 kHz rate range to 6 aida modules in the sub 20 kHz rate range by the end of the tests.
- We set the slow comparator thresholds on the n+n side to 0xf and the p+n side to 0xa. Subsequently, we observed improved noise conditions in many of the channels. The still noisy channels are aida01, aida04, aida08. (Attachment 29, 30). This might be because we have introduced an additional ground loop with the pusler tests, but could also mean that the ground loop that is currently being used is not sufficient or poorly implemented.
|