General Meeting Summary 8/3/23

PARTICIPANTS: Simona Malace, Alexandre Camsonne, Ben Raydo, Bryan Moffit, Carlos Munoz-Camacho, Charles Hyde, Christine Ploen, Ed Kinney, Tanja Horn, Hamlet Mkrtchyan, Hao Huang, Jiwan Poudel, Julie Roche, Malek Mazouz, Mark Jones, Mark Mathison, Mitch Kerver, Vardan, Wassim, Yaopeng Zhang, Yeran Ghandilyan, Zheng Huang, Chanfan Ghosh, Deb Biswas, Erik Wrightson, Sanghwa Park, Paulo Medeiros, Rolf Ent, Aaron Brown

CALORIMETER TRIGGER DISCUSSION (Ben)

• Discussed three main topics as listed below

CLUSTER RANGE

• clustering on VTP modules, look at all 3x3 regions in parallel; trigger on clusters that exceed the cluster threshold

• requires a local maximum - if no local maximum then no trigger
• local maximum depends on calibration of calorimeter - if calibration off because crystals lose transparency, then shift center to another crystal and 5x5 will be done about another 3x3 cluster (not the maximum as should have been)
• timing coincidence parameter: time evolution of hits (at zero one hit, and then additional hits) - seed hit must also be local maximum in both space and time

• if trigger cluster then readout 5x5 full waveform centered on clusters - second threshold: read out all clusters that are greater than the readout threshold

• can the cluster range be extended to 7x7? - yes, but this increases data volume and might saturate the data link (saturation limit 20kHz, 50 channels in one crate), threshold-dependent; need to check data volume for larger clusters
• what drives 7x7? - if centered on crystal then get 5x5 easily, but if at corner (or edge) then could get larger than 5x5; 6x6 may be acceptable

• Action item: implement 6x6 and a flag that allows to chose 5x5 or larger clusters

TIMING EXTENSION

• online clustering is 3x3; when have a 5x5 (6x6) then have to send bits to the adjacent VTP (not sending full waveform, but the bits)

• pulse processing - what are the parameters, e.g., the time window?

• seed threshold of 2 and hit delta t = +/-8ns (programmable up to +/- 16ns, the limit right now; in principle can add 32 or 38ns)
• can the nsa time window can be increased? - yes, in principle it can be increased, but there is concern about the data rate and dead time --> 32ns dead time (single channel)
• if nsa increase then overestimate integral (background + signal), but timing is that of background; downside: 16ns or less coincidence, if increase to 48ns, then background huge and may also shift local maxima by background
• if want low background --> small timing coincidence
• if accept huge background --> large timing coincidence
• since all is programmable could make changes online

• Actions:

• check out parameters that are presently in code - stored with the data run
• see npslog entry for how to get VTP configuration file: https://logbooks.jlab.org/entry/4164991)

• HIT_DT currently set to +- 9ns (in discussion above was +- 8ns)

• make run plan to check out things as taking data, e.g., when taking cluster data (example of such a run: https://logbooks.jlab.org/entry/4155466); good to include a screenshot of the configuration files in these runs

MULTI CLUSTER TRIGGER

• Correlation of clusters in different VTP

• Action: add bit that indicates these multi-clusters

ADDITIONAL DISCUSSION

• Event == any time doing a read out

• every nth event reading out all channels
• can disable a readout if don't want it

• DiagGUI

• tool to look at signals: what do noise and pulses look like - it's like an oscilloscope
• runs in parallel to DAQ
• monitors only calorimeter channels - only for a particular fADC (all 16 channels of it), cannot correlate other fADCs, i.e., cannot trigger on external channels

MECHANICAL UPDATE (Paulo)

• All progressing well and on schedule - new clam shell around beam pipe ordered (should be in within 1-2 weeks)

• Magnet installed

• installing beamline supports and components

• also installing lead by Q1

RUN PLAN (Charles)

• Run plan being prepared in Overleaf - link to the document on the Hall C Wiki

• Send any comments to Charles (hyde@jlab.org)

NEXT 1 HOUR LONG MEETING: THURSDAY 31 AUGUST AT 9:00AM ET