General Meeting Summary 4/15/21
PARTICIPANTS: Aaron Brown, Alex Camsonne, Fernando Barbosa, Bogdan Wojsekhowski, Carlos Munoz-Camacho, Carlos Yero, Rolf Ent, Hamlet Mkrtchyan, Paulo Medeiros, Tristan Anderson, Vardan Tadevosyan, Tanja Horn, Hakob Voskanyan, Marie Boer, Steve Wood
ANODE CURRENT (Fernando)
- Tests carried out with the divider to investigate origin of noise (divider itself or measurement conditions) for different dynode shortening scenarios (dynode 9 and dynode 10)
- background: dynode 9 is safest for PMT anode current based on previous studies reported at 4/1 NPS meeting, but noise was a major issue.
- found that source of the noise is the divider itself
- Path forward: design a new divider with the changes observed so far and with a new, externally powered amplifier
- Design new divider
- Finalize production by June
- Connection divider to socket and testing
- Impact of magnetic field on low HV PMT - no difference to before since main impact is distance to first dynode and that does not change.
- Action Items
- Prepare a summary of studies and optimized divider design based on them at the next meeting including:
- Anode current for optimized design
- Zero dynode? - generally the goal is: to stabilize gain as function of intensity (depends on the maximum current); need to determine if this is needed or not
- Working with Chris C. to develop a way forward on the few HV qualification issues that remain. No resolution yet.
- LED driver boards are in Procurement. The initial quotes were a little high - a few more vendors will be contacted. Should still be able to get them fabricated and on-site by this Fall though -- so no schedule impact.
DSG UPDATE (Aaron)
- HV supply cables tests - no load tests
- overall stable - good!
- Module Voltage vs. current tests - all channels
- Variation ~10mV
- Detector cart ready for procurement
- Layout for detector cabling ready
SOFTWARE (Carlos Y.)
- Finished clustering algorithm (Cellular automata) - based on Hall A DVCS
- Compared total number of clusters with different algorithms (new cellular automata vs. standard hcana)
- main difference is in definition of neighboring cells
- new method produces more clusters
- dependence on granularity
- Optimizations ongoing
- Action items:
- calclulate centroid of the cluster
- project to target - line to target since it is neutral
- correct for vertex
- NPS simulations input to algorithm - Carlos M. will check with Ho San
NEXT MEETING: THURSDAY 29 APRIL AT 9:00AM ET