Sphenix-calibration-l mailing list

[Hugo Pereira Da Costa <hugo.pereira-da-costa AT cea.fr>]

Hi all,

the notes from Tuesday's meeting are at 
https://indico.bnl.gov/event/9302/attachments/30251/47296/notes.txt and 
below.

As usual, comments/corrections are welcome,


Ross and Hugo

-------------------------------------------------------------------------------------

1/ Instrumentation of the central membrane for reading currents (Tom)
This would provide a direct measurement of the primary*Gain*IBF, rather 
than using Anode currents (primary*Gain) and multiply by an estimated IBF.
Problem:
- what are the expected currents (rough estimate would give 25uA on the 
full anode, Evgeny will work on getting more accurate estimate).
- which segmentation is desirable (balance between distortions lumpiness 
and current magnitude)
- are the Zagreb picoammeter up to the task ? <- can measure very low 
currents but need integration over 'large' times, to average the noise out.
We need measurements with a granularity ~few ms. <- no averaging. Will 
critically depend on the noise level
To be studied in more details

Q: how would that play along with the current induced by the diffuse 
lasers ?
A: the latter is likely negligible, but needs double check.
TODO: at some point would be nice to have the amount of charges 
generated in the TPC (and the current in the central membrane), by the 
diffuse laser, properly documented and compared to charges and currents 
from collisions, since the question keeps coming up.

2/ Discussion on Micromegas integration (Klaus, Maxence)
Q: how final are Micromegas drawings ?
A: not final at all. Everything can be changed

Q: is it possible to mount (possibly clip) the detector on the TPC 
carbon rods ?
A: in principle yes.

Q: what is the purpose of these rods ?
A: they relieve some twisting stress on the central membrane.

Q: what would be the preferred orientation of the Micromegas (long side 
along z, or along phi)
A: on the design point of view, both are possible. The long side however 
is slightly longer than the inter-rod distance. For the short side, the 
detector would fit between the rods. Would need extenders on the side to 
be clamped on the rods
On the tracking POV, having the long side along z is preferred, 
especially if the detector is placed at mid-rapidity, on both sides of 
the central membrane.
It is not necessary to have the full azimuth coverage of a given TPC 
sector to get information on the full sector, due to track bending.

Q: the occupancy on the detector is quite large (14% on average, 35-40% 
for most central collisions). How doable is it to cut the strips in half ?
A: it is doable provided that we have enough electronics. Other 
solutions could be studied: reducing the pitch, while at the same time 
reducing the amplification and drift space, as well as optimizing the 
gas, in order to keep the cluster size constant, would also help. 
Detailed studies ongoing by Maxence.

Q: will the Readout electronic be SAMPA ?
A: yes: no choice, validating any other electronic will take too long. 
Beside, SAMPA is probably perfectly suited, and comes for 'free' thanks 
to successful pre-production

Other aspects to consider:
- cooling. Will be necessary. You do not want a "local" hot spot on the 
outside of the TPC
- available radial space
- alignment of the detector


3/ Update on distortion map calculations (Ross)
several updates on SC maps: now have the right scale with respect to 
past estimates (within a factor 2 at worse). Will double check agreement 
with finer granularity

Q: with new maps, phi distortions are larger than r distortions. correct ?
A: yes. for r > 30cm you get 1mm max in r, and 4mm max in phi

Q: which distortions has the largest impact on momentum resolution
A: phi distortion directly impacts the momentum measurement. R 
distortion i) is smaller ii) only impacts momentum measurement at second 
order.

TODO: there is a remaining scale factor missing between charge maps 
generated from model by Ross and those from Evgeny. (this is now 
resolved.  Evgeny's current generation of maps uses total charge per 
bin, not charge density per bin)

Discussion on gas mixture:
Q: right now everything is done with Ne/CF4 90/10. Final gas will be 
Ne/CF4 50/50. Should one update
A: yes.
Will need to re-run GEANT, and redo the space charge maps

Q: how long does it take to generate the distortion maps:
A: 10 charge maps takes ~6hours. But can be submitted to the batch 
system. Distortion maps significantly faster.

4/ impact on distortions on momentum/mass resolution
- even if old, manually-scaled map is used, phi distortion is of similar 
magnitude (possibly x2 smaller) than what is expected with new, 
corrected maps
Q: are the momentum/mass resolutions w/ distortions after additional 
corrections ?
A: no. No attempt at correcting the distortions. The idea is to quantify 
the impact of not doing anything for the distortions

Q: does this include fluctuations ?
A: there are some spacial fluctuations, because the map is based on real 
HIJING events, but not time fluctuation, because the same map is used 
for all events. It can only get worse when including event/event 
fluctuations

Q: do you have equal amount of positive and negative charges in the 
momentum resolution plots ?
A: yes. Will show the residuals for both separately, to verify that the 
shifts go in opposite directions.

5/ Update on SC maps generation and anode currents (Evgeny)
- working on generating anode currents from SC maps
Q: what are the fluctuations you add in the process of doing so ? (for a 
given event, SC depend on primary, gain and IBF, whereas anode current 
depend only on primary x gain).
A: none for now. Will be added in a later version.

Q: is the description of the pads correct ?
A: in R1, the pads start at r = 30cm and are twice shorter than in R2 
and R3. Concerning the width (=pitch) of the pads, it is 2mm on average. 
This is probably too high a segmentation for reading the currents (the 
model to convert the currents into distortions, from Ross, will likely 
be much coarser. The detailed description of the pads is therefore 
largely irrelevant (unless you also want to add electronic noise and 
electronics response in the chain)

Discussion on R1 "test" pads (for r < 30cm): will be used to read the 
currents only. Segmentation still to be decided: does not need to be as 
finely segmented as for the tracking part, but at the same time one must 
not saturate the SAMPA, otherwise the current reading is meaningless.
Need inputs for Ross/Evgeny. Heuristic model can estimate charge density 
on those pads, new HIJING in correct gas will be better.

TODO list:
1/ understand discrepancy between Ross and Evgeny on SC maps
2/ change gas mixture
3/ produce time averaged map from heuristic model
4/ produce per-event maps from HIJING
5/ produce time-averaged map from HIJING
6/ subtract 5 (or 1) from 4/ to get fluctuations on top of time averaged 
maps
7/ re-run 4/, 5/ and 6/ on single particle simulations to better 
understand the impact of time-averaged distortions and of fluctuations 
on the momentum/mass resolutions, separately.
8/ try reconstruct and correct for the distortion, using tracks + 
Micromegas, on HIJING events