Date: 18th Jan 2017
Time: 15:00 UK time
Attendance: Adam, Chris H, Sophie, Melissa
Agenda and Minutes¶
Action: Ed to run MC and then look at ideal helix reconstruction
Action: Paul and all to look at Adams code and run with unit test with error corrected.
Action: Ed to add radius as a func of z and phi/z
Action: Melissa to contact Sumitomo UK and find lead time on a repair on a failed cold head in an emergency, (2-3 days to warm and 2-3 days to dry which must be added to the lead time).
Action: Ed, Spare Boards still need temp on RH boards and bias calib on LH boards. Will do it in a week in January.
ACTION: Chris R will make and send to Adam (1. Geometry needs updating with 4 stations,
2. Need to switch LiH absorbers in current geom to plastic. BOTH should be simulated (Ken) so that we can say what we loose using plastic (cheap) rather than LiH (£100k).)
Data Taking Summary MAU, EO, PK, AD¶
- Reconstruction of real data MAU
- What can we tell with field PK
- MC data comparison CHeidt and EO, MAU
Efficiency DS Round table discussion,¶
(From Dec: It looks as though there are 7% US and 15% DS 5 spacepoint tracks (doublets (20% DS) and triplets) that are not reconstructed. This seems to be happening in Pat rec (hence Kalman never gets the tracks from pat rec) and is more prevalent at low momentum. //Looking at Pat rec loosening the circle cut makes no difference, loosening sz (distance travelled around the circle a straight line a s function of z) cut makes a big difference. //Pat rec makes a circle cut to contain 5 spacepoints in a circle, then calculates how many rotations a track has made through the tracker, then makes a straight line sz fit. It is this rotations code that is not working well. Adam has never been able to find a good solution. Adam has sent the code that has the problem ~100 lines using ratio of distance between tracker stations ratio of turning angle (angle of a clockface) between spacepoints. Even in ideal magnetic alignment it doesn't work well.
From Dec: Adam has tightened the sz cut and Ed is investigating this, he sees some improvement but not much. We are also still not sure why this problem is on the sz cut. 10mrad rotation observed and this is excluded. Is field non-uniformity a part of the problem and how much is it? Chris R has sent some back of the envelope calculations which suggest it could be 10% Ken does not agree it is at that level. Need to look at Adams data file, trims on trims off data and Victoria's data. Francois, has taken a look at the alignment again there is some confusion with his results downstream... Also pat rec does not agree.)
Adam has not yet implemented CMM offsets will do after demo stuff.
Melissa has looked into potential bugs with the efficiency and event selection algorithms. At present it seems as though the major cause of inefficiency in the DS for straight tracks is scattering where 5 SP exist but MAUS reconstructs a 3SP track as the tracks scatters between stations 1 and 3. This clearly needs to be included cleverly in the code. For now a simple verification must be done to ensure that the inefficiency is just this and not real. Work ongoing to prove this in a simple time effective way before the true scattering will be implemented.
Ken will look at the errors tomorrow with Adam. We do not think error propagation is discussed in paper but he will send that to Ken.
Chris R has done some back of the envelope calculations, he is of the opinion you need to understand the momentum first. Ken is unsure at this.
Next meeting Friday at 2.30 pm.
Descoped cooling Tracker design - AD¶
Melissa has calculated dark current rates expected from MTA tests and calculated Tracker fibre deterioration (discolouration) due to photons as a function of this. Chris and Melissa are talking and Melissa will write a note on her findings.
Chris H is using a 3 station geometry from Chris which includes a secondary absorber and a vacuum window. He simulates and electron and photon beams from 0-3 MeV photons. For electrons (34000) less than 10 left after vacuum window and secondary absorber. His feeling is that with one or either or both in place electrons will not be an issue in this energy range. using same geometry but with photons of 0-3 MeV it looks (from a study of 35k) 85% make it through the windows and there is a 0.3% chance of oscillation. He feels electrons not a problem photons more so.
Chris and Melissa to work together to get real simulation and then parameterise.
RF runs 1ms per target dip it 1ms/sec.
Adam has been looking at performance. He has a to do list.
1. Geometry needs updating with 4 stations,
2. Need to switch LiH absorbers in current geom to plastic. BOTH should be simulated (Ken) so that we can say what we loose using plastic (cheap) rather than LiH (£100k).
3. Add in Chris H's dark current mapper.
4. Calculate Eff US and DS (3 and 4 point tracks).
5. Performance plots.
Chris R will make 1 and 2.
Status of spares: EO CMW
- cold heads and boards
2. Software: PK, AD, CHunt, CHeidt¶
- Performance and Analysis
- Validation script
- Calibrations, Configurations
- Monte Carlo
- Pattern Recognition
- Online Displays
Next Meeting Feb 1st.
Phone Details: International Toll-Free Dial-in numbers¶
- Switzerland: 0800705351
- UK: 08004960576
- US: 1-866-740-1260
- Other numbers
- Access Code: 5673390