Choose timezone
Your profile timezone:
INTT Weekly meeting
→
US/Eastern
Description
[Genki]
- Collisions is on-going now.
- 12x12 bunches for 66522.
- 10 chips are masked since they don't get bias. This issue was already seen in pedestal data taken in February.
- Ask Martin to let us know whenever they change something on GTM, so that we can check if the timing is changed or not.
- DAC0 scan analysis needs DST with hot channel mask. Genki will produce the channel mask and announce it at Mattermost.
[Itsuka]
- Cheng-Wei's interval cut is not applied in the analysis.
- If the trigger rate get higher, the fish bone structure will appear.
- Currently DAC0 is 35 comparing to DAC0=18 for Run23.
-- Even a similar DAC0 value, Run25 data looks less noisy, probably thanks to the fact that better hot channel mask is applied.
- No vertex cut is applied to the plots.
[Ryotaro]
- The timing scan has been performed with 6 x 6 bunches without beam dump.
- L1delay = 107 is better than 113 because GTM is sometimes off by 1 clock (=6 BCOs) due to GTM update.
- The last year default busy window is ~20 BCOs but now increased to ~30 BCOs, which is requested by TPC group.
- Yuko will do chip saturation study.
[Mahiro]
- The crossing angle of beams for Run66522 was zero.
- You shouldn't apply such a tight theta cut (dtheta~0.4) which is cutting so many combination making some biased structure.
- Use X-Y position from silicon seeding but not INTT vertex finder.
- Z-vertex resolution got worse with Ncluster cut but the shape is reasonable within the resolution.
- X-Y vertex from silicon seeding is not applied event-by-event, but just one integrated mean position.
[Jaein]
- Tracking is made by matching between the silicon seeding and TPC seeding.
- Z-vertex position is shifted from 0.
- No vertex cut is applied to the plots.
[Ryotaro]
- The timing scan has been performed with 6 x 6 bunches without beam dump.
- L1delay = 107 is better than 113 because GTM is sometimes off by 1 clock (=6 BCOs) due to GTM update.
- The last year default busy window is ~20 BCOs but now increased to ~30 BCOs, which is requested by TPC group.
- Yuko will do chip saturation study.
[Mahiro]
- The crossing angle of beams for Run66522 was zero.
- You shouldn't apply such a tight theta cut (dtheta~0.4) which is cutting so many combination making some biased structure.
- Use X-Y position from silicon seeding but not INTT vertex finder.
- Z-vertex resolution got worse with Ncluster cut but the shape is reasonable within the resolution.
- X-Y vertex from silicon seeding is not applied event-by-event, but just one integrated mean position.
[Jaein]
- Tracking is made by matching between the silicon seeding and TPC seeding.
- Z-vertex position is shifted from 0.
-- XY has satellite peak around 0. This may be just artificial. Check Joes code later.
- This analysis is based on DST and TTree made by Joe.
- Slide 3, this plots contains all tracks selected by Jaein's algorithm.
- Slide 4 should be a very peripheral event.
- Silicon seeding doesn't require TPC to make XYZ point.
- MVTX will check data with 111x111 with streaming, trigger.
[Takahiro]
- The plot shown here must be log scale, but need to double check.
- Ladder control works well.
- Channel 127 has some entry even after masked, which must be excluded in the offline analysis.
- Digital control script is same as last year, but it somehow works good with beam data.
- The next step is to create new digital control maps.
-- This has to be in the DBC database.
- Jaein will give a list of half entry created by 2024 data.
-- Intt7 has more than 10 chips with half entry issue.
- Try to take data with chip-by-chip control.
- Decoders (to make private DST or Joseph's decoder) take a long time to process digital controlled data. Maybe too much clone hits in it?
- Slide 3, this plots contains all tracks selected by Jaein's algorithm.
- Slide 4 should be a very peripheral event.
- Silicon seeding doesn't require TPC to make XYZ point.
- MVTX will check data with 111x111 with streaming, trigger.
[Takahiro]
- The plot shown here must be log scale, but need to double check.
- Ladder control works well.
- Channel 127 has some entry even after masked, which must be excluded in the offline analysis.
- Digital control script is same as last year, but it somehow works good with beam data.
- The next step is to create new digital control maps.
-- This has to be in the DBC database.
- Jaein will give a list of half entry created by 2024 data.
-- Intt7 has more than 10 chips with half entry issue.
- Try to take data with chip-by-chip control.
- Decoders (to make private DST or Joseph's decoder) take a long time to process digital controlled data. Maybe too much clone hits in it?
There are minutes attached to this event.
Show them.
