1 / 10

TRD Raw Data Simulation with Zero Suppression and MCM simulation

TRD Raw Data Simulation with Zero Suppression and MCM simulation. Oct. 9, 2007 Offline Week @ CERN Ken Oyama, Heidelberg. TRD FEE+DAQ Overview. 200 MB/s/SM. Multi Chip Module(MCM) with 18 ch. preamplifiers 21 ch. ADCs digital filters tracking processor. GTU (Global Tracking Unit). DAQ.

tiger-nolan
Download Presentation

TRD Raw Data Simulation with Zero Suppression and MCM simulation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. TRD Raw Data Simulation with Zero Suppression and MCM simulation Oct. 9, 2007 Offline Week @ CERN Ken Oyama, Heidelberg

  2. TRD FEE+DAQ Overview 200 MB/s/SM Multi Chip Module(MCM) with 18 ch. preamplifiers 21 ch. ADCs digital filters tracking processor GTU (Global Tracking Unit) DAQ TRD Chamber

  3. Data Flow on the Real System • So far: only PASA+ADC were simulated in the AliRoot  kept as digits. • We need simulation of tracking, digital filter, and zero suppression. digital filter, online tracking, & zero suppression MCM Not Simulated

  4. New AliTRDmcm Class Structure (updated) Raw signal (sdigits) AliTRDmcmSim • New class: AliTRDmcmSim • Interface to mcmSim is AliTRDrawData. • Filter part is using Bogdan’s version. Clemens is checking and adjusting so that it will be same as TRAP. Remapping 21 ch ADC data Digital-Filters MUX filtered data Tracklet Calc. ZS-Clustering Tracklet (trigger) ZS-mapping Prod. Trk stream Prod. Raw stream ZS-map GTU sim. DDL (trg) DDL (raw)

  5. ZSMapping in AliTRDmcmSim for RawVer=V3 Algorithm similar to the one implemented in FEE. • A lookup table defines hits with three conditions: cond1: peak Qi > Qi-1 AND Qi > Qi+1 cond2: cluster size Qi-1 + Qi + Qi+1 > Threshold cond3: absolute size Qi > Threshold + a flag to add neighboring ADC channel (neighbor sensitivity). • Combination of conditions are programmable by LUT (default: only cond3). • Full time-bin of ADC with at least one hit will be read out. 29 MCM Header +10 % @ 100 % occupancy - 60 % @ 25 % occupancy time bin I.R. MAP Channel Mask (21 bits) ADC06 data ADC07 data ADC08 data 0 0 20 ADC channel

  6. Development Status • Version 3 (test version) raw data read/write with zero suppression completed. • A new class “AliTRDmcmSim” completed with: • Filter() … apply digital filters: FilterTail(), FilterGain(), FilterPedestal(). • ZSmapping() … apply zero suppression mapping with same way as in TRAP. • ProduceRawStream() … Produce MCM raw data. • Reading part was done by Mateusz … AliTRDrawStreamV2.

  7. # of entries (a.u.) ADC value Data Integrity • Digits integrity and mapping have been fully tested by comparing before and after zero suppression algorithm and full simulation chain including date.root integrity with help of Mateusz.

  8. Z.S. Performance • Tested with standard p+p (PYTHIA) & Pb+Pb (HIJING std. b=0~10 fm) bench. • Single indicator threshold scanned from ADC>1 to ADC>7. with neighbor sensitivity w/o neighbor sensitivity

  9. Compression Efficiency (cont.) • Black event size: 42.5 MB for full super-module. • p+p event size example [kB] (%): threshold neighbor sensitive neighbor not sensitive >1 43,264 (99.5) 37,304 (85.8) >2 33,688 (77.5) 17,580 (40.4) >3 7,792 (17.9) 3,180 (7.31) >4 1,168 (2.69) 732 (1.68) >5 580 (1.33) 580 (1.33) >6 580 (1.33) 580 (1.33) >7 580 (1.33) 580 (1.33) • 580 kB (32 kB/SM) is the theoretical limit due to header words in data. • DDL: 200 MB/s with assuming 100 kB/Sm/ev  2 kHz event rate @ p+p achievable GTU to DAQ.

  10. Future Work • Large differences from real so far are: 1. tracklet is not simulated yet. Code exists for trigger study (Bogdan). 2. tail cancellation simulation should be reviewed and closer to real TRAP (will be done by Clemens). New class AliTRDtrapAlu (programmable precision arithmetic logic unit) proposed by Clemens for the first step of filter simulation study. • Careful parameter adjustment is necessary with realistic noise and cluster finding efficiency study. • GTU and trigger simulation are being studied (AliTRDgtuSim by Minjung).

More Related