AHCAL and BIF Test Beam at DESY (May 2016)

We had 2 weeks of beamtime at DESY, TB22, from 02/05 to 14/05. Elog is available here : https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/18/04.05&pos=2016-05-05T10:55:00

Contents

Elog is available here : https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/18/04.05&pos=2016-05-05T10:55:00

DAQ : Monitoring improvement (DQM4HEP), EUDAQ improvements (BIF data, LED automatic scans) and speedups

A wiki for the new monitoring system is up here : http://flcwiki.desy.de/AHCALMonitoring_DQM4HEP

Here you can find a wiki about the EUDAQ for AHCAL and BIF (wiki construction is in progress).

Electronic Setup: AHCAL + BIF

The setup is very similar to the one used in previous testbeams. See for example http://flcweb02.desy.de/flcwiki/AHCALTestBeamCERN2015/#Electronic_setup_and_cabling:_diagrams_and_pictures . I the present case, we only use two scintillators (placed in front of the detector) and we also used the BIF.

We are using two scintillators placed between the beam and the detector as validation signal. Both scintillators are thin sticks of about 1cm width and 15 cm lenght and both are crossed in order to set horizontal and vertical position with high precision. The coincidence signal is fed into the CCC to be used as validation signal.

The two signals from the scintillators and from the coincidence is conducted to the BIF also.

The coincidence signal is also fed into the two t0 channels, in the two setups setups with big layers (TDC calibrations, MIP calibration of new HBU4).

Setup 1 (4th - 9th of May): TDC calibrations using Electron Showers for 4 big Modules setup

biglayer_airstack.jpg whiteboard_map_biglayers.jpg

The mapping file is available here

Using the same configuration files from the CERN testbeam, we put the big layers in a airstack configuration (left orientation). We connected all T0s/Cherenkows (6 in total, as the ones on Module 13 are gone). And the BIF is also added in the setup, this will enable us to perform a cross-check of the T0s as reference time in the CERN testbeam and the BIF time as absolute reference. We can also perform a TDC calibration using the BIF time (cross-check with parameters extracted from SPS data). To perform the calibration, an absorber in Aluminium (3 X0) is put in front of the stack.

Modules:

4 big layers (72x72 cm^2) with individually wrapped tiles : shower profile, correlation of hit times.

Phys.
order

Module#

power
cable

HDMI
cable

MPOD

LDA
port#

name

1.

12

13

12

u204,u304,u504

10

"Ketek"

2.

13

14

13

u205,u305,u505

11

"Ketek new"

3.

14

15

14

u206,u306,u506

12

"Sensl"

4.

15

16

15

u207,u307,u507

13

"Sensl"

T0s channels

Same used as in the Testbeam at CERN.

List of T0 channels

Name

Module#

ChipId

Channel

Comments

Ketek

12

169

29

-

Ketek

12

177

23

-

SenSL

14

201

29

noisy in few runs

SenSL

14

211

6

broken

SenSL

15

217

23

-

SenSL

15

227

6

-

Absorber block

We place the absorber in front of the detector. The absorber is a block of aluminium of a length of 3 X0 (see picture).

absorber.jpg

Run List

More detail are available here https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/18/05.05&pos=2016-05-05T16:53:00

Runs

Comment

41000-41149

Test Runs, Wrong validation signal - Do not use!

41150-41192

Stage Position 5869.6 H - 285.5 V (middle), Trigger per run ~38000

41193-41307

Stage Position 5735 H - 270 V (middle), Overnight run, Trigger rates ~45500

41308-41370

Stage Position 5915 H - 90 V (top-right)

41371-41459

Stage Position 5915 H - 450 V (top-left), Overnight run

41461-41537

Stage Position 5555 H - 450 V (bottom-left)

41538-41641

Stage Position 5555 H - 90 V (bottom-right), Overnight run

41642-41712

Stage Position 5555 H - 90 V (bottom-right), /!\ New DIF firmware (faster) - offset might be different /!\

Due to a bug in the EUDAQ AHCAL and BIF Producer or crashes of the BIF software, some runs have a de-synchronization of the BIF ROC and EUDAQ ROC (shifted by one ROC from the start of the error - errors are into the log of the EUDAQ). After counting the total number of events and rejecting runs with problems, we have still around 2 Mio ROC valid (eq. 16 Mio events minimum). A full table can be found here.

Reconstruction using EUDAQ Files (includes BIF timing)

The latest Calice software (version 4664) includes the new EUDAQ event builder (EUDAQEventBuilder2016 Processor) for the new particular format of the data. The EUDAQ Data format is now :

i:CycleNr:i:BunchXID;i:EvtNr;i:ChipID;i:NChannels:i:TDC14bit[NC];i:ADC14bit[NC]. The Gain/Hit Bit are encoded in the TDC/ADC (both are available now) and can be retrieved by doing

//Get HitBit
int HitBit = (ADC[ichan]& 0x1000)?1:0;
//Get GainBit
int GainBit = (ADC[ichan]& 0x2000)?1:0;

Also the BIF Data is available in the ROC of the EUDAQ now. The format is always a sequence of 4 unsigned ints: Type | EventCnt | TS_Low | TS_High.

More details are available here : http://flcwiki.desy.de/HCalMeetings2016?action=AttachFile&do=view&target=20160517_JK.pdf Let's see an example :

collection name : EUDAQDataBIF
parameters:

--------------- print out of LCGenericObject collection ---------------

flag:  0x0
parameter DataDescription [string]: i:Type:i:EventCnt:i:TS_Low:i:TS_High,
parameter Timestamp [string]: Wed, 11 May 2016 13:14:54 +0200,
parameter TypeName [string]: CaliceObject,

[   id   ] i:Type:i:EventCnt:i:TS_Low:i:TS_High - isFixedSize: false
--------------------------------------------------------

[00000731] i:50331648; i:449; i:91639648; i:200578; i:16973824; i:49744725; i:91645249; i:200578; i:16973824; i:49744726; i:92107790; i:200578; i:16973824; i:49744727; i:92204049; i:200578; i:16973824; i:49744728; i:92524077; i:200578; i:16973824; i:49744729; i:92591582; i:200578; i:16973824; i:49744730; i:93075374; i:200578; i:16973824; i:49744731; i:93097880; i:200578; i:16973824; i:49744732; i:93225387; i:200578; i:16973824; i:49744733; i:93274140; i:200578; i:16973824; i:49744734; i:93315393; i:200578; i:16973824; i:49744735; i:93359158; i:200578; i:33554432; i:449; i:91639648; i:200578;  -----
---------------------------------------------------

The format is always the following but can be of variable sizes (LCGenericObject as a vector<int>) depending on the number of BIF triggers. The 4 first numbers are : Start | CycleNr | TS_low | TS_high . They correspond to the Time of the start Acquisition and should always be there.

50331648 is 0x03000000 which is StartAcq and to get the TimeStamp value one as to put together TS_low and TS_high as the following : 91639648 is 0x05764F60, 200578 is 0x00030F82 - To combine we have 0x00030F82|05764F60 which gives 861476041936736 (uint64_t).

The 4 last numbers are : Stop | CycleNr | TS_low | TS_high . They correspond to the Time of the stop Acquisition and should always be there.

Using the same method, this gives 3554432 is 0x02000000 which is StartAcq and the !Timestamp here is the same due to a bug (corrected now). This give us the full time when the detector was open during a ROC.

Then the rest is the BIF triggers with Trigger Type | TriggerNb | TS_low | TS_high always repeating. 16973824 is 0x01030000 indicates that it is an external trigger (0x01030000) and on the 4th lemo connector (0x01030000) of the mini-TLU. The timestamp of the trigger is relative to the start of the acquisition. The format of the data in that way is a bit complicated and during the event building, a new format is created and BXID correlation of the BIF Triggers with the data is done, at the end the new format look like this :

collection name : BIFData
parameters:

--------------- print out of LCGenericObject collection ---------------

flag:  0x80000000
parameter DataDescription [string]: i:InputSource; i:BXID:f:Time,
parameter Start_Acquisition [string]: 336809120124544,
parameter Stop_Acquisition [string]: 336809120124544,
parameter TypeName [string]: BIFBlock,

[   id   ] i:InputSource; i:BXID:f:Time - isFixedSize: true
--------------------------------------------------------
[00062768] i:3; i:62; f:2.652e+03;  --------------------------------------------------------

The new format keeps the input source (can be modified in the future for indicating Scint/Cherenkow/Other triggers), the BXID of the BIF trigger and the absolute time of the BIF Trigger in ns. The BXID is obtained by : ((TSparticle - TSstart) - BIF OFFSET)/5120. The Time of the Trigger by : ((TSparticle - TSstart) - BIF OFFSET)%5120 * 0.78125 [ns]

Each time, there is a new configuration (especially DIF firmware, power-pulsing mode, ILC/TB mode, Beam instrumentation cabling), the BIF OFFSET has to be calibrated. Here is a table with the BIF OFFSET depending of the type of runs we took :

Type

Offset

Electron showers (Big layers)

16870

MIP (new DIF) with POD reset

68080

PowerPulsing (Ton = 100 / 20 us)

93610

A procedure for calibrating the offset is being developed and soon to be available in the calice_calib package.

Calibration of BIF Offset

A new processor is available in the calice_calib package. Ahc2BifOffsetCalibrator enables to take a single run in lcio and scan the bif offset in order to get the best value. In the steering parameters, you have to specify the start, stop and step of the scan (e.g. 0, 100000, 1).

As a benchmark, it takes about 6 mins to scan 100000 points on a single run (Run 41150). The output is a rootfile containing the graph of the scan with on the x axis the bif offset and on the y axis the number of correlated events. What you want to look for is when the slope suddenly changes (here around 16870 / second highest peak). The maximum of the curve is not the interesting point because of miss-assignments of BXIDs thus increasing the number of correlated events.

BifOffset_scan.jpg

After all the reconstruction, the bif data can be added into the rootfile (using the new BifWriteEngine) and then correlation plots can be done.

BIFTDC_Reco.jpg

Setup 2 (11th May & 13th May): MIP Calibration of the new HBU4

Setup

We disassemble one of the big layer (Module 15 - SenSL) and we used the last HBUs (HBU3_5 & HBU3_12) in a single layer configuration. We put the layers in the small airstack with the new Interface Boards. We measured the HV (adjusted by software) for getting close as possible to the original voltage (32.1 V). HBU3_5 : 32.11 V & HBU3_12 : 32.09 V.

The Layers are position in Top readout but with the tiles faced to the beam. The mapping file is available here . For reconstruction a temporary database is available in /cd_calice_Ahc2/TestbeamMay2016/

From 11th of May and 13th of May

Phys.
order

Module#

power
cable

HDMI
cable

MPOD

LDA
port#

name

1.

HBU3_5

1

1

u0,u100,u400

0

"SenSL"

2.

HBU3_12

2

2

u1,u101,u401

1

"SenSL"

3.

SM_100

3

3

u2,u102,u402

2

"SMD MPPC"

After reconstruction :

MIP.png

Note : 13th of May the MIP Scan was performed in powerpulsing mode (212 us time on) and with a PA of 600 fF for the SM_100

Run List

11th of May https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/19/10.05&pos=2016-05-10T18:27:49 Was performed with a PA of 650 fF

Runs

Comment

41758-41766

MIP scan chips 223, 231, 239

41767-41775

MIP scan chips 224, 232, 240

41776-41788

MIP scan chips 222, 230, 238

41789-41798

MIP scan chips 221, 229, 237

41799-41901

high statistics MIP in the center x=253.9, y=5752.2 / Overnight run

13th of May https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/19/13.05&pos=2016-05-13T09:25:07

/!\ Was performed in PowerPulsing mode (212 us Ton) and with a PA of 600 fF for the SM_100. This is enough to not see any PowerPulsing effects /!\

Runs

Comment

42789 - 42798

MIP scan at 600fF for Mainz board, chips 221, 229, 237

42799 - 42807

MIP scan at 600fF for Mainz board, chips 223, 231, 239

42808 - 42814

Automatic running (20000 cycles) x=105.0, y=5782.0

42815 - 42823

MIP scan at 600fF for Mainz board, chips 224, 232, 240

42824 - 42832

MIP scan at 600fF for Mainz board, chips 222, 230, 238

Setup 3 (11th - 12th May): Power-pulsing

whiteboard_map_singlemodules.jpg

/!\ One has to be careful as the first PowerPulsing test where actually without due to temperature readings switching off/on chip /!\ https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/19/11.05&pos=2016-05-11T18:06:15

Run List

More detail are available here https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/18/05.05&pos=2016-05-05T16:53:00

Runs

Comment

41911 - 41945

Reference runs /!\ no powerpulsing /!\

41954 - 41957

Ton:100 (Labview setting) = 21us

41959 - 41963

Ton:500 (Labview setting) = 106us

41964 - 41965

Tests with ILC bunch structure 1 ms data / 199 ms idle

41966 - 42044

Ton:100 (Labview setting) = 21.6us / Overnight run

42045 - 42538

Bad runs - EUDAQ/BIF crashed

42539 - 42547

Ton=100 (20 us). Position: X=253.9, y=5752.2

42548 - 42551

Spill tests

42552 - 42557

Ton=100 (20 us). Position: X=253.9, y=5752.2

42558 - 42559

Ton=100 (20 us). Position: X=165, y=5752.2.

42620 - 42646

Ton=100 (20 us). Position: X=165, y=5752.2.

42647 - 42676

Ton 21.6us. Position: y 5752, x 344

42677 - 42788

Overnight run same position, same settings / 42736-42741 no beam

Setup 4 (14th to 16th May): TDC Calibration in HG/TDC and Intercalibration HG/LG with absorber

https://ttfinfo.desy.de/CALICEelog-sec/show.jsp?dir=/2016/19/13.05&pos=2016-05-13T18:59:41

The running was the following :

/!\ Was performed also in PowerPulsing mode with Ton= 212 us /!\

Run List

Runs

Comment

42833 - 43321

AT_IC Runs

43322 - 43458

AT Runs with TDC

Data location (dCache)

The data is uploaded on the dCache at the usual space, the raw data (binary as well as slcio and txt files) :

Theme

To feed the shifters urges of artistic beautiness and compensate the lack of natural light in the hut, a set of wallpapers was prepared and loaded in the PC's.

For this reason, Miro's paintings and pictures taken at the interior and exterior of the Miro's Fundation in Barcelona (in a sunny day...) accompanied us during the testbeam installation, debugging and data taking phases.

AHCALandBIF_TestBeamDESYMay2016 (last edited 2016-10-07 13:37:02 by AmbraProvenza)