Level-0 Trigger

Overview

The BGV trigger consists of scintillator plates with dimensions 30x30cm2, arranged in 3 stations. A pair of scintillators is located before the gas target and is used to veto interactions occurring upstream of the target chamber. A second trigger plane is placed downstream of the scintillating fiber (SciFi) stations and provides the trigger signal. A third trigger plane has recently been installed and will be commissioned during the LHC 2017 run. It will be used in coincidence with the second plane to reject background from the other beam and confirm a crossing particle produced in the gas volume by the correct beam.

Position of the L0 trigger scintillators around the BGV experiement :

Drawings of the scintillator plate and the light guide:

• Drawing2_scintillator.pdf
• Drawing1_Light_guide.pdf

The PMT used is one from Hamamatsu : http://www.hamamatsu.com/us/en/product/category/3100/3002/H1949-51/index.html

Photos of a scintillator in the test lab:

With this setup, it is possible to measure signal from cosmic particles, from the LED and to test the cards (high voltage and acquisition).

Measurement of the length of the cables in the tunnel :

• L0CableLength.xlsx

Tunnel installation

Network devices information in the tunnel :

• cfv-ua43-bgv : The control card of the vme crate
• cfvm-ua43-bgv : The vme crate
• Information on the configuration of the fan tray of the crate can be found here
• More general information on front-ends can be found here, also mentioning the process for remote reset.
• To login to these devices one needs to be a member of dscdev unix group

Cables connections :

Cable Id	Channel	Scintillator (2016 Run)	Scintillator (2017 Run)
1425 362	Channel 3	Signal Bot	Confirm Top
1425 363	Channel 4	Signal Top	Signal Top
1425 431	Channel 1	Veto Bot	Veto Bot
1425 432	Channel 2	Veto Top	Veto Top
1425 355	Channel 3 HV	Signal Bot	Confirm Top
1425 356	Channel 4 HV	Signal Top	Signal Top
1425 358	Channel 1 HV	Veto Bot	Veto Bot
1425 359	Channel 2 HV	Veto Top	Veto Top
1425 434	LED	Veto Top	?
1425 435	LED	Veto Bot	?
1425 365	LED	Signal Top	?
1425 366	LED	Signal Bot	?

The length of each cable path can be seen below. Based on the 2016 setup we have a delay of 26ns, therefore L0 Latency setting of 115 on ODIN should be OK (to be verified).

VME Crate organisation :

• Slot 1 : CPU/control card : cfv-ua43-bgv
• Slot 7 : BRAN-B card (acquisition)
• Slot 9 : High-Voltage card
• Slot 12 : BOBR card

CAEN Modules

Two VME modules from CAEN were purchased in order to better control the trigger. These were the V812 constant fraction discriminator (CFD) and the V2495 programmable logic board.

Information on the firmware of the V2495

• The documentation can be found here.
• In order to install the firmware to the board a Windows/Linux machine is needed and a USB-miniUSB cable to connect to the front panel of the V2495
• [Linux] - Install CAENVMELib, CAENComm
  • for an ubuntu14 machine also openjdk-8-jre [jdk] and a 64bit version of fdt2xx was also needed)
• Then install CAENUpgrader utility
  • Then from the CAENUpdater select the following:
    • Action->Upgrade Firmware, Model->V2495, choose appropriate .rpd file / Connection type-> leave this as is - no VME Base address / Flash Memory-> Application 5, Skip Verify ticked
    • Upgrade - wait few minutes - should be OK !

Setting up the boards

For these examples we are assuming base addresses for V812 as 0xeeff0000, and for V2495 as 0x32100000 - as it is currently the case.

Setting up V812

Register	Address	Value	Comment
Threshold Ch.0	0xEEFF0000	0xC8	Set threshold of ch0 (-5->-255 mV)
...	...	...	...
Threshold Ch.15	0xEEFF001E	0xC8	Set threshold of ch15 (-5->-255 mV)
Output pulse width	0xEEFF0040	0x64	Pulse width for ch0-7 (~20ns)
Output pulse width	0xEEFF0042	0x64	Pulse width for ch8-15 (~20ns)
Dead Time	0xEEFF0044	0x0	Dead time for ch0-7 (150ns)
Dead Time	0xEEFF0046	0x0	Dead time for ch8-15 (150ns)
Pattern Inhibit register	0xEEFF004A	0xFFFF	Enable (all) channels

Setting up V2495

Register	Address	Value	Comment
Input configuration Ch.0	0x32101800	0x0	Set threshold of ch0 (Enabled - not inverted)
...	...	...	...
Input configuration Ch.5	0x32101814	0x0	Set threshold of ch5 (Enabled - not inverted)
Control Register	0x3210181C	0x0	External LHC clock / Mon7=Coincidence / NIM

Data Acquisition

The data is stored on EOS, in the folder : /eos/bgv/data/L0Trigger

Two types of files are created : TriggerData _(unixTime).dat and BunchData _(unixTime).dat

The unixTime stamp tells from when the data starts in this file. The files are text files and are organised like this :

In order to monitor the data, the following script is present in the folder /afs/cern.ch/project/lhcbgv/L0Trig/scripts :

• FullAnalysis _EOS.py : Read the TriggerData and the BunchData for a given period. Create many plots and can save them if asked.

The script needs to be run from lxplus, there are 3 different ways to use it :

• python FullAnalysis_EOS.py -b <start time> -e <stop time>
• python FullAnalysis_EOS.py -b <start time> -l <time length [sec]>
• python FullAnalysis_EOS.py -f <fill number>

The arguments <start time> and <stop time> can be either unix : 1436190224

or UTC time in the format : "16/07/15 15:25"

Results example :

Here I show the results of a fill of 298 bunchs at 6.5 TeV. The high voltage of the PMT are all 1400V, the thresholds of the downstream scintillators are 200, for the upstreams it is 50. The logic is such that we create trigger signal only when signal is detected in both downstreams and nothing in the upstreams. This should correspond to interactions happening only in the gas tank.

To get the next plot, I used the command : python FullAnalysis_EOS.py -b "15/07/15 23:00" -l 30000

On the left plot, we see that the higher rate for the the Downstream scintillitor, which agree with the fact that they have the smallest thresholds. The difference between the two can be due to slightly different gain. The "Veto" rate corresponds to the upstream scintillators in coincidence.

Next are the two Downstream scintillators which have bigger threshold and then "Hit" rate corresponds the downstreams in coincidence.

Finallly, the trigger rate is the one defined by the logic. It is around 400 Hz but without gas injected. With gas injected, the pressure should increase by a factor 10^3, so we can expect the rate to be around 400 kHz.

The script gives also as results the settings of the trigger system during the selected period of time.

All the plots and the trigger settings for all files can be found in the folder /afs/cern.ch/project/lhcbgv/L0Trig/Plots

The control panel :

This panel gives a live view of the results and control the acquisition card and the High-Voltage card.

One can access this panel with the following procedure:

• ssh -X <username>@cs-ccr-dev1
• /afs/cern.ch/project/l/lhcbgv/L0Trigger/Control_Panel

Other info

Notes from meetings:

• 2014/06/26

LHCb Documents:

• Specification of the L0DU - ODIN link (R. Jacobsson)