Transcript M. Ellis

Readout Electronics & DAQ
 Introduction
 DAQ Spreadsheet





VLSB_DAQ Sheet
Initialisation
Bias control
Cryo control
Data taking
 VLPC Cassette Characterisation



Process
G4MICE Applications
Status
 Station 5 Test


Software required
Hardware Required
 Conclusions
M. Ellis - MICE Tracker Meeting - 12th February 2007
1
Introduction
 Talk will discuss readout and analysis tools for Tracker 5 QA
and VLPC characterisation, this is not the same as firmware
and software for final MICE tracker!
 I should have spent this weekend writing documentation, so
have tried to use writing this talk as a start to documenting
the new software.
 Reminder about readout system:






One MICE cryostat with one or two cassettes.
Currently can read out 2 AFE IIt boards synchronised via a patch
cable and firmware from Kwame.
LED Pulser used for VLPC characterisation
Spreadsheet provides 3 triggering modes, one useful for LED and
cosmic data, the other for radioactive source data.
Data is produced in a plain text format, save a factor of 10 by
zipping.
G4MICE can read data and 2 applications currently written to
analyse it (3rd in preparation).
M. Ellis - MICE Tracker Meeting - 12th February 2007
2
DAQ Spreadsheet
 Excel spreadsheet with Visual Basic code to send commands to
the 1553 and VLSB modules via a BIT3 driver (Bob Angstadt,
FNAL).
 Most AFEIIT operations involve sending lists of commands
which are found on different tabs of the sheet.
 To make control easier for non-expert users, I’ve put all the
functions together on one sheet with some Visual Basic to set
the parameters automatically.
 VLSB_DAQ sheet can initialise up to 16 AFEIIt boards (FPGA
programming and TRIP register setup) as well as control VLPC
bias voltage, set temperature control and initiate temperature
monitoring as well as take data with the VLSB.
 Should not be too difficult to incorporate VB code to drive
the stage that will house the source for the station QA work.
M. Ellis - MICE Tracker Meeting - 12th February 2007
3
VLSB_DAQ Sheet
M. Ellis - MICE Tracker Meeting - 12th February 2007
4
Initialisation
 The process to initialise the boards starts by establishing
communication with the VME crate via the BIT3.
 Then the boards details are entered in the coloured region of
the sheet.
 Each board should have the correct board ID (written on a
sticker on the side of the board near the handle), the correct
RT (0 = LHB on LHC, 1 = RHB on RHC, 12 = LHB on RHC, 13
= RHB on RHC)
 The cassette number can be found on the cassette just above
where it emerges from the cryostat.
 The VLSB base address in hex should be entered and will be
converted to decimal by the spreadsheet.
 The last two booleans control whether or not the VLPC bias
should be turned on and whether the temperature control
should be initiated.
 At D0, with 2 boards, this process takes about 3 minutes and
15 seconds.
M. Ellis - MICE Tracker Meeting - 12th February 2007
5
Establish communication
with the VME crate via
the BIT3 interface
This should say “ok!”
Enter the number of boards to be controlled in 15G
Then starting on row 17, enter the Board ID, RT,
VLSB base address, cassette number and set TRUE
or FALSE for Bias and Cryo control
Press this button and all of the boards will be setup
including bias and cryo if requested
M. Ellis - MICE Tracker Meeting - 12th February 2007
6
Bias Control
 Once the correct voltage has been determined for each
cassette (see later in this talk), the spreadsheet will have
these values saved.
 The sheet also has the calibration information for every MICE
board (12 at the moment, information will be added as more
boards become available).
 In order to do bias scans, the cell I15 allows a global offset
to be selected by the user.
 For Station QA, this should be left at 0.0
 For bias scans, this is changed from 4.0 – 8.0 in 0.1V steps.
 The voltage can be set (or changed) by pressing the “Set Bias
Voltage Only” button.
 If the Set Bias? column is set to TRUE for each board and
there is no need to change the bias, then the “Init All
Boards” will set the voltages correctly, so there is no need to
use the “Set Bias Voltage Only” button.
M. Ellis - MICE Tracker Meeting - 12th February 2007
7
Set the offset to the bias voltage here.
For Tracker QA data taking this should be 0.0
For VLPC characterisation this is changed from
4.0 to 8.0 in 0.1 V steps
Enter TRUE to have the
bias set for the corresponding
board or FALSE for no bias
In order to set the voltage only
(assuming the boards have already
been initialised) or to change the
M. Ellis
- MICE
voltage,
press
thisTracker
buttonMeeting - 12th February 2007
8
Cryo Control
 In a similar manner as for the bias, there are sheets to setup
the VLPC cassette temperature control.
 A database of the known cassettes and AFEIIt boards allows
the R(9K) values and board calibrations to be converted into a
setpoint.
 As for the bias, a TRUE or FALSE can be entered to initiate
control on a board by board basis.
 Even if TRUE is set for a left hand board, nothing will be
done.
 There are four buttons that can be used to set the heaters
on a cassette to be off or on full. These are used for testing
the production cryostats.
 Clicking on the Monitor Temp button will change to the
monitoring sheet and start logging the cassette temperatures.
M. Ellis - MICE Tracker Meeting - 12th February 2007
9
This button will set temperature
control on the selected cassettes
It’s really only useful if you initialise
the boards without control and
then later want to start itThese four buttons will allow you to
select the heaters on a Left or Right
Cassette to be full on or off individually.
Enter TRUE to have the
This is really only useful for testing
temperature control switchedthe production cryostats
on for the cassette or FALSE
for no temperature control
This button will
change to the
Monitoring sheet
and start logging
the VLPC Cassette
temperatures
M. Ellis - MICE Tracker Meeting - 12th February 2007
10
Temperature Monitoring
 During normal operation of a cassette, we do not have
temperature monitoring running on the AFEIIt board as this
produces too high a load on the PIC for stable data taking.
 We can, however send 1553 commands to readout the current
temperature being monitored by the feedback loop on the
board as well as the heater current for each of the 8 modules
on a cassette.
 The TMonitor sheet will read the temperature and heater
currents for both cassettes and plot a graph of these versus
time.
 The recorded values are converted into a temperature in
Kelvin and the heater values into percentage of total power
(1.5W per cassette for all 8 heaters on).
 Note that for temperatures away from 9 Kelvin, the
calibration becomes less reliable.
 To stop monitoring, press the “End” key on the keyboard.
M. Ellis - MICE Tracker Meeting - 12th February 2007
11
Processed Values for the Left Cassette
Set Points for the Left Cassette
Right Cassette Temperatures are in Green
Heater Currents for the Left Cassette
Left Cassette Temperatures
in Red
0 = Off, 255 are
= Full
on
Mean Cassette Temperature for Left and Right Cassettes
Processed Values for the Right
Cassette
converted
Kelvin
RMS of Cassette Temperatures
for Left
and RighttoCassettes
% Power on Cassette Heaters
M. Ellis - MICE Tracker Meeting - 12th February 2007
12
Data Taking
 Once the AFE boards have been initialised and the bias
voltage and temperature control set, it is possible to take
data with the VLSBs.
 The output filename is selected in cell B16.
 The number of events is selected in cell B17.
 The trigger mode is set in cell G10:



0 = External trigger. This is used for LED and cosmic ray data.
1 = Software trigger. This is only used for debugging purposes.
2 = Self trigger. With the correct synchronisation cable, this will
cause the two boards to readout every time at least one of them
has a discriminator fire. This will be used for the Station QA.
 Once the “VLSB DAQ LOOP” button is pressed the number of
events requested will be taken.
 To stop the data taking early, press the “End” key.
M. Ellis - MICE Tracker Meeting - 12th February 2007
13
Output Filename including full path
Number of Events to Record
Press this button to start the DAQ
M. Ellis - MICE Tracker Meeting - 12th February 2007
14
Appearance During Data Taking
Number of Events Recorded
Number of Events Requested
M. Ellis - MICE Tracker Meeting - 12th February 2007
15
DAQ Spreadsheet
 Can be found on the web here:

http://www.mice.iit.edu/modules/tracker/VLSB/
 Current version is “a” (i.e. 10).
 Version 9 had a bug that was only discovered and fixed last
week that meant it was only reading out one of the two VLSBs
and writing that data twice.
 This has been fixed in version a.
M. Ellis - MICE Tracker Meeting - 12th February 2007
16
VLPC Cassette Characterisation
 In order to understand if Cassette 112 will be useful for
MICE we have taken data with the LED pulser and are in the
middle of analysing it.
 Data is taken at 41 voltage settings (4.0V – 8.0V in 0.1 V
steps) with the LED on and off.
 A G4MICE application (VlpcCassettes/ReadData) is used to
read this data and produce a ROOT file with one histogram
per channel.
 Another G4MICE application (VlpcCassettes/Calibration) reads
in a pair of ROOT files (LED on and off) for a given voltage
and fits the single or multi peak distributions and extracts the
pedestal, gain, mean number of photons and pedestal width.
 A third application (under development) will be used to produce
summary plots of gain, QE, etc as a function of bias and to
determine the number of dead and broken down channels as a
function of bias in order to optimise the bias on a module by
module basis.
M. Ellis - MICE Tracker Meeting - 12th February 2007
17
Peak Fitter
 The Calibration application uses the ROOT TSpectrum class to
find peaks in the histograms.
 The positions and heights of the peaks are used to seed a
special function that was developed by Yury Pogorel for fitting
the distributions in D0.
 Some example plots follow to show what the distributions and
fits look like for:



Low bias, voltage too low for the VLPCs to operate and only a
pedestal is seen in both LED on and off data.
High bias, voltage too high for correct operation and the VLPCs
are breaking down. This is sometimes seen in just the LED on
data, or when the voltage is even higher, both the LED on and
LED off data.
Normal operating voltage, where a multi peak spectrum can be
fitted.
M. Ellis - MICE Tracker Meeting - 12th February 2007
18
Cass 112 - Low Bias (4.0 V)
M. Ellis - MICE Tracker Meeting - 12th February 2007
19
Cass 112 – High Bias (8.0V)
M. Ellis - MICE Tracker Meeting - 12th February 2007
20
Cass 112 – High Bias (8.0V)
M. Ellis - MICE Tracker Meeting - 12th February 2007
21
Cass 105 – 6.0V
M. Ellis - MICE Tracker Meeting - 12th February 2007
22
Cass 112 – 6.0V
M. Ellis - MICE Tracker Meeting - 12th February 2007
23
Applications
 VlpcCassettes/ReadData:



./ReadData cards.in <input file1> <inputfile2> ... <inputfileN>
Produces a file adcs.root
Histogram ID = BBBMCC where BBB is the board number without
the leading digits 574, M is the MCM number (1-8) and CC is the
channel number (1-64).
 VlpcCassettes/Calibration:




./Calibration –ledon <ledon.root> -ledoff <ledoff.root> [-noplot]
[-bias <voltage>]
If –noplot is included then no plots will be shown.
If –bias and a number are passed then the output information will
include the voltage, otherwise it will be listed as 0.
I’m currently trying to get ROOT to write out a postscript file.
Once this works, the –ps option will cause this to happen.
M. Ellis - MICE Tracker Meeting - 12th February 2007
24
Demonstration
<< show demonstration here if time permits >>>
M. Ellis - MICE Tracker Meeting - 12th February 2007
25
Analysis
 Work is currently in
progress to determine a
set of criteria to
identify good and bad
channels.
 The operating voltage
will be set to maximise
the mean number of
photons (and thus the
QE) while minimising the
number of channels that
are unusable either due
to noise, no signal or
breakdown.
M. Ellis - MICE Tracker Meeting - 12th February 2007
26
Station 5 Test
 Requirements for Station 5 test:



DAQ Spreadsheet
Calibration information for chosen cassette and AFE boards
Cryostat, cassette, 2 AFEIIt boards, waveguides, station, gas
system, pumps, AFE power supply, VME crate, 1553, 2 VLSBs,
DG2020, etc...
 For the tracker QA this list is almost identical, except as we
will not be using an external trigger there will be no need for
the DG2020.
 The DAQ spreadsheet should now be ready for Station 5 and
the Station QA.
 I took some data on Friday that should confirm that the bug
with reading out multiple VLSBs is fixed and just need to get
a spare couple of minutes to analyse it.
 The following slides will indicate what else is still missing.
M. Ellis - MICE Tracker Meeting - 12th February 2007
27
Software Required

For whichever cassettes is put in the prototype cassette, we will
want to perform the MICE characterisation process on it.

The data taking for this will take one day and the analysis probably
a further day or two assuming nothing odd is found.

The third application for this analysis is not yet ready, but will be
written soon.

Once the correct bias values are determined, the spreadsheet will be
updated with these values in the database.

The only other missing piece of software will be whatever application
is required in G4MICE to perform the analysis of the data. This
should be a relatively quick job to write, copying from existing
applications and could serve as a good path into G4MICE for a new
user...
M. Ellis - MICE Tracker Meeting - 12th February 2007
28
Hardware Required
 The first production cryostat was cooled down last week and
we looked at the cassette temperatures on Friday.
 Russ is working on this one and in parallel the next one is
being finished.
 Over the coming few months, the existing AFE and cryo
infrastructure will be needed for the cryostat testing and
cassette characterisation.
 If the Station 5 test is to be conducted in the UK, then a
complete second set of hardware capable of operating on UK
mains will need to be setup.
 This is currently in progress, but not complete.
M. Ellis - MICE Tracker Meeting - 12th February 2007
29
Conclusions
 The software tools for the VLPC cassette characterisation are
almost complete.
 A first working version of the Excel DAQ spreadsheet capable
of reading out multiple VLSBs synchronised with a patch cable
and internally or externally triggered is now on the web.
 A first look at Cassette 112 indicates that it is producing
good data, however further analysis (of both halves!) will be
required to understand the performance of the different
types of VLPC in the cassette and decide whether or not it
will perform well enough to be used in MICE.
 There is still a fair bit of hardware that needs to be set up
and tested before it will be possible to run two VLPC systems
in parallel.
M. Ellis - MICE Tracker Meeting - 12th February 2007
30