Transcript HF+ phi3 timing

ROBOX with Multi Anode PMT’s
Status and plans
A. Mestvirishvili, I. Schmidt, S. Sen, F. Ozok
University of Iowa
ROBOX
• ROBOX with 24 new m-anode PMT’s assembled and being
tested in B904 test facility
• Box is operational, test runs were taken – SPE vs HV, LED vs
HV, phase scan. Analysis is in progress
• Initially planned to be installed during January this year, just
before HF leaves garage
• Postponed due to observed problem with baseboard.
Particularly with the PMT socket
• New schedule for installation mid February
ROB with new adapter boards and recessed connectors
3
New Base board
Baseboard with adapter board
Baseboard – view from socket side.
Sockets are custom designed at FNAL
Baseboard with PMT’s
R7600 LED vs Channels - 2012
550 V
600 V
8 channels are missing on
the plot for HV=700. This is
due either reconstruction
program can not treat
correctly
saturated
channels, or not proper run
was used to produce this
plot
650 V
700 V
Slide borrowed from Sercan Sen’s presentation. Full version can be found
5
https://indico.cern.ch/conferenceDisplay.py?confId=174973
LED Signal for 700 Volt
Reflections seen here
appears to be feature of
the test stand and QIE.
We do not have light
guides and light injection
path is slightly different for
a test stand.
When
PMT
produces
large signal, more 10mA
QIE input circuit is not
correctly
impedance
matched to cable and we
get reflection
600Volt
750Volt
R7600 LED vs HVs - 2012
Chan
Chan
Chan
Chan
01
06
13
22
1. Channel
6. Channel
7. Channel
12. Channel
13. Channel
14. Channel
15. Channel
16. Channel
17. Channel
18. Channel
19. Channel
20. Channel
21. Channel
22. Channel
23. Channel
24. Channel
6.79401e+00
6.57143e+00
6.93455e+00
6.48874e+00
6.95571e+00
6.49086e+00
6.79958e+00
6.82936e+00
6.99471e+00
6.68241e+00
6.61797e+00
7.16156e+00
6.17717e+00
6.67851e+00
7.14078e+00
6.47633e+00
Slide borrowed from Sercan Sen’s presentation. Full version can be found
8
https://indico.cern.ch/conferenceDisplay.py?confId=174973
Last SPE Runs, Channel 13-24
Slide borrowed from Sercan Sen’s presentation. Full version can be found
9
https://indico.cern.ch/conferenceDisplay.py?confId=174973
Problems we hit during testing
Two from three base boards, which were shipped from FNAL in December were
not working properly.
Cathode Voltage – 600V
Voltagesr for last two dinodes – 95 and 47 V
Power supply was drawing too much current
and tripping finally with over current protection
when on a last dinodes nominal values were
set.
When power for last two dinodes were off,
divider was setting different values for those
two dinodes, then nominal
Measuring the resistances of divider chain
showed that last two dinode resistor values
were different than nominal.
PMT Sockets
Designed at FNAL
Pics from Sergey Los
Pins are press fitted to PCB and probably
this cracks the capacitors, which then
alerts the resistor values in a divider chain
and, as consequence changes division
coefficient
https://indico.cern.ch/conferenceDisplay.py?confId=174973
Problems we hit during testing - 2
1. New boards, which underwent more strict test were shipped from FNAL. All
three boards exchanged.
2. During the test one board failed again with exactly the same signature, as
previous two.
•Power supply for last dinode tripped with over current protection
•Operating the board without external voltages for last two dinodes resulted
voltages set by divider to be 88 and 33 Volt instead of 95 and 47 for 600
Volt for cathode
• Board was exchanged. ROBOX is operational now with new board
installed
Two different solutions
• First press fit the pins, then solder caps
Difficulties due to space unavailability
Second solution – remove caps from
socket and move them to baseboard
Difficulties - (according to Sergey Los)
“several extra centimeters in dinode
connection which is at the edge of 1ns
rise time”
Summary
• After a faulty baseboards were exchanged ROBOX is operational.
Different tests were done – SPE, LED, Phase scan etc.
• Analysis is in progress, but from those results which are already obtained
and provided baseboards are in normal working condition, everything
looks good
• ROBOX is ready to be installed on detector. Installation postponed for four
weeks, until we completely eliminate reason of failure
• Two different solutions – either first press fit the pins and then solder
caps, or just move caps from socket to baseboard
• In case of baseboard last or last two dinode caps fails it can be operated
without power for last two dinode, but this is not perfect solution since
extra noise can be introduced due to current leakage, gain change,
response linearity will be alerted.