Previous version can be found here. Updates in this version are shown in red, while unchanged responses are in blue. Backup documents are:
3879, 2310, 2535, 5777, 5331, 1808, 5256, 5246, 2633, 1343.6099
1. Looking at the layout the straws do not overlap
around the straw center. Can you provide plots from MC of the straw resolution
versus distance including electronics? (even if it is not exactly within
the scope to review the performance requirements).
Plots from simulation for time division resolution can
be found here. “V” is the coordinate along the
wire. Similarly, drift time resolution from simulation is presented here. “Reco drift radius” is drift distance based
on fitting the track over all hits, while “true drift radius” is drift distance
directly from the simulation.
The simulations assume the electron signal is 10x the rms noise and the
threshold set to 5x the rms noise.
a.
How do you
verify these with the prototypes?
An 8-straw prototype being tested at LBNL includes a
pixel detector. Although not it’s
primary function, this can also measure resolution. Initial results are
presented in document 6168. The noise level for these tests was higher than
expected and, not surprisingly, neither resolution (>200μm) nor
efficiency (<50%) meet our requirements. The test will be repeated in the
near future.
b.
Does the
MC include the discriminator threshold level achieved with the prototype
electronics?
Yes, but using thresholds achieved with bench tests. A
more complete understanding of noise and thresholds is in progress.
c.
What is
the gas gain and how has it been measured?
We expect to run ≲5·104. Gain is measured
using total current with Fe55 and using the known primary ions from
Fe55.
d.
Has the
noise level been measured?
Yes. With bench tests – straws attached, but no other
electronics present -- it agrees well with Spice simulation
e.
Can you
meet the required resolution and efficiency specs with present S/N?
Yes: With gas gain of 5·104 and assuming S/N is what we expect from
SPICE and measure in bench tests. We note we have not achieved this in a
realistic setup and there is a risk we will need to increase the gas gain.
2. Can you describe
in more detail the cooling system? Have you done any simulations or tests with
realistic geometry of heat loads (boards) and cooling interfaces? What is the
temperature of the coolant and electronics (overall temperature gradient)?
Documents 2068 and 2013 provide considerable
detail. Both simulation and bench tests have been done. See documents 4087 and
4085. Revisions since that time in panel design improve heat transfer (more
aluminum) and lower power. These have been analyzed but presented in a less
self-contained manner; see for example document 5208. The coolant will run ~25C
and gradient is ~3C.
3. Can you provide
details (drawings) of the vacuum sealing of the panels and how the assembly and
leak testing is carried out?
Hopefully we can go over this interactively with 3D
models.
4. What are the
characteristics of the fuses and have they been tested?
The fuse consists of a small torsion spring
soldered under tension to conduct HV. One side uses a low-temperature solder
which is melted using a heating resistor. It was tested extensively this summer
and the document for it will be uploaded soon.
5. How have you
reached the conclusion that the inside and outside of the straw do not have to
be connected?
No. Given the ambiguity of the conclusion, and the
difficulty of assuring there is no connection, we have (tentatively) decided to
make the connection.
6. Can you explain
in some more detail with a drawing or diagram the connections of the services
(HV, LV and gas) i.e., patch panels, position of connections and modularity?
We do not yet details of the layout. Planned
modularity is:
- One pair of LV lines per panel, 48V, with stepdown at the panel
- One pair gas lines per plane, with “fanout” outside the DS.
- We have left space for 960 HV lines inside the DS, fanned out from 192 HV
channels outside the DS. We have not yet settled whether we need so many or how
to distribute them if we do.
7. Can you describe
the plan in case of failure/problem with HV, LV, leaking straw and readout
problem?
The general short-term solution to any problem is
to disable the smallest possible segment of the detector and continue running.
Longer term the plane is removed and the panel replaced. The defective panel
can then be repaired while data taking continues. “Smallest possible segment”
is a panel for LV or readout; a single straw for (most) HV problems; a plane
for gas problems.
8. It would be good to produce a detailed circuit
diagram with characteristics and values of the components.
These exist, but perhaps should be coalesced into a
single document.
9. In the aging
test you have seen no change in gain up to 1C/cm. Have you seen any effects on
the straws?
Straws were cut open to examine for discoloration
or other visible damage. End to end resistance measurements were also done.
There was no indication of radiation-induced damage.
10. Can you outline the development plan towards the final electronics?
a.
What additional prototypes are planned?
The push right now is for 2-channel preamps. Work
on analog and digital motherboards is in progress. Prototypes for mother boards
will be produced by early CY16.
b.
What questions remain unanswered and when do they
think they will know enough to go into production?
Two vs single channel preamps, and details of the
preamp to straw connections, are major concerns. I expect them to be settled by
the end of this year. A lower probability, but higher impact, issue is
radiation damage. Tests will be performed at LBNL this fall (October).
c.
What risks do they see in the near
future?
The biggest risk is radiation sensitivity. In the
unlikely event it is a problem, it could require major changes.
d.
How much schedule and cash contingency
do they have in the baseline plan?
We cannot begin with production electronics till
CY17, which is adequate time. We have ~40% contingency. Sadly not cash, we have
to beg and plead to get it released.
Comments:
1.
It is
important to do the radiation test of the electronics.
2.
It is
recommended to repeat the discharge tests with the smaller blocking cap.
Comments:
This
section summarizes the comments and conclusions by the review committee after
the first two video meetings:
1.
It
is the review opinion that the project is in good shape in view of the present
schedule.
2.
The
review committee thinks it would be useful to establish a clear grounding and
shielding plan that is based on their present design status.
3.
A
simple drawing of how signal and HV currents flow might be instructive (and
also help define the grounding and shielding plan).
Some information regarding HV is
provided here. Low voltage is
described in document 6099. The baseline plan uses exclusively fiber optic
communication, so there is no grounding or current issues. We are looking at
copper and determining how to create a ground break will be part of that study.
4.
Interaction
at an early stage with the FNAL safety team in order to get the green light for
the planned electrical scheme and choice of components may avoid surprises and
problems in the future.
An electrical integration
team, headed by Karen Byrum and Gary Drake (both of Argonne), is working to
ensure the various systems have consistent grounding scheme. The design will
then be reviewed for safety.
5.
The
reviewers felt that the layout and choice of components should not be left
entirely to the DAQ group - the reliability problems with commercial
electro-optical converters (let alone radiation hardness issues) are not to be
ignored. The possibility to move the sensitive parts/components outside the detector
to a more accessible area with little cost, should be considered
We work closely with the DAQ
group as part of the ROC and integration tests. We are also looking at
non-optical communication to bypass problems with electro-optical converters.
6.
The
idea of using the CERN rad hard I/O equipment, as a fall back is plausible, but
expensive. However, this option also contains a reliability worry and it needs
to be examined in more detail, if considered as a real option.
7.
Concerning
the fuses, a document describing design and validation is required.
8.
It would be to useful to establish a plan
including the different tests and validations foreseen before CD3b e.g.,
electronics tests, radiations tests and a full system test.
Electronics
·
Test 2‑channel
preamps with cosmics using the test chamber at LBL. Nov ’15
·
Instrument
(including digitizers) eight straws on panel v1.5. Test with cosmics and, if
available, test beam. Dec 15
·
Finalize ROC
and digitizer. Oct-Dec 15
·
Test revised
ROC+Digitizer board. Jan 15
·
Neutron
damage (total dose) tests. Coordinate with test of plastics used for
construction (below).
·
Depending on
outcome of above, explore vacuum rated coax as an alternative to fiber. (If
electro‑optics looks ok this may be postponed till after CD3c).
We do not want to design
a full-size Analog Mother Board (AMB) which is not compatible with v2.5.
Depending on how fast v2.5 design is finalized, we may instrument all of panel
v1.5 before CD3c.
Full plane test is
post-CD3c.
Mechanical
The immediate push is to
test design changes from v1.5 to v2.5. See document 6225.
9.
A
link to the plots from the simulations (by Dave B) including momentum
resolution should be provided. The space resolution close to the wire is
usually not so good and there is a gap between the straws in the double layer
of ~1.3mm.
10. A lot of work has been done
on testing and validation. A link to beam test set-up, noise measurements, gain
measurements and results would be useful.
See documents 5777 for proton
beam tests. See 6168 for LBL’s cosmic ray setup and preliminary results, and
6265 for updated results.
