Powered by Indico
v3.2.5
Peter Kluit is inviting you to a scheduled Zoom meeting.
Topic: Lepton Collider Zoom Meeting
Time: Sep 2, 2024 10:30 AM Amsterdam
Join Zoom Meeting
https://nikhef.zoom.us/j/96374549010?pwd=Q56MOmSAIukEUaybCILsYMJhXVozOl.1
Meeting ID: 963 7454 9010
Passcode: 752349
Kees Ligtenberg and Fred Hartjes contributed to the 24 chip module and the DESY testbeam
eviewers' comments:
Reviewer #1: Please find minor line-by-line comments below:
50: readout -> read out
50: SPIDR -> "SPIDR"
61: In this paper,
Table 1: was water content measured or determined from drift velocity measurements are described below?
232, 241: Please give values for the errors even if they are small.
279: Specify which hits were excluded - radius around pillars?
289: Due to the presence of the dike,
290: Therefore,
312: Because of limited statistics,
340: bias -> basis
349: Telescope -> telescope
363: place reference at end behind number to improve readability
367: Time Projection Chamber -> TPC - already defined above
Reviewer #2: The manuscript entitled "Towards a Pixel TPC part I: construction and test of a 32 chip GridPix detector" reports on the construction and first test beam results of a Time Projection Chamber (TPC) read out by 4 x 8 GridPix chips, 256 x 256 pixels each, and a maximum drift length of 40mm. Building upon previous detectors employing a single or four GridPix chips, the results presented in the manuscript are significantly new and important, a priori justifiying publication in Nucl. Instr. Meth. Before doing so, however, a number of issues, related to the presentation and the analysis should be addressed.
Major comments:
a) The manuscript is a bit sloppy about the use of the terms "resolution and residual". I insist that this is corrected. The resolution is related to the intrinsic performance of the detector. It does not include external uncertainties like the uncertainty of the reference track. The residual is the difference (measured - expected) hit positions for a single track/plane. The residual distribution is the distribution of residuals over many tracks. Being a distribution, it has a mean and a width. In this sense, e.g. Eq. (2) gives an expression for the width of the residual distribution, not the resolution. Figure 5 shows the measured width of the residual distribution (could be called residual width, if defined like that). Figure 7 shows the mean of the residual distribution (could be called residual mean, if defined). Similar examples can be found all along the manuscript. In addition, it is not explained in unambiguous terms, how exactly the residuals in xy and z are calculated. For every hit, one normally calculates the 3D-vector of the closest distance to the track. This vector projected onto the xy plane gives the residual in xy, its projection onto the z axis gives the residual in z. What value is used for the z coordinate (hit z position or track z position at DCA)? Please explain in the text. In Section 5.4, the authors talk about "tracking precision in the middle of the TPC". Please clearly define the term and explain how the numbers were obtained. Uncertainties (or rms values of the corresponding distributions) should be given for the numbers quoted.
Major comments:
a) The manuscript is a bit sloppy about the use of the terms "resolution and residual". I insist that this is corrected. The resolution is related to the intrinsic performance of the detector. It does not include external uncertainties like the uncertainty of the reference track. The residual is the difference (measured - expected) hit positions for a single track/plane. The residual distribution is the distribution of residuals over many tracks. Being a distribution, it has a mean and a width. In this sense, e.g. Eq. (2) gives an expression for the width of the residual distribution, not the resolution. Figure 5 shows the measured width of the residual distribution (could be called residual width, if defined like that). Figure 7 shows the mean of the residual distribution (could be called residual mean, if defined). Similar examples can be found all along the manuscript. In addition, it is not explained in unambiguous terms, how exactly the residuals in xy and z are calculated. For every hit, one normally calculates the 3D-vector of the closest distance to the track. This vector projected onto the xy plane gives the residual in xy, its projection onto the z axis gives the residual in z. What value is used for the z coordinate (hit z position or track z position at DCA)? Please explain in the text. In Section 5.4, the authors talk about "tracking precision in the middle of the TPC". Please clearly define the term and explain how the numbers were obtained. Uncertainties (or rms values of the corresponding distributions) should be given for the numbers quoted.
b) It is somewhat surprising that the statistical uncertainties on the measured diffusion coefficients are so small. What is the reduced chi2 of the fits in Figs. 5 and 6? The authors say a few lines later that "the values of the diffusion coefficients depend on the humidity that was not precisely measured". In Line 192f, they write about "changes in the relative humidity of the gas volume due to leaks". Why does this not affect the extraction of D_T and D_L from the data? In addition, the statements in Line 234ff are hard to understand. Do the authors want to say that they obtained an estimate for the humidity from a comparison of the measured drift velocity with Magboltz, and that these humidity values were then used to obtain the \pm 4% quoted for the Magboltz "prediction" in Line 233? In any case, uncertainties (statistical or systematic or both) have to be attached to the measured values of the diffusion constants.
c) It is somewhat difficult to reconcile the results for the resolution, the deformations and the efficiency, because it seems that they are based on different data sets or at least on different cuts applied to the data. For example, the track selection seems to be much stricter for the resolution studies in terms of fiducial cuts and amplitude (ToT) than for the efficiency. I understand that there may be reasons of statistics, etc. to do so, but I still want to express my concern that in the end, for physics measurements, it is the combined performance for a given set of cuts that counts, not the best possible value for each single parameter, obtained under different conditions.
d) The procedure leading to Figs. 8 and 10 is not understandable, at least not to this referee. The text in Lines 263 - 301 and in Lines 323 - 336 has to be completely rephrased in my opinion. Examples of phrases to be improved are:
- "the module was regrouped in four 256x256 pixel planes put side by side on the horizontal axis",
- "A bias in the mean residual..." => shouldn't this statement also be true for Figs. 7 and 9? If yes, it should be moved there.
- "due to the presence of the dike pixels..." => dike pixels have not been defined
- "the region near the edge of 5 pixels was removed"
- "a region of 10 pixels was removed"
e) Generally, the presentation towards the end of the manuscript (End of Section 5 and Section 6) seems to decline in quality with respect to the other Sections. I suggest that the authors have a closer look at these sections. In Section 5.4, no details are given on how the numbers are defined or extracted (see also comment a) above). In Section 6, the difference between the number of hits (corresponding to number of ionization electrons) and the time over threshold (corresponding to charge after amplification in the Micromegas grid) should be made clear. In addition to the distributions of the number of hits, also the distributions of the ToT could be shown. The authors mention that the measured mean number of hits is in agreement with the prediction from ref. [13]. But given the Landau fluctuations of the number of ionization electrons and the large tail, wouldn't the most probable value be a better number for this comparison? Alternatively, one could show the predicted distribution along with the measured one.
Additional comments:
Abstract:
- Line 19f: quoting the result for transverse and longitundinal diffusion coefficients without specifying the gas does not make sense.
- Line 20: remove "D_T is"
- Line 22: the phrase "the diffusion measurements have negligible errors" is not valid. Every measurement must be accompanied by an estimate of the uncertainty!
Introduction:
- Line 36: what clusters are meant here? Ionization clusters?
- Line 43f: with an ENC of 70 e-, why is the threshold set to more than 7 sigma?
- Line 45: be quantitative instead of claiming "high efficiency"
- I did not find the information on the number of pixels per TPX3 chip in the paper
- Line 50: "readout" => "read out"
Section 2:
- Line 72 and 89: "guard" => "guard electrode"
- Line 86: "diameter"
- Line 86: check use of "and"; what are the "guard strips"? Can they be indicated in Fig. 2?
- Line 92: "two 50 um thick Kapton windows"
Figure 1:
- should be enlarged
- labels and a scale should be added
- "rendering"?
- stick to either "32-GridPix module" instead of "8-quad module" throughout the text
Figure 2:
- add labels, e.g. for guard wires, guard strips, guard electrodes, pillars for field wires, etc.
Section 3:
- Line 114: "At DESY, the Mimosa26 silicon..." (add comma and remove parentheses)
- Line 124: "parameters"
- Line 125: add "and" before "oxygen"
- Line 134: "systems"
- Line 137: "of" => "with respect to"
Figure 3:
- is this a top or side view?
- add labels
Table 1:
- is the number of runs important?
- are the beam momenta relevant? (see also corresponding comment below)
Section 4:
- Line 155: check language in "Telescope tracks were selected with at least 5 out of 6 planes on the track" => "Telescope tracks were required to have hits in at least 5 out of the 6 planes"
- Line 156: why is the track chi2 cut so large? What is the distribution of the reduced chi2?
- Line 176: with respect to what has the acceptance window for GridPix hits been defined?
- Line 177: "was" => "were"
- Line 179: check language in "quadratic track B=1 T model"
- Line 180: give values for the "expected uncertainties". How were they obtained?
- Line 181f: the phrase "outlier removal at respectively 10, 5 and 2.5 sigma level" is hard to understand, please rephrase. What sigma do you refer to here?
- Line 183: why do only 25% of all hits lie on the track? Figure 4 looks much cleaner. Please explain.
- Line 185: what orientation does the "plane in the middle of the TPC" have?
- Line 199: add "from" before "run"
Figure 4:
- red and blue lines are hard to distinguish
- green points are hardly visible
- "driftplane => "drift plane"
Section 5:
- Line 208: "Secondly" would require a "firstly" before
- Line 220: "staying 20 pixels away from the chips edges" => are all 4 chips edges meant here?
- Line 221: see above comment on "resolution"
- Eq. (2): give a reference, e.g. [Yonamine et al., JINST 9 (2014) C03002]
- Line 228f: "sensors", "windows"
- Line 244f: "resolution ... in the drift plane" => should be "residual width in z direction"? The expression "drift plane" has not been defined.
- Lline 249: why is the ToT cut chosen for the z residuals so much higher than the one applied for the xy residuals (0.6us vs 0.15us)? How does this affect the efficiency?
- Line 257: what do the authors conclude from the discrepancy between the measured value of D_L and Magboltz? Is Magboltz data wrong, or are there systematic effects that were not taken into account?
- Line 274f: it would greatly facilitate reading if the chip numbers quoted in the text were visible in Figs. 7 and 9, without having to go back a few pages to Fig. 2.
- Line 302: "electrons will drift mainly along the magnetic field lines" seems not entirely correct. What is the value of \omega\tau? I suggest to remove the sentence, as it is not needed.
- Line 312: mention that these are now biased residuals, in contrast to the ones for B=0T, where the external track was used as reference
Figures 7, 9:
- add labels and units for z axis
- is the binning really 8 x 16 pixels? Zooming in, it seems that there is an equal number of bins in x and y for each chip, which would imply an 8 x 8 binning.
Figures 8,10:
- add labels and units for z axis
- what is the "regrouped expected hit position"?
Section 6:
- Line 350f: check language in "the B=0T analysis selects the ...." => "For the analysis of the data with B=0T, the chips ... were selected" or similar?
- Line 352: why were chips 12, 13, 20, and 21 excluded for the B=0T analysis? For a proper comparison of B=0 and 1T data, wouldn't it be advisable to use the same data set?
- Line 362: typo in "possibility"
Figure 11:
- typo in "per per"
- check language in caption
Section 7:
- the authors emphasize here again that data were taken at two different beam momenta, but throughout the analysis in the previous sections, no mention is made, which beam momentum the data correspond to. Have both momenta been used, e.g. also for Section 6? Probably the authors should state for each Section, which beam momenta were used
- Line 380f: this statement does not make much sense, unless an explicit public link to the data is given
Language and format:
- put symbols for physical quantities in italics: B, x, y, etc.
- do not put units in italics, e.g. GeV, cm
- use mathematical symbols where applicable, e.g. "\times" instead of "x"
- use SI units where applicable, e.g. there is no need to define T as Tesla when talking about magnetic field strength
- check use of hyphens (missing very often, especially in compound modifiers like "32-chip module", "high-precision tracking",...)
- references: should be in square brackets, not in parentheses; use [1,2] instead of (1), (2).
- format tables 1 and 2 for better readibility
- slided => slid
- avoid use of jargon, e.g. "quads", "guard", etc.
- avoid unphysical statements like "great precision", "high efficiency", ...
- check use of commas
EIC Gridpix project - test beam in the US - is approved
https://indico.jlab.org/event/659/
The extension is also approved https://indico.jlab.org/event/751/
The quad detector has been shipped to Yale