EC Tests at UVa, 2016

Summaries and Useful Links:

1. Summary of Tests in 2014.
2. Summary of Tests in 2015.
3. Preshower tile radiation hardness test page in 2016.
4. Location of materials, PMTs and detectors in room 2 of the Detector Lab at UVA.

October 31^st - December 22^nd, 2016

1. 2-fiber Tests of Irradiated Preshower Tiles from CNCS and Kedi (Part 2):
   Using the Kedi and CNCS Preshower hexagon tiles with two 1.0-mm diameter, 1.2-m long Y11 (200 ppm) fiber. The fiber ends were cut off so that only about 20 cm of fiber is not embedded on each fiber end. These tiles were tested after being irradiated at Jefferson Lab between February and April 2016. Click here.
   
2. This set of measurements is a follow up to those conducted in June and July 2016.
3. Details involving the electronics, DAQ and HV settings can be found here.
4. A summary of the test measurements can be found at the following link.
5. In general, the higher radiation dosage the preshower tile received, the worse its performance was after being irradiated. However, Kedi #2, which was upstream of the target, but received about 185 kRads, appears to perform better than the tiles placed downstream of the target, which also received over 150 kRads.
6. It appears that PMT S/N 24561's quantum efficiency may have degraded some over the past two years. The results with this PMT even for the IHEP, which was not irradiated, consistently performed worse compared to the earlier measurements. This PMT has been extensively used over the past three years, and hence, we can expect that its QE should be slowly worsening.
7. Typically the SPE peak location can be determined within about 2 QDC channels. This dominates the uncertainty of the measurements to be at roughly the 9% level, though for some runs when there are fluctuations near the SPE peak, the uncertainty can be worse by a factor of two.

August 29^th - September 24^th, 2016

1. Large and Forward Angle SPD tests with UVA GEM Chambers:
   The NIM and VME electronics were moved to the GEM lab at UVA in order to perform cosmic tests with three GEM chambers. The VME crate was coupled with a GEM crate in order to readout both the VME for the QDC and TDC and SRS modules for the GEMs. This process involved a month long process to readout the VME crate using VxWorks with a CentOS 7 operating system.
   
2. The tests were conducted using the test stand in room 106D. The test scintillator was placed below the three GEM chambers in the GEM stand. The test scintillator was centered underneath the GEMs and the trigger scintillators. The trigger scintillators consisted of two bars at the top of the detector stack, and one bar at the bottom of the detector stack. In both cases, the front scintillators were used in the trigger, and the back scintillator bars were unused. The data was then readout using CODA on srscoda. The trigger was formed by taking an "AND" of the left and right PMT signals. An "OR" was used for the top scintillator bars (front and next one in from the front). Then an "AND" was formed between the "OR" of the top bars and the "AND" of the bottom front bar. The discriminator threshold was set to -10 mV.
3. Dimensions of trigger scintillators:
   Top scintillators: 160 cm (L) x 22.2 cm (W for two bars) x 1 cm (thickness with wrapping)
   Bottom scintillator: 198 cm (L) x 24.8 cm (W) x 1 cm (thickness with wrapping)
4. Here is the HV channel map along with the HV settings used during the tests, and here is the signal channel map. In the last column, the cabling was connected so that each detector had the same QDC and TDC channel. Hence, only one column is listed for both the QDC and TDC, even though two separate modules were used.
5. A few measurements were conducted in early September without the GEM readout to determine the baseline for the timing resolution measurements with one of the 5-cm thick trigger bars and the LASPD bar with the 55-degree lightguide, runs 294 and 319, respectively.
6. For the LASPD bar, the same PMTs were used as in the tests from 2015. To keep the PMTs connected firm to the scintillator surface, special connectors were made via 3D printing to fix the PMTs to the scintillator surface. The interface between the bar and PMT were coated with optical grease to improve the light yield. The connectors and PMTs were then wrapped in vinly and black electrical tape. The wrapping was completed in January 2016, by the time the test was conduced in September, the tape wrapping was no longer as firm as when it was first wrapped, which probably resulted in a somewhat reduced light yield compared to the tests in 2015.
7. For the FASPD bar, one 1.0-mm diameter, Y11 (200 ppm) fiber was used. The fiber was cut from the 500-m spool with a length of 212 cm. After the fiber was embedded into the scintillator groove, the fiber ends were cut off so that only 20 cm of fiber was not embedded in the tile. The fiber ends were then polished with polishing paper from JLab.
8. The fiber inside the groove has 3 turns for an approximate embedded length of 168 cm and a total length of 208 cm, which includes the length of fiber outside the scintillator.
9. The WLS fibers were coupled to a XP2262 PMT by using the outer plastic casing of the PMT, and the area from the scintillator to the PMT was made light tight by using black tape and tedlar. The WLS fibers ends were held ridged by using a machined piece of black derlin. Optical grease was placed on the PMT surface.
10. A first attempt to test the FASPD was done using black vinly tape, but it was soon realized that the vinly tape is partially tranparent. This did not affect any previous test results, since they were all conducted inside of a dark box. The early measurements when the detector-PMT were not light tight involved runs 371 and 372.
11. Here is a list of relevant runs and a brief discription of them.
12. Conclusions and Observations:
1. Run 294 using the 5-cm thick trigger bar was analyzed to study the timing resolution of the long scintillator bars in the detector stand. It was found that the timing resolution was poor, between 0.786 and 1.111 ns. The value of the timing resolution depends on wherther or not one assumes the top and bottom scintillator bars are identical. Since they are not identical, the consevartive estimate of 1.111 ns may be more accurate. The 5-cm thick bar was assumed to have a timing resolution of 58 ps as observed in 2015.
2. The FASPD test conducted with the GEMs can be compared with the results from 2015. During those tests, the number of photoelectrons achieved was 9.5 to 11. The difference between those and the results presented here can probably be attributed to using a different coupling method for the PMTs and the WLS fibers. Also these tests were conducted outside of a dark box.
3. The WLS fibers probably need to be glued into place to ensure adequate coupling between the PMT and fibers. Due to lack of time, modifications to the fiber coupling to the PMT were not attempted.


Table 7: FASPD MIP Response, not including GEM analysis



run#	Bar Length (cm)	# fiber turns	fiber length embeded (cm)	measured single p.e. position	measured MIP position	measured # p.e. using single p.e. position	comments
373	23.2	1 fiber 3 turns	168	18.2	QDC Spectra: 125.7	6.9	XP2262 PMT, HV = -1950 V
375	23.2	1 fiber 3 turns	168	26.4	QDC Spectra: 165.8	6.7	XP2262 PMT HV = -2000 V

June 7^th - July 11^th, 2016

1. 2-fiber Tests of Irradiated Preshower Tiles from CNCS and Kedi:
   Using the Kedi and CNCS Preshower hexagon tiles with two 1.0-mm diameter, 1.2-m long Y11 (200 ppm) fiber. The fiber ends were cut off so that only about 20 cm of fiber is not embedded on each fiber end. These tiles were tested after being irradiated at Jefferson Lab between February and April 2016. Click here.
   
2. Optical grease was coated onto the surface of the PMT's cathode, and the four fiber ends were placed up against the coated cathode. The fibers were held rigid using a DDK ferrule.
3. Before taking the tiles to Jefferson Lab, optical grease was applied in the grooves of the tile. Then the tiles were wrapped in a new piece of tyvek, which was then wrapped in a piece of black tedlar.
4. For the measurements at UVA, the 5-cm thick scintillator counters formed the trigger. The HVs remained the same for these counters. For the preshower tiles, one XP2262 PMT with S/N 24561 was used with HV = -2100 V.
5. Conclusions and Observations:
1. Quickly, the tests revealed that the photoelectron yield was vastly different compared to the test results in 2015.
2. Applying fresh grease into the tile grooves did improve the light yield, but the results were still significantly lower than seen previously before irradiating the tiles. Since it wasn't clear if the lower yields were due to the irradiation or something else, the IHEP tile, which was not irradiated, was placed in the system to check for reproducibility.
3. Unfortunately, the results with the IHEP were siginficantly lower compared to the earlier result from just April 2016. During run 1629, it was discovered that the PMT base had come away from the PMT iself. Reseating the base to the PMT, appeared to improve the preformance of the IHEP tile. However, the result of 94 photoelectrons was not reproducible. The PMT was changed twice to test if the original PMT (S/N 24561) had degraded. However the two PMTs tested did not result in any better than about 78 p.e. One thing that was clear is that the PMTs SPE peak location was difficult to isolate cleanly, which produced large variations in the test results for runs 1631 to 1639. After run 1632, the signal from PMT S/N 23963 became noisy. This was later traced to a broken wire on the PMT HV divider. For run 1639, the SPE peak is quite wide at almost 20 QDC channels and the location significantly different compared to the earlier runs with PMT S/N 20673. This run was conducted over 10 days, so perhaps the gain shifted during the run.
4. When the grease was replaced in the IHEP tile, the fibers previously were optically cemented into a ferrule (from DDK). It was difficult to reinsert the fibers into the grooves, and it is possible the fibers may have some slight damage due to the embedding.
5. The optical grease had expired in March 2104, so the hypothesis was that the optical grease was old and no longer providing reliable results. In July, new optical grease from Eljen was ordered to repeat the tests.


Table 6: CNCS and Kedi Tiles, Multi-fiber (two Y11, 1mm dia) test results after irradiation at JLab



run#	Tile	# fiber turns	fiber length embeded (cm)	measured single p.e. position	measured MIP position	measured # p.e. using single p.e. position	comments
1617	CNCS # 3	2.5 each fiber	151	31.2	QDC Spectra: 1206	38.6	XP2262 PMT, S/N 24561 HV = -2100 V Post irradiation in Hall A Interrupted by power outage
1618	CNCS # 3	2.5 each fiber	151	30.9	QDC Spectra: 1266	41.0	XP2262 PMT, S/N 24561 HV = -2100 V Tile repositioned wrt trigger bars
1619	CNCS # 3	2.5 each fiber	151	31.0	QDC Spectra: 1370	44.2	XP2262 PMT, S/N 24561 HV = -2100 V Applied fresh grease to tile grooves and polished fiber ends
1620	CNCS # 3	2.5 each fiber	151	28.2	QDC Spectra: 1454	51.4	XP2262 PMT, S/N 24561 HV = -2100 V Applied a thinner layer of grease to PMT surface
1621	CNCS # 3	2.5 each fiber	151	29.8	QDC Spectra: 1570	52.7	XP2262 PMT, S/N 24561 HV = -2100 V Cut and repolished fibers
1622	CNCS # 1	2.5 each fiber	151	30.1	QDC Spectra: 1106	36.8	XP2262 PMT, S/N 24561 HV = -2100 V Post irradiation in Hall A
1623	CNCS # 1	2.5 each fiber	151	29.4	QDC Spectra: 1340	47.3	XP2262 PMT, S/N 24561 HV = -2100 V Applied fresh grease in tile grooves
1625	IHEP	2.5 each fiber	155.4	22.7	QDC Spectra: 1367	60.3	XP2262 PMT, S/N 24561 HV = -2050 V Reproducibility of Run 1608
1628	IHEP	2.5 each fiber	155.4	22.6	QDC Spectra: 1373	60.8	XP2262 PMT, S/N 24561 HV = -2050 V Changed QDC gate width from 62.5 ns to 100 ns
1629	IHEP	2.5 each fiber	155.4	25.3	QDC Spectra: 1952	77.3	XP2262 PMT, S/N 24561 HV = -2050 V Adjusted discriminator thresholds to -50 mV Reseated PMT base
1630	IHEP	2.5 each fiber	155.4	25.8	1817	70.5	XP2262 PMT, S/N 24561 HV = -2050 V Recentered trigger bars wrt prehower tile
1631	IHEP	2.5 each fiber	155.4	28.3	QDC Spectra: 2005	70.5	XP2262 PMT, S/N 23963 HV = -2000 V Changed PMT
1632	IHEP	2.5 each fiber	155.4	29.1	QDC Spectra: 2266	78.0	XP2262 PMT, S/N 23963 HV = -2000 V Cleaned and applied fresh grease on PMT and fiber connector
1636	IHEP	2.5 each fiber	155.4	33.9	QDC Spectra: 2556	75.4	XP2262 PMT, S/N 20673 HV = -1850 V Changed PMT
1637	CNCS # 1	2.5 each fiber	151	28.6	QDC Spectra: 1855	64.9	XP2262 PMT, S/N 20673 HV = -2000 V
1638	CNCS # 1	2.5 each fiber	151	32.2	QDC Spectra: 1798	55.9	XP2262 PMT, S/N 20673 HV = -1850 V PMT was shifted under trigger bars to enhance SPE peak
1639	IHEP	2.5 each fiber	155.4	38.5	QDC Spectra: 2331	60.6	XP2262 PMT, S/N 20673 HV = -1850 V Cleaned applied fresh grease in tile grooves

May 10^th - May 23^rd, 2016

1. 1.5 mm Scintillator Hedgehog Test with lead (Part 3):
   Using 1.5-mm thick scintillator tiles from China with 96x 1.0-mm diameter, approximately 40-cm long Y11 (200 ppm) fiber. The fiber ends were polished with polishing paper from JLab.
   
2. Details of the test setup can be found here.
3. For the first test, the scintllator stand was disassembled so that lead pieces were placed between each scintillator plate with tyvek paper surrounding both the top and bottom sides of the scintillator. We only had enough tyvek paper to stack 15 scintillator layers, compared with the 25 layers in the previous measurements.
4. Before adding the bottom scintillator plate, a piece of tyvek paper with holes punched through it was placed first. This piece of tyvek was cut with hexagonal sides so that they could be folded up to reflect light coming out of the sides of the scintillator plates. After all the tiles were stacked, a separate piece of tyvek paper was placed above the stack to reflect light back from the fiber ends, and an aluminum plate was placed on top of this to hold the paper in place.
5. Conclusions and Observations:
   
   1. Prior to this test two of the WLS fibers broke from repeated manipulation of the fibers. If the light yield is directly proportional to the number of fibers, then this would result in about a 2% loss of light compared to having all 96 fibers.
   2. For run 1615, the fiber coupling to the PMT surface was attempted to be optimized, but instead the light yield dropped. When additional tyvek sheets were available, the additional 10 scintillator and lead layers were added. For run 1616, the light yield did improve compared to only 15 layers, but it was still significantly lower than expected.
   3. Something significant happened between run 1614 and 1615 to produce a drop in light yield. This problem would presist into the preshower tile tests that began in June 2016.
   
   
   Table 5: Hedgehog test results with reflectors and lead layers
   
   

   run#	Fiber length [cm]	measured single p.e. position	measured MIP position	measured # p.e. using single p.e. position	# p.e. per layer	comments
   1614	40	34.5	QDC Spectra: 558.5	16.2	1.08	XP2262 PMT, HV = -2100 V 15 layers Tyvek paper above and below each scintillator
   1615	40	33.0	QDC Spectra: 397	12.0	0.8	XP2262 PMT, HV = -2100 V 15 layers Tyvek paper above and below each scintillator
   1616	40	32.6	QDC Spectra: 588.9	18.1	0.72	XP2262 PMT, HV = -2100 V 25 layers Tyvek paper above and below each scintillator

   
   

April 15^th - May 8^th, 2016

1. 1.5 mm Scintillator Hedgehog Test (Part 2):
   Using 1.5-mm thick scintillator tiles from China with 96x 1.0-mm diameter, approximately 40-cm long Y11 (200 ppm) fiber. The fiber ends were polished with polishing paper from JLab.
   
2. Details of the test setup can be found here.
3. For the first test, the scintllator stand was disassembled so that mylar pieces were placed between each scintillator plate. The mylar sheets were produced by the UVA machine shop in a hexagonal shape with 96 holes punched through the material.
4. Before adding the bottom scintillator plate, a piece of tyvek paper with holes punched through it was placed first. This piece of tyvek was cut with hexagonal sides so that they could be folded up to reflect light coming out of the sides of the scintillator plates. After all the tiles were stacked, a separate piece of tyvek paper was placed above the stack to reflect light back from the fiber ends, and an aluminum plate was placed on top of this to hold the paper in place.
5. Conclusions and Observations:
   
   1. The bend radius of the fibers from the stack to the PMT is probably too small. The minimum bend radius should be no less than 10 cm. This can result in substantial light loss from the fibers. For upcoming measurements, we will attempt to adjust the bend radius to make it more ideal.
   2. The data from runs 1600 and 1601 is reproduced in Table 4 from Table 1.
   3. When the assembly was taken apart, it was noticed that one of the WLS fibers had kinked. The fiber was left as is in the assembly.
   4. The light yield with the mylar between the scintillator layers (run 1609) is quite low compared to the light yield using either no reflectors or printer paper. The experimental setup was checked, and obvious reasons for the poor light yield was found. However, the the repeated assembly and disassemly of the stack probably has an adverse affect on the WLS fibers. Over time, the fiber quality worsens due to repeated movements of the fibers.
   5. The yield with the tyvek paper (run 1610) is much improved over the mylar, though it is still lower (11.5%) than without any reflectors. Before this test, it was noted that five fibers now show kinks in the region where the scintillator stack is held together, though there could be a few more that were not identified. The result is slightly better (~ 4%) compared with printer paper between the layers. It should also be noted that the tyvek paper either had blue or red ink printed on one side of some of the sheets. This probably also has some effect on the effectiveness of the reflectors.
   6. Still using tyvek paper between the layers (run 1612), the edges of the scintillator plates were wrapped with mylar, which was tapped in place. Around this layer of mylar, black vinly tape was wrapped to hold the mylar close to the edges. Before this test, one of the WLS fibers had broken so that 95 out of the 96 fibers collected light to the PMT. The result is consistent with using tyvek along the six edges. However, the tyvek was not held tightly to the edges of the hexagon plates.
   7. For a baseline measurement, the mylar wrapping for the edges was removed, and the edges were left completely uncovered during run 1613. In this case, a 15.5% reduction in light yield was seen without wrapping along the edges.
   
   
   Table 4: Hedgehog test results with reflectors
   
   

   run#	Fiber length [cm]	measured single p.e. position	measured MIP position	measured # p.e. using single p.e. position	# p.e. per layer	comments
   1600	40	34.2	QDC Spectra: 1353	39.6	1.58	XP2262 PMT, HV = -2100 V 25 layers No reflectors
   1601	40	34.1	QDC Spectra:1144	33.6	1.34	XP2262 PMT, HV = -2100 V 25 layers Printer paper between each layer
   1609	40	34.3	QDC Spectra:827.2	24.1	0.96	XP2262 PMT, HV = -2100 V 25 layers Mylar sheets between each layer
   1610	40	33.6	QDC Spectra:1177	35.0	1.4	XP2262 PMT, HV = -2100 V 25 layers Tyvek paper between each layer
   1612	40	34.6	QDC Spectra:1202	34.7	1.39	XP2262 PMT, HV = -2100 V 25 layers Tyvek paper between each layer Mylar wrapped tightly around the edges
   1613	40	34.4	QDC Spectra:1021	29.6	1.18	XP2262 PMT, HV = -2100 V 25 layers Tyvek paper between each layer No wrapping around the edges

   
   

April 7^th - 15^th, 2016

1. 2-fiber Tests of Preshower Tiles from CNCS and Kedi:
   Using an additional two CNCS and two Kedi Preshower hexagon tiles with two 1.0-mm diameter, 1.2-m long Y11 (200 ppm) fiber. The fiber ends were cut off so that only about 20 cm of fiber is not embedded on each fiber end.
   
2. Optical grease was coated onto the surface of the PMT's cathode, and the four fiber ends were placed up against the coated cathode. The fibers were held rigid using a DDK ferrule.
3. Optical grease was applied in the grooves of the tile.
4. The 5-cm thick scintillator counters formed the trigger. The HVs remained the same for these counters. For the preshower tiles, one XP2262 PMT was used with HV = -2100 V.
5. Conclusions and Observations:
1. After run 1602, the fibers were repositioned onto the PMT cathode, since they appeared not to be flush against the cathode's surface. This had the intended effect and the NPE yield improved considerabley for run 1603.
2. The surfaces of the Kedi tiles appear to be roughed up: photo #1 and photo #2, which is next to a square tile from SDU. Even after repositioning the fibers on Kedi tile #6, there were no improvements to the NPE yield.
3. The NPE yield from the CNCS tiles #5 and #6 are comparable to those for #1 to #4 tested last year, though they are also systematically lower.
4. Run 1608 is a reproducibility run with a well-tested IHEP tile with similar dimensions and fibers. The result is consistent with the achieved number of photoelectrons in 2015.
5. Table 2 summarizes the test results from these tiles.


Table 2: CNCS and Kedi Tiles, Multi-fiber test results (two Y11, 1mm dia)



run#	Tile	# fiber turns	fiber length embeded (cm)	measured single p.e. position	measured MIP position	measured # p.e. using single p.e. position	comments
1602	CNCS # 6	2.5 each fiber	151	37.7	QDC Spectra: 1988	52.8	XP2262 PMT, S/N 24561 HV = -2100 V
1603	CNCS # 6	2.5 each fiber	151	32.64	QDC Spectra: 2426	74.3	XP2262 PMT, S/N 24561 HV = -2100 V Fibers repositioned to Cathode
1604	CNCS # 5	2.5 each fiber	151	31.88	QDC Spectra: 2500	78.4	XP2262 PMT, S/N 24561 HV = -2100 V
1605	Kedi # 6	2.5 each fiber	151	26.0	QDC Spectra: 1468	56.6	XP2262 PMT, S/N 24561 HV = -2050 V
1607	Kedi # 5	2.5 each fiber	151	26.0	QDC Spectra: 1352	52.0	XP2262 PMT, S/N 24561 HV = -2050 V
1608	IHEP	2.5 each fiber	155.4	23.2	QDC Spectra: 2197	94.6	XP2262 PMT, S/N 24561 HV = -2050 V Reproducbility Check

February 20^th - April 7^th, 2016

1. 1.5 mm Scintillator Hedgehog Test:
   Using 1.5-mm thick scintillator tiles from China with 96x 1.0-mm diameter, approximately 40-cm long Y11 (200 ppm) fiber. The fiber ends were polished with polishing paper from JLab.
   
2. The 5-cm thick scintillator counters formed the trigger. The HVs remained the same for these counters as used in 2015. For the fiber readout, a XP2262 PMT was used with HV = -2100 V.
   
3. Optical grease was coated onto the PMT's cathode surface, and the fiber ends were optically glued into a plastic holder made of delrin. After hardening, the fibers ends were cut and the entire holder was polished using sandpaper and the polishing paper from JLab. Then the fiber holder was placed up against the cathode surface.
4. The scintillator plates were stacked by aligning the alignment hole for each of the plates. 2-mm diameter brass rods were used to hold the plates in place, while the plates were stacked. The other ends of the fibers were then collected and pushed through the holes of this plastic holder with the handle attached until the fibers were all the way through all plates in the assembly.
5. Before adding the bottom scintillator plate, a piece of tyvek paper with holes punched through it was placed first. This piece of tyvek was cut with hexagonal sides so that they could be folded up to reflect light coming out of the sides of the scintillator plates. After all the tiles were stacked, a separate piece of tyvek paper was placed above the stack to reflect light back from the fiber ends, and an aluminum plate was placed on top of this to hold the paper in place.
6. Conclusions and Observations:
   
1. For run 1594, the light collection was not optimal. The fiber ends were not flush againt the PMT surface. Also the 50-50 splitter was still in place, so only about half of the signal was sent to the QDC. The SPE should have been closer to about 16.5 QDC channels, but clearly in the QDC distribution the SPE peak cannot be cleanly separated from the MIP peak. See figure in Table 1 below. For this test, there was not paper between the layers, and printer paper was used below the bottom layer and above the top layer.
2. For run 1596, a new holder made of delrin was used to hold the fibers to the cathode's surface, though the fibers were loose inside the holder. For this test, tyvek paper was used between the layers and above and below the top and bottom layers, respectively. The 50-50 splitter was bypassed so that all the PMT signal went to the QDC. The SPE was clear in this data, but the MIP was hard to discern.
3. For run 1599, the fibers were optically glued inside the delrin fiber holder. This fixed one end of the fibers. Then 21 additional scintillator plates were stacked on top of those in the previous configuration. The bottom four layers had tyvek paper in between them. For run 1600, the stack was disassembled and the tyvek paper between the layers was removed. Then the stack was put back together without any paper in between the layers. The results from these two runs are very consistent with each other, indicating the few layers of tyvek didn't change the results and that complete disassembly and reassembly of the stack produces reliable results.
4. For run 1601, the stack was once more disassembled and reassembled. However, this time printer paper was added between each of the layers. In this case, the photonelectron yield dropped by about 15% from the previous tests, which indicates that no paper allows some of the lost light to be collected by one of the neighboring scintillator plates. The reflectivity of printer paper is about 80%, which would appear to be consistent with the yield drop seen.


Table 1: Hedgehog test results



run#	Fiber length [cm]	measured single p.e. position	measured MIP position	measured # p.e. using single p.e. position	# p.e. per layer	comments
1594	40	21.2	QDC Spectra: 57	2.7	0.7	XP2262 PMT, HV = -2100 V four layers Configuration not optimal for light collection
1596	40	33.1	QDC Spectra: ----	N/A	N/A	XP2262 PMT, HV = -2100 V four layers MIP not clearly isolated
1599	40	33.2	QDC Spectra: 1315	39.6	1.6	XP2262 PMT, HV = -2100 V 25 layers Tyvek paper between bottom four layers
1600	40	34.2	QDC Spectra: 1353	39.6	1.6	XP2262 PMT, HV = -2100 V 25 layers No reflectors
1601	40	34.1	QDC Spectra:1144	33.6	1.4	XP2262 PMT, HV = -2100 V 25 layers Printer paper between each layer