Primary Collimator Live Page
GEM rotator movie #1
GEM rotator movie #2
Tony's Status Report
Klaus's Status Report
Michael's Status Report
Yongguang's Status Report
Qweak Simulation Forum
Keep Out Zones a la Allena
Previous Reports
Virginia Tech Update for Tracking Group,
Unless otherwise stated, all dimensions are in cm, and rates in MHz/octant. You can view a larger version of most plots by clicking on the picture, and get back to this page by using your browser's back button. All of the results shown are with the minitorus on and helium as the global volume, unless otherwise stated. To see a somewhat chronological version of this summmary, see the page that I made for Dave as we were iterating back and forth for_dave.html .
There have been two major developments since the last tracking group meeting. One is that Dave has chosen a rectangular shape for the cerenkov detectors. Based on that decision, we were able to work towards a final collimator acceptance. There are ep profiles for the full acceptance through QTOR (with lowered upper theta) at the collimator optimization live page, but a final collimator choice is not too far away now.
For reference, the scheme we are proposing is shown below (See Figure 1).
At the last meeting, we chose the downstream location for the primary collimator (See that report). The maximum acceptance through QTOR was found, and then using a reasonably sized bar at the focal plane, an upper theta was chosen that did not increase the error bar. Since then, Dave Mack has worked closely with us to help us further refine the collimator opening. We made plots of error on Qweak and inelastic percentage as a function of upper x of the bar for 5 lengths, 3 widths and 5 different z locations (See those plots). Dave then chose a z location that seemed to have good focus (z = 570cm), and eventually an upper x (328cm) and bar length (2m) and width (18cm), with cost being taken into consideration as well. A rectangular bar is made feasible with a less conservative estimate for the tolerable inelastic percentage (more on this later).
Choosing a collimator shape
There are several relatively simple collimator shapes that cover much of the maximum acceptance defined by QTOR. Of greatest concern was having edges "cut into" the collimator acceptance. First the acceptance was trimmed to cut out the worst of the spilling over at the ends of the bars. We then chose a shape that was simpler, ie. had less sides and no edges cutting into the acceptance (See Figure 8a,b). We then trimmed even more to cut out all of the events not making it onto the bar (See Figure 9a,b).
|
|
Figure 2a - Collimator definition with maximum acceptance through QTOR (green). |
Figure 2b - ep profile at focal plane for new collimator definition. |
|
|
Figure 3a - Trimmed version (blue) superimposed on version with |
Figure 3b - ep profile at focal plane for collimator definition |
Figures 2a and 3a showed the ep profile for the maximum acceptance through QTOR at the z location of the first upstream QTOR support bar. The following leftmost plots (Figures 4a-8a) show the simpler shapes that we tried for the collimator cutout. The plots are at a z location right after the primary collimator.
|
|
Figure 4a - Collimator version 1 (red) superimposed on trimmed version (black). |
Figure 4b - ep profile at focal plane for collimator Version 1. |
|
|
Figure 5a - Collimator version 2 (red) superimposed on trimmed version (black). |
Figure 5b - ep profile at focal plane for collimator Version 2. |
|
|
Figure 6a - Collimator version 3 (red) superimposed on trimmed version (black). |
Figure 6b - ep profile at focal plane for collimator Version 3. |
|
|
Figure 7a - Collimator version 4 (red) superimposed on trimmed version (black). |
Figure 7b - ep profile at focal plane for collimator Version 4. |
|
|
Figure 8a - Collimator version 5 (red) superimposed on trimmed version (black). |
Figure 8b - ep profile at focal plane for collimator Version 5. |
The table below summarizes the rate information for the full acceptance which makes it through QTOR with
the simpler version of the collimator and it's trimmed version. This is an abbreviated table - for more versions
and trimming steps, see the other versions Figures 13a-e and the two that were trimmed Figures 18a-h at the
page that I made for Dave.
Description |
ep rate (MHz) |
inelastic rate (MHz) |
inelastic percentage (%) |
Q2 |
Error on Qpweak |
Full acceptance through QTOR |
978 |
1.22 |
.12 |
.0265 |
4.01 |
Above, trimmed for 18cm x 2m bar |
941 |
.674 |
.072 |
.0265 |
4.05 |
Version 5 |
787 |
.571 |
.073 |
.0286 |
4.14 |
Version 5, trimmed to fit on bar |
761 |
.554 |
.073 |
.0285 |
4.18 |
|
|
Figure 9a - The proposed collimator cutout (red) superimposed |
Figure 9b - The ep profile at the focal plane |
Region II sanity check
An intermediate step in this process was to choose versions 3 and 5 of the simplified collimators and look at
the rates and errors, as well as the acceptance at the Region II chambers and the minitorus
(See Figure 10). Norm felt that there was no real difference between those two versions as far as the
Region II chambers were concerned, and at the entrance to the minitorus the conclusion is similar. Even the
moller rates for version 3 and 5 at the middle chambers are comparable at 33 kHz/nA and 28 kHz/nA respectively.
|
Figure 10 - ep profile at the entrance to the minitorus, |
The final motivator for choosing a trimmed version 5 is that it may fit better on an 18cm wide bar due to the smaller goatee. But when we began looking into this, we realized that there also may be events at the top of the focal plane profile that are being poorly focused. The only way to really know is to look at the unradiated profile at the focal plane, to see if it fits on the cerenkov bar that we choose. For an 18cm bar it seems to fit fairly well.
|
|
Figure 11a - ep profile for Version 5 at the focal plane. |
Figure 11b - ep profile for Version 5 at the focal plane, unradiated. |
Inelastic Ratio
This is the reasoning behind the "maximum tolerable inelastic fraction" of 0.08% that we are currently specifying.
We used to have this set down at 0.02%, but we think this was probably unreasonably conservative. The new number
is still rather conservative.
Reasoning: The new upper limit of 0.08% corresponds to an increase in the overall Qweak error bar from 4.10% to
4.15%. Assuming the inelastic asymmmetry is about 10 times larger means the contribution to the asymmetry will
be 0.8%. If we assume that we can measure/estimate this to ± 50% of itself, then the fractional error on the
asymmetry from this effect is 0.4%. The error on Qweak from this (after multiplying by 1.64) will be 0.7%. That
added in quadrature with 4.10% gives 4.15%.
Additional Items of Interest
Some other things that have been considered at various stages along the way are the effect of changing
the main torus current by +/- 1 %, turning the minitorus off, running with air instead of helium, and
the effect of increasing the target endcap thickness on the general shape at the focal plane. For those
results, click on the appropriate link below. Bear in mind that the results are for rough comparison to
the "standard" listed in the individual table.
main magnet current changes
turn minitorus off or thicker target endcap
turn minitorus off or change GLOB into air
1. Front-end electronics problem:
As we set up to read out the 32 channels of the drift chamber into the TDCs we noticed a problem. When we connected the ribbon cables to the Nanometrics amplifier/discriminator boards we saw what I characterize as oscillations of the amplifiers. The frequency was typically 10s of MegaHertz but varied as we moved cables/touched grounds etc. After much trouble-shooting to exclude external radiation sources, varying grounding schemes and shielding, and adjusting of the threshold level, - all to no avail - the solution we arrived at was to add a 1000pf capacitor to ground on the input of the amplifier. This allowed us to reduce the threshold a factor of 30 from where it was last fall with no oscillations. It remains to be seen what affect this capacitor has on the timing resolution of the chamber. We will investigate this.
2. VME DAQ:
We have been plagued for awhile with problems with our VME DAQ. We finally traced it to the controller card on the VME side of the VME-PCI interface. As we were preparing to ship it out for repairs the problem (garbled bits in the address word) seems to have cured itself. It has worked fine for a couple days now, so we're writing the code needed to read out the CAEN V767 128 channel multihit TDC in MIDAS. Then we will start to look at drift time distributions in the prototype chamber.
