A simple DEAR setup was installed at the DAFNE machine, in the FINUDA pit, on 14 October 1998, in order to perform background measurements with and without beams circulating in DAFNE.
The setup consists of 2 CCDs (I411 and I412), with two readout boxes (RB5, RB6), with an aluminium window of  50 mm, placed in the interaction region (IR) at a distance from the pipe of about 60 cm in the horizontal plane and of about 10 cm in the vertical plane. This setup stays in an asymmetrical position with respect to the Interaction Point (IP), namely is  located in front  of the flanges between the central quadrupole and the first quadrupole of the DAY-ONE IR - shifted of about 36 cm upstream the electrons-coming-in direction.
 

Experimental results for "non-machine" background
 

Without circulating beams, data were taken with a duration per run of 15 and 30 minutes; occupancies of the CCDs during this run-period were less than 1%.
The results we obtained can be summarized as follows:
 

• for what concerns the CCD-blindness a number of  (300 ± 30) clusters/hour/ half-CCD was measured;

• for what concerns the X-ray background a number of (0.85 ± 0.07) X/hour/CCD/keV in the region of interest (6.5 keV) was measured;

• the ratio between the total number of clusters and X rays in all the dynamical range (2 - 18 keV) was (50 ± 5).

The X-ray spectrum during this period was smooth, not showing any structure. It is worthy to mention  that the X-ray background  in the pit is smaller by about 40% than the one measured in the DEAR Laboratory.
 

Background measurements with circulating beams
 
 

Electrons or/and positrons beams circulated at DAFNE from 19 October 1998 to 2 November 1998. During this interval DAFNE staff did many tests and adjustments of the beam parameters. Injections in the rings were made quite often and conditions were rather unstable. However, there are few periods in which beam lifetime measurements were performed and, consequently, beam was rather stable for longer periods (at the level of few hours).
The electron and positron rings are not in the same condition: namely,  the average pressure in the positron ring is actually better than that in the electron ring (few units 10^-9 mbar).  This situation is reflected  by a shorter beam lifetime for electrons, due to the larger losses from beam-gas interactions.
When injections were frequent and not efficient we had ``occupancy-problems",  in the sense that CCD images were almost white, even for read-outs of about 30 seconds.
Stable beam-conditions, useful from the point of view of performing an analysis of the results, occurred during the following periods:

• 21 October, beam of positrons, with initial intensity of about 2 mA;

• 26-27 October, beam of positrons, with initial intensity of about 9 mA;

• 28 October, 2 beams of electrons, with initial intensity of about 23 mA;

• 30 October, beam of positrons, with initial intensity of about 38 mA.

We report below, as an example, the DEAR run of 21 October.

In the night of 21 October '98 a measurement of the beam lifetime of a positron beam was performed at DAFNE. A positron beam with an initial intensity of about 2 mA was injected and then left to decay.
We performed measurements with the test setup during a time of 3.75 hours with runs of 15 minutes each.  In Fig. 1 the beam intensity, as measured by DAFNE staff, is reported, while in Fig. 2 the number of clusters versus time, as recorded by DEAR, is presented (a cluster of hit pixels occurs whenever a charged or a neutral particle deposits energy in the CCD).
Numerical results relative to this beam are:

• Ratio between total number of clusters and X rays in the (2-18) keV region  (mean value): (53 ± 5);

• X-ray background in the 1 keV region of interest (mean value): (1.67 ± 0.42) X/hour/CCD/keV.

Figure 1: Positron beam intensity (values reported must be corrected by subtracting the pedestal), 21 October 1998.

Figure 2: Total number of clusters for runs of 10 minutes as a function of time, with the test setup, during the night of
              21 October 1998.
 
 

Summary of the results of the first period of background measurements on DAFNE
 

During this period of data taking, we observed that for the electron beam, the CCD closer to the coming-in direction of electrons registered (20-30)% more signal than the CCD further away.
In the same time for a positron beam the external halves CCDs registred (10-15)% more signal than the internal ones.
In Figure 3 an overall X-ray energy spectra representing data taken during this period (with electron and positron beams, as well as without beam) is shown. This spectra is without structures, except for the K_a peak of electronic Aluminium (at 1.49 keV).

In summary we learned:
 

• with machine turned off, the X-ray background in the energy region of interest is of the order of 0.85 counts/CCD/keV/hour; this value is 40% lower than the results obtained in the DEAR laboratory;

• when a positron beam with intensity 2-15 mA is circulating with stable conditions, the X-ray background in the region of interest is enhanced very little: it is on average (1.7 ± 0.4) counts/CCD/keV/hour;

• the number of clusters in the CCD varies approximately linearly with beam current; therefore the beam lifetime can be measured to a precision of about 20%;

• the ratio between the total number of clusters and the X-ray background in the (2-18) keV dynamical range varies as a function of beam current from about 40 to 100 for a positron  current variation from 2 to 30 mA;

• background events have a different cluster size in case of electrons and positrons; this feature deserves further studies;

•  if the beam is lost during injection, a large  X-ray peak appears at 1.49 keV, corresponding to the electronic K_a X-ray line in aluminium;

In the present commissioning phase of the machine, the circulating beams are of the order of few mA and the dynamical vacuum, consequently, does not exceed 3 x10^-9 torr. In these working conditions, the dominant cause for particle losses from the primary beams is the so-called Touschek effect.

In Table 1 (from ref.[1]) the rates of lost particles due to Touschek scattering are reported.
 
 

Table 1: Particle lost due to Touscheck scattering in the DEAR interaction region ( N=8.9 x 10¹⁰ particles per bunch ).

Particles with negative deviation (d < 0) are lost in the last quadrupole before the interaction point. Those with positive d are lost in the quadrupole after the interaction point.
 

Experimental data
 

We took into consideration the data taken in the run of 21 october, when a positron beam was performed at DAFNE. A positron beam with an initial intensity of about 2 mA was injected and then left to decay.
We performed measurements with the ``no-target 2 CCDs" setup during a time of 3.75 hours with runs of 15 minutes each.
With reference to Figure 2, one deduces that, for a current of about 1.6 mA (the mean value of the current during our 15 minutes data taking), the total number of clusters for 2 CCDs in 15 minutes was about 600; 290 clusters were generated by the ``non-machine" background, the remaining 310 by the lost particles from the primary positron beam.
Concerning the X-ray background, the statistics was too poor to have an analogous figure. However, a ratio of about 60 between the total number of clusters and the number of X rays (in the 2-18 keV dynamical range) was observed.This would result in a total number of X rays, observed in 15 minutes as accompanying the 600 clusters, of about 10. About 5 of these X rays are generated by the ``non-machine background', the remaining 5 can be attributed to the machine background.
Normalizing to 1 hour the experimental results, we obtain:
 

• Total number of clusters for 2 CCDs: (2400 ± 100) hour^-1;

• Total number of clusters after subtracting the "non-machine" background: (1240 ± 70) hour^-1;

• Total number of X rays for 2 CCDs (in the range 2-18 keV): (40 ± 10) hour^-1;

• Total number of X rays for 2 CCDs  after subtracting the ``non-machine" background  (in the range 2-18 keV): (20 ± 5) hour^-1.

DEAR Monte Carlo results
 
 

The elements considered in the DEAR Monte Carlo [2] were:
 

• Day-One interaction region, with the central quadrupole included;

• pipe: DAFNE pipe in the central region - Aluminium, 2 mm;

• CCDs placed at the distances of the pipe above mentioned and surrounded by an aluminium box of 1 cm thickness.

As input for the calculation the values reported in Table 1 were used. In the Table, the particles lost are referred to a number of particles per bunch of 8.9 x 10¹⁰, which gives a single bunch current of 45 mA (Project value). The machine was working without scrapers, with a 1.6 mA average current during the 15 minutes of our data taking. The beam measured lifetime in this period was about 130 minutes to be compared with the lifetime at full current, reported in Table 1, equal to 97 minutes. Therefore the rate of lost particles, input of the Monte Carlo, is about 1 x 10⁵ x 1.6/45 x 97/130 » 2.6 kHz.
Lost positrons were generated with the kinematics corresponding to the Touschek scattering (ref [3]).
A total number of 1.420.000 positrons lost from the primary beam were generated. Being the rate of lost particles due to the Touschek effect equal to about 2.6 kHz, this corresponds to about 546 seconds of beam time. The obtained results at the level of CCDs were:
 

•  total number of particles depositing energy in CCDs: 178 particles (positrons and electrons);

• X rays with energy below 18 keV: 3 X-rays.

This gives the following  results:
 

• Total number of clusters generated by a 1.6 mA positron beam circulating in DAFNE: (1173 ± 86) hour^-1;

• Total number of X rays in the range 2-18 keV generated by a 1.6 mA positron beam circulating in DAFNE: (20 ± 12) hour^-1.

The comparison between the experimental results of background measurements with circulating beams in the DAFNE machine, using a simplified version of the DEAR experimental setup, with a calculation performed with the DEAR Monte Carlo shows a fairly good agreement.

This makes us confident - within the limits of a comparison done for a simplified setup and machine conditions - in the reliability of the background evaluations performed for the DEAR setup at full luminosity and, specifically, in the results obtained in background suppression using different shielding configurations.
 

References

[1] S. Guiducci, Background calculation for the DAFNE experiments, Proceedings of the 5th European Particle Accelerator Conference (EPAC'96), Sitges (Barcelona), Eds. S. Myres  et al., Institute of Physics Publishing, Bristol and Philadelphia, 1996, pag. 1365.

[2] R. Baldini, C. Guaraldo et al., DAFNE Exotic Atom Research (The DEAR Proposal), LNF-95/055 (IR) (1995).

[3] S. Guiducci (DAFNE Machine staff), private communication.

Details of the calculation can be found in: C. Petrascu, Monte Carlo calculation for Touschek lost positrons hitting a simplified DEAR setup, DEAR Note, 2 Nov. 1998.