Professional Documents
Culture Documents
The MIDAS Experiment: A Prototype For The Microwave Emission of Ultra-High Energy Cosmic Rays
The MIDAS Experiment: A Prototype For The Microwave Emission of Ultra-High Energy Cosmic Rays
Recent measurements suggest that extensive air showers initiated by ultra-high energy cosmic rays (UHECR)
emit signals in the microwave band of the electromagnetic spectrum caused by the collisions of the free-electrons
with the atmospheric neutral molecules in the plasma produced by the passage of the shower. Such emission
is isotropic and could allow the detection of air showers with 100% duty cycle and a calorimetric-like energy
measurement, a significant improvement over current detection techniques. We have built MIDAS (MIcrowave
Detection of Air Showers), a prototype of microwave detector, which consists of a 4.5 m diameter antenna with
a cluster of 53 feed-horns in the 4 GHz range. The details of the prototype and first results will be presented.
1
2
trigger logic, and preliminary test results are dis- sent to the trigger master board to perform a sec-
cussed below. ond level trigger (SLT) at the level of the whole
camera. If the SLT condition is satisfied, the sys-
tem trigger pulse is send to the 4 ADC boards for
2. THE MIDAS PROTOTYPE
the readout. These first and second level trigger
The main goal of the MIDAS experiment is conditions are implemented as follows:
to detect and characterize the microwave emis- FLT: A channel is marked as triggered if the
sion from EAS. For this purpose, we have built running sum of 20 consecutive bins exceeds an
a wide field of view telescope with a camera of adjustable threshold. This status is extended for
53 channels of 0.8 GHz bandwidth in the 4 GHz 10 µs to allow coincidences between channels in
microwave band. The center of the camera is lo- subsequent trigger requirements. The threshold is
cated at the focus of a 4.5 m diameter parabolic continuously regulated to compensate the chang-
antenna. Each channel is equipped with a C-band ing background conditions and to ensure a con-
commercial satellite television receiver, i.e. a stant trigger rate per channel of 100 Hz. The
combination of feed horn, low noise amplifier, and procedure is illustrated in figure 1.
frequency down converter. The signal from each
of these receivers is then transformed to a volt-
age level using a fast power detector with loga-
rithmic response to the input RF power. The DC
output signal (ranging from 0.5-2.1 V) is continu-
ously digitized in custom 14-bit 20 MHz Analog-
to-Digital Converters (ADC) modules. Digital
samples are processed locally with on-board Field
Programmable Gate Arrays (FPGAs), where a
trigger logic allows discrimination of events. A
trigger master board is used to synchronize all
ADC boards, and to generate a global trigger.
It provides a system clock for simultaneous sam-
pling and receives the local trigger from each Figure 1. Response of one of the channels to a
ADC board via front LVDS-panel connections. pulse generated by a transmiting antenna. The
After a trigger, a 100 microsecond time stream ADC trace (black line), running sum (blue line)
of data (25 µs corresponding to data before the and threshold (red line) are presented.
trigger) is packed and buffered for readout via a
VME back panel and stored for later analysis. A
VME GPS module (Hytec 2092) is also used for SLT: The FLT information is sent to the Mas-
precise timing information. ter Trigger board where another on-board FPGA
is used to evaluate trigger conditions at the level
2.1. Trigger and DAQ systems of the whole camera. The SLT block is designed
The 53 receiver channels are serviced by 4 iden- to search for 4-fold patterns of FLT channels com-
tical ADC boards. On-board FPGAs allow local patible with an UHECR shower track. A SLT
data analysis, based on a trigger level system de- rate of much less than 1 Hz is expected from ran-
veloped to select candidate events. The trigger dom coincidences.
system is based on the same principles as the one High level trigger conditions have been also
used for the fluorescence detector of the Pierre developed to deal with the different background
Auger Observatory [7]. conditions. For example, the threshold regulation
A first level trigger (FLT) at the channel level, and data acquisition are inhibited when SLT rate
has been designed to identify peaks in the FADC is higher than a certain limit.
trace. The FLT information of all channels is then The DAQ system is continuosly monitoring the
3
3. CONCLUSIONS
The MIDAS experiment has been assembled
and commissioned between October 2009 and
February 2010 at the University of Chicago. The
first period of data taking has been focused on
the understanding of the detector and of the noise
conditions. The sensitivity of the prototype was
found to be adequate for detection of UHECR.
Some improvements are being implemented to in-
crease the duty cycle to close to 100%. In order
Figure 2. Average baseline (black line) of the cen- to crosscheck candidate events with already es-
tral channel of the camera during 5 hours. The tablished detection techniques, we foresee the op-
threshold (red line) follows the baseline structure eration of MIDAS at the southern Auger Obser-
except when the Sun enters in the field of view vatory.
due to the high trigger rate.
REFERENCES
2.2. Calibration 1. R. U. Abbasi et al. [HiRes Collaboration], As-
An absolute calibration of the antenna is pro- trophys. J. 622 (2005) 910.
vided by the Sun, whose emission in the frequency 2. N. Chiba et al. Nucl. Instrum. Meth. A311
of interest is measured every day by several ra- (1992) 338.
diotelescopes around the world. Figure 2 shows 3. J. Abraham et al. [Pierre Auger Collabora-
the signal in ADC counts generated by the Sun tion], Nucl. Instrum. Meth. A 523 (2004) 50.
crossing the field of view of the central channel 4. R. Abbasi et al. [HiRes Collaboration], Phys.
of the camera. The system temperature can then Rev. Lett. 100 (2008) 101101.
be derived from the difference in ADC counts be- 5. J. Abraham et al. [Pierre Auger Collabora-
tween the peak induced by the Sun and the base- tion], Phys. Rev. Lett. 101 (2008) 061101.
line. A patch antenna located at the center of 6. P. W. Gorham et al., Phys. Rev. D78 (2008)
the 4.5 m dish is used to generate RF signals of 032007.
known intensity, thus performing a relative cali- 7. Pierre Auger Collaboration, “Pierre Auger
bration of the individual feed channels. Also, it Project Technical Design Report”.
allows the study of the relative timing, which was
found to be well below the 50 ns sampling.