Quentin A. Kerns

Photomultiplier Single-Electron Statistics

Our measurements of the amplitude distribution of photomultiplier anode pulses due to the emission of single-electrons from the cathode consistently show a peak. It is significant that the peak position agrees with that of a calculated distribution based on a Poisson distribution of secondary electrons at each dynode. The integral distribution, obtained by counting single-electron pulses, tends to show a plateau. In low-light-level counting applications, one can set the discriminator so that a majority of photomultiplier single-electron pulses will be counted. Further increase in the sensitivity will eventually increase the noise rate faster than the counting efficiency. The techniques for measuring photomultiplier single-electron statistics are useful for obtaining comparative collection efficiencies. By single-electron measurements, one can adjust focusing-electrode potentials to maximize overall collection efficiency. It is believed that there is some correlation be. tween the amplitude and time distributions. Further work is necessary to show the extent of the correlation.

GENERATOR OF NANOSECOND LIGHT PULSES FOR PHOTOTUBE TESTING

This mercury‐capsule light‐pulse generator was developed to test and evaluate the high‐speed features of multiplier phototubes and low‐level image tubes. Light pulses and electrical trigger pulses are generated simultaneously in an arc discharge at a usual repetition rate of 60 per second. The electrical pulse is used as a time reference for the light pulse. The time required for both the light and the electrical pulse to rise from 10 to 90% of peak amplitude is less than 5×10−10 sec. The light pulse rises to a maximum and falls to 50% peak amplitude in less than 1.5×10−9 sec. The electrical pulse power available is large. Time resolution of 10−10 sec is typical of measurements made with conventional fast oscilloscopes, while elaboration of technique permits relative time measurements that are better in some cases by at least three orders of magnitude. The light is emitted from a region a few mils in diameter, and thus may often be considered to come from a point source. An S4 photosurface subtending 0.1 ...

Evaluation of the C-70045A High-Speed Photomultiplier

Measurements were made of the time spread, pulse response, and other characteristics of the C-70045A 14stage photomultiplier which was developed for high-speed timing applications with scintillation and Cerenkov radiation detection. Rise times and time jitters are given. Some measurements of the scintillation decay time of two plastic scintillators are reported. (D.L.C.)

Pulse shaping and standardizing of photomultiplier signals for optimum timing information using tunnel diodes

Abstract Accurate timing signals have been derived from fast photomultiplier pulses by differentiating the photomultiplier pulse in order to produce a zero-crossing signal whose time of zero crossing is fixed over a large dynamic range of light level. Detection of the zero-crossing point is accomplished with a balanced-bridge tunnel-diode discriminator, which produces a standardized output of 100-mV amplitude, and a half-width of 4 nsec. A balanced bridge is used to prevent the zero-crossing signal from appearing at the output. For a twenty-to-one range of photomultiplier-input light level, the time shift of the output pulse may be as low as 0.5 nsec with a change in the output amplitude of less than a factor of two. When this discriminator is used with a 14-stage photomultiplier, the sensitivity is great enough to allow direct operation from single photo-electrons. The above numbers are an indication of what one can obtain with the tunnel diode and transistor combination chosen at the time of the original development almost two years ago. Since that time, particularly in the tunnel-diode field, numerous improved devices have become available. We are currently evaluating these in zero-crossing circuits. The zero-crossing photomultiplier pulse has been generated in two ways. The first method uses an overdamped LC -tuned circuit built into the base of a 6810-A photomultiplier. Here the time of zero crossing is controlled by the frequency of the LC -tuned circuit. The decay time of the scintillation plastic used and the rise time of the photomultiplier determine what this zero-crossing time should be. A clipping stub has also been used to produce the zero crossing remotely from the photomultiplier. By this method, one can control the zero-crossing time by both the length and impedance of the clipping stub.

A triggered nanosecond light source

Abstract A nanosecond light source based on a corona discharge started by field emission has been developed for attachment to scintillation counters. These individual light sources, placed to illuminate each phototube and fired by a pulse generator through splitting transformers and cable delays, have facilitated the coincidence timing of an array of photomultiplier detectors. Pulse generation, a pulse-splitting transformer, light-source construction, and timing properties, as well as spectral intensity are discussed. An optical attenuator designed for the source is described.

A SPARK-GAP TRIGGER SYSTEM

The construction and operation of a trigger system designed to fire a 30-kV 5000 A spark gap with a minimum delay following the arrival of a small signal pulse is described. In this particular experiment a 150-MeV/c muon is detected with scintillators on three 6199 phototubes, and the output pulse of the attached tunnel-diode triple-coincidence circuit is amplified and used to trigger the gap. Approximately 32 nanoseconds are needed from passage of the muon to the coincidence output, and approximately 25 nanoseconds are required from the coincidence output to the time of complete breakdown of the gap. These delays represent the shortest times that we could achieve with the particular boundary conditions under which the circuit had to operate. Sufficient detail is given to show how additional savings of nanoseconds could be made under different operating conditions.

Electromagnetic Deflector for the Beam of the 184‐Inch Cyclotron

The removal of the ion beam of 190‐Mev deuterons from the magnetic field of the 184‐inch cyclotron is complicated by the fact that at large radii the ion increases its radius very little in one revolution. The usual type of deflector did not appear to be usable. Consequently, a pulsed electric deflector was designed that would give an additional radial oscillation to the ions which could be as much as seven centimeters. This permits deflection of the ions outward to a magnetic deflector which is outside the range of the circulating ions. The field is sufficiently reduced inside the magnetic deflector to permit the ions to leave the magnetic field of the 184‐inch cyclotron. Approximately ½ percent of the circulating beam appears in the external beam.

Test routines and monitoring for a large counter experiment

Abstract Some recent multichannel scintillation-counter experiments are so complex that the use of test routines and fault monitoring becomes essential. In a recent experiment, it was desired to initially align the system, provide continuous calibration during the experiment, and rapidly locate the cause of threshold drift or catastrophic failure. A test and monitoring system was devised in which synthesized pulses are injected into several points of the system, and the operation is automatically compared with the desired response. If the system performance does not agree within prescribed limits, an alarm sounds. Through a program that varies the amplitude of the test signal, one can check coincidence-circuit threshold and feed-through characteristics. By programming delay between input pulses, the resolving time can be measured. Several novel monitoring devices have been developed which allow one to assess the display of a large volume of information at a glance. One unit for reading time delays involves the use of a cathode-ray-tube raster display, another unit for reading binary information employs a transistor-driven incandescent lamp panel.

Ferrite Measurements for Synchrotron RF Accelerating System Design

Ferrites provide a reliable and flexible means of electronically tuning particle accelerator rf cavities without wasting a disproportionate amount of rf power. A systematic design procedure is presented for a ferrite tuned rf resonator. Radio frequency considerations, and use of salient ferrite parameters are the prime considerations of the design example. Using requirements set forth in the rf design, a bias supply design is discussed with respect to optimization of overall system design.

A solid-state chronotron for digitizing time delays☆

Abstract This solid-state chronotron measures the relative time delay between two shaped photomultiplier pulses, and quantizes this time delay into seven time intervals with widths of 3 to 7 nsec. These intervals represent the difference in time of flight of two particles in a large scattering experiment, and may be varied at will by charging front panel cables. The quantized time delay is then stored in a buffer storage until readout, at which time the system is recycled. Because the chronotron was used with a system having a 40-μsec cycle time, a maximum repetition rate of 100 kc/s was selected. However, this rate could be increased considerably if necessary. The time intervals are formed by splitting the two shaped photomultiplier signals into parallel diode coincidence circuits, with the time delay of each coincidence circuit set for the center of its respective time interval. Since the coincidence circuits are identical, both the position and width of each time interval are determined only by the delay cables. Thus the coincidence circuit with the greatest output voltage indicates the correct time interval. To detect this voltage the stretched output of each coincidence circuit has a common ramp voltage added to it, and together these signals begin driving a bank of blocking oscillators. As soon as one blocking oscillator triggers, all others are disabled and the blocking oscillator that triggered sets up its binary coded number in the buffer storage. As the ramp generator always insures that one blocking oscillator triggers there is never a hole between time intervals, even if the edges of the intervals drift, because when one interval expands its neighbors will contract. Long-term stability of the time interval centers is a few tenths of a nanosecond, and the 24-hr stability of the edges of the intervals is ±0.5 nsec. With the aid of a TV monitor used in checking the system, the chronotron can be held within the above limits indefinitely by simple daily adjustments.