Dynamic Range of the Picoframe Camera

The dynamic range of a UV/X-ray-sensitive framing camera is of prime importance when the system is required to resolve weak luminous events in close proximity to intense ones, as in the case of a laser produced plasma. Typical X-ray photon liberation from a laser produced plasma may be of the order of 10^2 photons per joule of incident laser radiation for a solid target. This photon flux may be liberated (in

a temporal period roughly equal to the duration of the laser pulse) from a region similar in size to the focal spot, while weak 'coronal' X-ray emission from the plasma edge may yield detailed information on the plasma evolution. It is therefore essential that the dynamic range of the camera be known before being used as a diagnostic under these conditions.

The electron-optical image framing camera operating in conjunction with a high gain intensifier such as the Mullard type XX1330A (50/40) with a maximum gain of -2000 and Ilford HP5 photographic film combination ensures the ability to record single photoelectrons emitted from the photocathode. Under these conditions the

minimum photon flux density arriving at the photocathode must be sufficient to liberate sufficient photoelectrons within the frame exposure time in order to define the required resolution element. The number of photoelectrons required is independent of the type and sensitivity of the photocathode. Providing there is no loss of photoelectrons liberated from the photocathode (not entirely true as the mesh alone intercepts some 30%) as a first approximation, we may evaluate the required photoelectron density at the cathode.

The signal-to-noise ratio (SNR) of the system may be defined as SNR = | - (2.5)

where S is the average signal (photoelectrons emitted firom the photocathode) per pixel or resolution element and is the total root-mean-square noise appearing at the phosphor screen of the image intensifier. If we assume that the intensifier is noise free (reasonable if operated at usual moderate gains) and the complete system is capable of single photon detection, then the only noise within the system is statistical noise of the photons arriving at the photocathode. Thus

SNR = / s ' (2.6)

If the SNR is arbitrarily (but usually) chosen as 10 then we must have at least 100 photoelectrons emitted per resolution element from the photocathode. Given that a gold X-ray-sensitive photocathode has a typical quantum efficiency of 5% (ie.l electron emitted per 20 incident photons) [13] for a photon energy of ~1.5keV, the total photon flux density required is -4000 per resolution element.

The maximum current density has been found empirically to be -7x10-2 A/cm^ at the photocathode [14], limited by the space charge effects within the image tube which cause electron packet spatial and temporal distortions. This evaluation was undertaken on a standard electron-optical geometry of a Photochron II type streak camera but it is believed that the results are generally applicable to the electron-optical design of the

Picoframe type geometry. Thus the dynamic range (or more precisely the contrast ratio) of the camera may be estimated from

Dynamic Range = _{I .}

rmn

where and are the maximum and minimum current densities at the

photocathode. 1 ^ may be calculated knowing the cathode area (6mmx6mm), the frame exposure time (lOOps FWHM) and the desired spatial resolution (10 Ip/mm in both directions) to be -6.5x10'^ A/cm^ yielding a dynamic range of -10. This dynamic range can increase to -40 if the resolution restriction is relaxed to 5 Ip/mm. This calculation defines the dynamic range of the image tube if the entire cathode area is illuminated (it is assumed that little degradation in resolution occurs due to finite photocathode resistance due to this being negligible when using UV and X-ray-sensitive gold photocathodes) which is not the case if a USAF resolution test chart is employed. With this chart only some 25% of the cathode area is actually illuminated which would infer an increase in dynamic range to about 40 at a spatial resolution of 10 Ip/mm in both directions.

The effect of image definition due to photon statistics was demonstrated

practically by operating the framing camera in Static' mode under pulsed illumination conditions. The illumination was derived from a Q-switched and mode-locked frequency-quadrupled Nd: YAG laser employed during the dynamic testing of the

UV-sensitive system (chapter 2). This system delivered up to 2mJ of radiation at 266nm in 60ps FWHM. Figures 1, (a) to (e) show the static, pulsed images recorded when successively decreasing the laser radiation intensity and increasing the intensifier gain to maintain the film exposure at a (roughly) constant level.

HI

m

tus

(c) (d)

T* 5^! j

;y ■{] - , ..’-w. 4:

(e)

Figure 1, (a) to (e) Loss of spatial contrast due to progressive reduction in optical input intensity to the Picoframe camera. Intensifier gains were 50, 100, 200, 400, and 1000 for (a) to (e) respectively.

The corresponding intensifier gains employed were approximately 50, 100,200,400 and 1000 for images a toe respectively. Assuming that in dynamic operation, the electron packet axial dispersion due to Coulomb repulsion may be ignored due to the compensation deflectors operation, the pulsed static dynamic range should be

comparable to the operational dynamic range. Analysis of these images yields that at an intensifier gain of 100 the 10 Ip/mm resolution elements are not quite resolvable by eye (although the 8 Ip/mm elements definitely are). The 10 Ip/mm elements are easüy resolvable by eye at an intensifier gain of 50 however. Further increase in the photon flux density arriving at the photocathode lead to the 10 Ip/mm resolution elements becoming unresolvable and image distortion became noticeable,but it was not possible to record this effect due to the photographic film becoming overexposed even when operating the image intensifier at minimum gain (50). This then allowed an

experimental value of the dynamic range to be deduced to have an upper limit of 50 for a resolution of 10 Ip/mm. This figure represents the experimental dynamic range with a signal to noise ratio of -10 due to the ability to resolve -5% modulation by eye [15].

The 'corrected' theoretical dynamic range to be expected when measuring the dynamic spatial resolution of the Picoframe camera with the USAF resolution test chart agrees most favourably with the experimentally arrived at value.

It may be concluded from the above analysis that the camera should be operated in conjunction with an intensifier with a gain no greater then 100 in order that the dynamic range and spatial resolution be optimised.

Calculations based upon this idea have led to a predicted minimum photon density of 3x10^ photons per 36mm2 (cathode area). This figure implies a pulse energy of -200pJ at 266nm in lOOps (the frametime). While the laser employed is quite readily able to provide this energy level, it must be realised that the cathode should be evenly illuminated, requiring the use of a diffuser and so much of the energy is lost.

Photoelectrons 'lost' within the image tube itself due to mesh absorption, intersection with apertures due to high initial emission energies and angles will exacerbate the problem, but sufficient election densities are quite easily achievable at the phosphor within the frame-times achieved. Shorter fiame-times may prove problematical for this very reason although at present the parallel plate deflectors will limit the maximum usable sweep speed due to fringing fields (see chapter 3.7.1).