Image Intensifiers and Image Converters

Image intensifiers of generation I were realized based on vacuum tubes of different kinds:

• Devices with proximity focus essentially consist of a photocathode and a phosphor screen in close proximity (e.g. in less than 1 mm distance). One can apply a high voltage between photocathode and screen, so that photoelectrons are accelerated towards the phosphor. As a result, for each photoelectron one obtains a substantial number of electrons emitted by the phosphor. Because of the small distance, the photoelectrons produced on one point of the photocathode are sent to a tiny area just opposite to that point. A limiting factor is that the voltage cannot be particularly high, because otherwise one would get electrical breakdown due to the small electrode distance.
• Substantially higher voltages and therefore higher light output of the screen becomes feasible by using a substantially larger distance between photocathode and screen. One can then use suitably shaped electrode structures to obtain accurate imaging of photoelectrons to the corresponding points on the screen. With such electrostatic focus, one can also obtain some amount of image magnification.
• An alternative approach is magnetic focus, where the imaging is obtained with a strong magnetic field.

Although no electron multiplication process is employed, there can be substantial image intensification (e.g. 30 times) because of the high quantum yield of the phosphor, which is obtained for high electron energies. For higher intensification, to all three such stages can be used in a sequence. Such devices get relatively long; they have been used mostly in military vehicles.

Such devices have been largely replaced with those based on microchannel plates, as described in the following. However, some of those devices are still used in applications where a lower amount of image intensification is sufficient.

A microchannel plate is a compact type of multi-channel electron multiplier, which is well suited for image intensifiers and was introduced in their generation II. Because moderate electron energies of a few hundred electron volts are sufficient for the input of such a plate, proximity focus as explained above can be applied. After the microchannel plate, a higher voltage of the order of 5 kV is applied for sending the output electrons to the metal-coated phosphor screen.

The image on the screen may be directly observed with the eye or can be sent to an image sensor, e.g. of CCD type – either directly or with some further imaging to adapt the image to the sensor size. In case of direct viewing, one may need to apply an image inversion in order to compensate for the inversion from the objective. For such purposes, one can use traditional lens optics or alternatively a fiber bundle or a fiber-optic plate. While fiber bundles lead to a lower spatial resolution, they allow for a more compact setup. For adapting to an image sensor size which differs from that of the optical output of the MCP, one can use a tapered fiber bundle. The achievable coupling efficiencies with lens optics (with a small f-number) or with fiber bundles are roughly comparable.

The initially used photocathodes (e.g. S25) had a relatively low quantum efficiency and quite limited lifetime. In generation III image intensifiers, one has introduced gallium arsenide (GaAs) photocathodes with much better performance.

Also, an additional ion barrier film (a thin metal-oxide coating, e.g. Al₂O₃) on the input side of the microchannel plate has been applied, which protects the photocathode against such ion bombardment. Unfortunately, it also somewhat degrades the quantum efficiency and thus introduces additional image noise. On the other hand, it eliminates image noise caused by ions.

The achieved lifetimes where now of the order of 15,000 hours, compared with a few thousand hours of older generations.

In later versions, one has reduced the thickness of the ion barrier to a minimum and still achieved a relatively long lifetime with other improvements, in particular with auto gating of the photocathode: the photocathode voltage is automatically switched on only for limited times, depending on the illumination conditions. One therefore avoids to generate any unnecessary photocurrents and potentials for attracting positive ions. This also allows for a accurate vision even with strongly varying illumination levels.

The spectral response is determined by the photocathode in combination with the optical window. Frequently, the long-wavelength limit is determined by the photocathode, while absorption in the optical window sets the shot-wavelength limit.

Note that for some applications it is vital not only to have a high responsivity but also a high quantum efficiency. This is particularly the case where photon counting is required: the quantum efficiency then determines how many of the photons are lost. The quantum efficiency may also be degraded by the optical window and by the microchannel plate: the latter may not react to every photon sent by the photocathode.

The overall amount of image intensification depends on the product of the following factors:

• the responsivity of the photocathode
• the gain of the microchannel plate
• the efficiency of the phosphor, which also depends on the electron energy at that point

For minimum noise, one requires a photocathode with high quantum efficiency and a microchannel plate with low-noise electron amplification. For the latter, the operation voltage should not be chosen to high, although that would increase the amount of intensification.

The dynamic range – the ratio of maximum to minimum input signal intensities – of an image intensifier should be considered in two different variants:

• Within a single image frame, the dynamic range is limited by the used image sensor. This is often of the order of 20 dB to 30 dB.
• The inter-scene dynamic range can be far higher – even well above 100 dB. This is because a number of different factors can be exploited: a variable f-number of the objective, the dynamic range of the image sensor, the electron gain of the multichannel plate (adjusted through the operation voltage) and the exposure time (through the duty cycle with time gating). Therefore, one can make imaging devices which can work in a very wide range of illumination conditions, e.g. from star light to daylight.

The amplified electron image is generated with a very high bandwidth, which is normally limited only by some spread of electron transit time (often well below 100 ps). However, the light emission of a phosphor occurs over a much longer time after irradiation – sometimes tens to hundreds of milliseconds. Another limiting factor is the frame rate of an image sensor.

For high-speed imaging (with high frame rate), one may use an image sensor with electron-beam input, not requiring a phosphor.

For high speed imaging, it can be important to have a fast gating capability. That is usually achieved by switching the voltage between the photocathode and the microchannel plate. Typical switching times are a few nanoseconds. That is of course far lower than the minimum possible frame period.

The spatial image resolution can be limited by different factors:

• the quality of the imaging objective
• the resolution of the microchannel plate
• the details of the image transfer e.g. from the phosphor to the image sensor (e.g. with lens optics, a fiber-optic plate or a fiber bundle)

The achieved resolution varies substantially between different devices.

The electrical power consumption of an image intensifier device can be quite low. A set of not too heavy batteries, powering a high-voltage source, may last for several days of uninterrupted operation. This is because the strip current is usually not that large, and one can use an efficient DC-to-DC voltage converter to produce the high-voltage.

A certain amount of performance degradation can occur during operation of an image intensifier. One factor is the degradation of the photocathode by bombardment with positive ions, which can be generated from gas molecules for a non-perfect vacuum, combined with ionization when they are hit by fast electrons. The positive ions are attracted by the photocathode and can introduce some damage, which gradually decreases the performance. That effect, however, can be largely suppressed with an ion barrier on the input side of the MCP.

Rapid photocathode degradation can occur under conditions of bright illumination. Particularly some older devices would even quickly be damaged under such conditions. Modern image intensifiers are usually much less sensitive to that, because they use auto gating and other measures for adjusting to a wide range of illumination conditions.

There is also some degradation of the microchannel plate during operation. It does not simply depend on the operation time, but basically on the accumulated electrical output charge: aging is substantially faster with input light than without.

Here you can submit questions and comments. As far as they get accepted by the author, they will appear above this paragraph together with the author’s answer. The author will decide on acceptance based on certain criteria. Essentially, the issue must be of sufficiently broad interest.

Please do not enter personal data here; we would otherwise delete it soon. (See also our privacy declaration.) If you wish to receive personal feedback or consultancy from the author, please contact him e.g. via e-mail.