Comparison of Vacuum to Solid State Camera Technologies

Thursday, April 26th, 2018 - Light, Photoconductors, Photovoltaic

Comparison of Vacuum to Solid State Camera Technologies

To carry out a comprehensive comparison of the vacuum versus solid state technology one would also need to include a comparison of the specifications for the various cameras available on the market, since there can be considerable performance variations between similar cameras made by different manufacturers. An in-depth comparison can therefore only be carried out by referring to the manufacturers data sheets and data manuals, some of which have been listed as further reading material at the end of the chapter. General guidelines from this comparison can, however, be drawn and are reported in this paragraph.

Vacuum cameras can suffer from geometric distortion due to the non-linearities in the electromagnetic scanning; solid state cameras exhibit a better geometric stability and linearity than vacuum ones and are therefore better suited for optical dimensional measurements, such as in the case of automatic arc welding control [Meta Machines 1985, Edling 1986].

The wavelength response of the two technologies is shown in Figure 1 (a) and 1 (b) from which it can be seen that solid state cameras are more sensitive to longer wavelength (such as infrared light) than Newvicons and Chalnicons but much more so than the most popular vacuum cameras, namely the Vidicon and the Plumbicon. I.R. sensitivity can be a drawback in terms of image focusing and resolution (since the camera optics are affected by the wavelength), and might therefore require the use of an l.R.filter, but can also be an advantage if non-visible illumination is required, such as in the case of imaging objects moving on a conveyer belt where strobed I.R. illumination may need to be used in order to avoid creating unpleasant conditions for factory workers.

Comparison of Vacuum to Solid State Camera Technologies

Figure 1 Typical spectral response of some vacuum and solid state cameras

Solid state cameras have a very low image lag and can therefore be used in higher speed applications than vacuum cameras. They also allow a greater flexibility in vision system design since solid state cameras come in different geometries (1-0 and 2-0 array devices are available) and with different resolutions (64 to 4096 pixels is a typically resolution range for linescan cameras). These features can be used to help improve specific vision system parameters, such as system cost or image processing complexity and speed, as demonstrated in a food processing plant where the use of medium resolution, linescan cameras enables the high speed inspection of corn cobs (Hollingum, 1984).

The smaller size and more rugged construction of solid state cameras, afforded by the absence of a vacuum envelope, also allows a wider range of applications compared to vacuum cameras, such as in the case of eye-in­ hand vision sensors (Pugh, 1982; Loughlin and Hudson, 1982; Van de Stadt, 1985).

Solid state cameras have a higher sensitivity than the Vidicon but, typically, a lower one than the Newvicon and Chalnicon. Solid state cameras do, however, have a less homogeneous sensitivity across the photosite due to the photosite physical size tolerances. A typical CCO photosite is shown in Figure 2; its dimensional tolerances are generated mainly by inaccuracies in the electron beam lithographic masks used during the device manufacture and yield a variable photosite active area which, in turn, produce variations in the individual photosite sensitivity: in the CCO case, for instance, the worst case photosite size error is ±0.065 J.lm for 1024 pixels at 13 J.lm pitch (Fairchild, 1984) which, as shown by eqn below, produces a photo site sensitivity error ΔSp of ±0.5% :photo site sensitivity errorVidicons, on the other hand, have a variable sensitivity which can bechanged by adjusting the target voltage (other devices have a fixed sensitivity) and a higher dynamic range; this latter, however, is afforded largely by the gamma factor non-linearity and the lower sensitivity and might not always be an advantage.

Typical photoelement (or photosite) dimensions

Figure 2 Typical photoelement (or photosite) dimensions for the CCDll1 linescan camera (courtesy of Fairchild, 1984)

Electrical noise tends to be higher in vacuum devices because of the  thermoionic, and therefore random, nature of the electron emission from the vacuum cameras cathode. Some noise is, however, also generated within CCD cameras as a consequence of the cross-talk that takes place between contigous photosites and during the charge transfers.

Ultimately cost may also playa part in the device selection even though the camera cost is typically only a small proportion of the overall vision system cost. The Vidicon is generally the cheapest area scan camera on the market (with the exception of DRAM cameras which however have a worse performance and do not find large use in the industrial robotics field) due mainly to the large number of devices manufactured for the television market, whereas the Plumbicon tends to be the most expensive device among those described. It should be noted that solid state linescan cameras are, in fact, cheaper than Vidicons but that a direct price comparison is somewhat meaningless in view of their different array geometries and specifications.

I hope this information about “Comparison of Vacuum to Solid State Camera Technologies” is easy to be understood.