GB2094548A

GB2094548A - A thermal imaging system

Info

Publication number: GB2094548A
Application number: GB8107605A
Authority: GB
Original assignee: UK Secretary of State for Defence
Current assignee: UK Secretary of State for Defence
Priority date: 1981-03-11
Filing date: 1981-03-11
Publication date: 1982-09-15
Also published as: IL62571A; JPH0241178B2; FR2501942B1; JPS57160158A; IL62571A0; GB2094548B; FR2501942A1

Abstract

A thermal imaging system including a biassed elongate detector element (3) of photoconductive material, over which an image of a thermal scene is scanned at a velocity that is matched to the ambipolar drift velocity of photocarriers generated in the element (3). In order to improve responsivity and detectivity the length of the detector element (3) or the magnitudes of bias and scan velocity, are selected so that the time taken to scan the detector element (3) from one end (5) to a read-out region (9) of the detector element (3) is greater than the lifetime of the photocarriers generated in the element (3). In order to avoid loss of resolution by photocarrier diffusion the photocarrier lifetime of the detector material is of necessity of relatively low value. In the example described the detector is of cadmium mercury telluride material sensitive to infra-red radiation in the 8 to 14 mu m band and has a characteristic photocarrier lifetime of around 2 mu s. The system may include several detector elements (3) arranged in parallel. <IMAGE>

Description

SPECIFICATION A thermal imaging system and a method of operating the same The present invention concerns a thermal imaging system and a method for its operation; in particular a system of the kind comprising: an elongate detector element of photoconductive material; a bias supply connected across the detector; a read-out circuit responsive to the detector; an optical assembly for focussing an image of a thermal scene on to the surface of the detector; and, a scan-mechanism for scanning the image along the length of the detector element; wherein for appropriate bias and scan velocity, photocarriers generated in the element are caused to drift along the element at an ambipolar drift velocity that is matched to the scan velocity.

A system of this kind is described in GB Patent No 1,488,258 (US Patent No 3,995,159), the contents of which are here imported by way of reference. As therein described the detector element is a strip of cadmium mercury telluride material having bias contacts at each end, and a read-out region provided between the bias contacts. Two forms of read-out region are described. One form, a passive read-out, comprises a pair of conductors in contact with the strip, spaced a short distance apart along the length of the strip. One of these conductors may be provided by the adjacent bias contact (the second of the two bias contacts). The associated read-out circuit includes a high impedance preamplifier and is connected across the two conductors.This circuit produces an output signal dependent on the voltage between the two conductors, a voltage which is modulated as the resistivity of the region between the conductors changes with variation of the local photocarrier density. The output signal provides an analogue representation of the thermal scene.The read-out region geometry (ie both the conductor spacing "a" and the strip width "w") is chosen as a compromise between signal level and spatial resolution, and depends on the extent of photocarrier diffusion (2art a r 2#2 and w -- 2n where AN/(Dt),tST "D': being the ambipolar coefficient for carrier diffusion at the operating teperature, "t" the time taken to scan an image along the detector element from the first bias contact to the read-out region, and "z" the lifetime of the photocarriers.

The alternative form of read-out, an active readout, comprises a p-n junction formed in the strip.

The associated read-out circuit includes a discharge impedance and a pre-amplifier connected to measure the discharge current. The read-out region and its circuit serve to sweep out photocarriers; an output signal providing analogue representation of the thermal scene is thus provided. In this case the time constant "T" of the read-out circuit is chosen to optimise spatial resolution:~ vTt2A.

v being the drift velocity of the photocarriers.

In the system described therein the spacing between the first bias contact and the read-out region is chosen so that the maximum transit time of the photocarriers is less than the photocarrier lifetime. In fact for the example particularly described, for material exhibiting a relatively long lifetime ( 10 ,us), the loss of resolution arising from photocarrier diffusion is limited by restricting this spacing to a value significantly less than the distance that would be travelled by photocarriers in a lifetime::~ 1 viz in fact 1 ~ 0.5 vz According to the present invention the system is characterised by detector material of relatively short photocarrier lifetime, and a spacing between the first bias contact and the read-out region of the detector element such that for the appropriate bias and scan velocity, the time taken to scan the distance between the first bias contact and the read-out region is greater than the photocarrier lifetime.

This system may, for example, include a plurality of detector elements arranged parallel to each other in the focal plane of the optical assembly, and each having a bias supply and a read-out circuit.

According to another aspect of the invention, a method of operating a thermal imaging system includes selecting the bias and the appropriate scan velocity such that the distance between the bias contact at one end of the detector element and the read-out region remote therefrom, is scanned in a time greater than the photocarrier lifetime.

Quite contrary to GB Patent No 1,488,258, it is found for material exhibiting relatively short photocarrier lifetime that it is not at all necessary to limit the bias contact to read-out spacing to give a short time scan time (ie t < T) With the scan time in excess of the photocarrier lifetime as here, the diffusive spread of photocarriers remains relatively constant:~

With material exhibiting relatively short photocarrier lifetime, this extend of difussion and the resultant limitation in spatial resolution is acceptable for imaging applications. Furthermore, it is in fact advantageous to use a scan-time greater than the photocarrier lifetime. Although photocarriers are lost by recombination, the photocarrier density does not fall.Instead, it continues to increase with continued exposure of the detector material approaching an equilibrium level where optical generation balances recombination. Both read-out signal level and signal-over-noise are improved as a result of the added time that is afforded for signal integration in the detector element.

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings of which:~ FIGURE 1: is a diagram partly in perspective form, and partly in circuit form of a single element detector and bias supply; FIGURE 2: is a plan schematic of a thermal imaging system showing the optical assembly and scan-mechanism in particular; FIGURE 3: is a graph showing the variation of detector responsivity as a function of detector length; and, FIGURE 4: is a plan view of detector comprising several parallel elements for use in a system such as is shown in Figure 2 above.

There is shown in Figure 1 a biassed detector 1 comprising a strip 3 of photoconductive material provided with ohmic end contacts 5 and 7 one at each end, and a passive read-out region 9 between these contacts 5 and 7. This read-out region is formed by one of the end contacts 7 and an additional contact 11, also ohmic, spaced from the end contact 7 by a short distance.

A high impedance pre-amplifier, part of a read-out output circuit 1#3 is connected between the two read-out contacts 7 and 11. The detector 1 is biassed by means of a constant current supply source 1 5.

In detail, the strip 3 is of instrinsic n-type cadmium mercury telluride material (CMT):- CdxHg x= 0.208 exhibiting a net donor concentration n of approximately 5 x 1014 cam~3. At liquid nitrogen temperature, CMT material of this composition has a long wavelength infra-red response cut-off in the 8 to 14 #m of the spectrum at around 1 1.5 to 12 m and thus will respond to radiation in this band at least up to this cut-off value. The strip 3 is 8 form thick (d), 62 ym wide (w) and the length (1) 700 Flm between the end contacts 5 and 7.The resistance of the strip is around 500 Q. It is mounted on a sapphire substrate (not shown) and the CMT surfaces are passivated with an anodic oxide. The contacts 5, 7 and 1 1 are of gold metal formed by evaporation. Under operating conditions, the material exhibits a photocarrier lifetime of about 2 ,us, a relatively short lifetime value. This corresponds to a diffusion length "A"' of around 25 'ism. To match the diffusive spread of photocarriers, to give compromise signal level and spatial resolution, the additional contact 1 1 is spaced a distance 50 ym from the nearest end contact 7.

The detector is located at the focus of an optical assembly, as shown in Figure 2. The optical assembly comprises a first pair of infra-red transmitting lenses 21 and 23 arranged to collimate and direct radiation towards a third lens 25. The detector 1 is located in the focal plane of this third lens 25 and is shielded from much of the background radiation by means of a cold shield 27, also part of the optical assembly. The detector lens 25 has an F-number (ie the ratio of focal length to diameter) of 3 and the aperture of the cold shield is matched to this lens F-number. A Joule-Thomson cooler is used to maintain the shield 27 and the detector 1 at liquid nitrogen temperature (770K).

It is noted that both the ambipolar mobility lua and the excess carrier lifetime T depend on electron and hole densities n and p respectively~ K = (n - p) ,u,#u,l(nEL, + p,up) and when electron-electron Auger recombination predominates n2 T 2TA1 n(n + p) where 'rAi is the intrinsic Auger lifetime and nj is the intrinsic carrier density.

The electron and hole densities depend somewhat on the background radiation flux.

Choosing a system F-number in the range 2.5 to 3 helps to limit the background flux and to achieve a sufficiently high mobility and lifetime, both factors on which optimum resolution depends. The material described, exhibits an ambipolar mobility of around 450 cm2v-s-1 and an excess carrier lifetime of around 2 ys for a typical background flux of 2 x 1016cm2s-' (ie the infra-red spectral band up to the cut-off).

A scan mechanism is interposed between the second and third lenses 23 and 25. This comprises an octagonal drum reflector 29 and a flapping mirror 31. As the drum reflector 29 is rotated, the image of the distant thermal scene focussed onto the surfaceof the detector 1 is scanned along the length of the detector strip element 3. When the bias current flowing through the strip 3 is adjusted so that photocarriers are driven towards the read-out region 9 at the same velocity as the image, the output signal from the circuit 13 provides a line signal, ie a time dependent signal representative of a strip of the thermal scene. As the drum 29 rotates and each of its faces 33 intersects the beam of collimated radiation, a new line scan is initiated. The flapping mirror 31 is advanced each time so that as each line scan is initiated, a different strip of the thermal scene is imaged. The thermal scene is thus scanned line by line until a frame is completed, a new frame is then initiated, and so forth.

In the system thus far described, the bias current is adjusted to give a bias field of 29 V cm-~ and the scan velocity chosen to match the ambipolar drift is approximately 1.3 x 104cm s-'.

The time taken to scan the strip from the bias contact 5 at one end to the read-out region 9 is thus 700 #m s 1.3 x 104cm s-' ~ 5.4 'its.

Thus in this example the scan time exceeds the lifetime by a factor of 2.7 or thereabouts. At this scanning rate the detector provides data at a pixel rate (ie V/w) of about 2.1 MHz. Under these operation conditions the responsivity and detectivity measured with 5000K blackbody radiation have been found to be around 1.5 x 105VW-1 and 1.7 x 10" cm HzW-' respectively.

Alternatively, operating the detector with a bias field of 68V/cm and scanning at a matched velocity of 3.1 x 104cm/s, a responsivity of 2.5 x 105 V/W-1 and a detectivity of 2.0 x 10" cm H > W-' have been measured. In this example the scan time exceeds the lifetime by a factor of 1.1 or thereabouts.

It is found that the responsivity R (defined as the voltage output produced by a radiation density of 1 watt per width squared incident on the detector) is given by the formula~ w7Tav R = [ 1--exp-exp(-1,v#) ] E##2dn'ia where 77 is the quantum efficiency for converting infra-red photons to photocarriers, for radiation in the spectral band of width EA up to cut off wavelength, Tthe carrier lifetime, I, w and d the dimensions of the strip; "a" the read-out spacing, jtta the ambipolar mobility, n the net donor concentration, and v the scan velocity. A quantum efficiency of 0.6 is typical.This can be further improved by using an anti-reflection coating, for example by spattering a thin layer of zinc sulphide on to the cadmium mercury telluride surface and over the anodic coating. As can be seen from figure 3, the responsivity of the detector at a specified scan velocity increases with the length of detector provided. The detectivity also is improved if a detector of long length is used:- D* cc [ 1 - exp (-1,v) ] + where the detectivity D* is a measure of the signal-to noise performance and in this case is defined as the signal-to-noise ratio in unit bandwidth, when a radiation density of 1 watt per width squared is incident on the detector, multiplied by the width.

Whilst better signal level (eg responsivity) and signal above noise (eg detectivity) is achieved by using long detectors, thus high I, is accompanied by a higher power dissipation. Ultimately the choice is governed by practical bias supply and temperature control limitations, and the choice of length I would be a compromise between these competing considerations. Where relatively high power dissipation can be tolerated, a detector having a length up to 5 to 6 times the value VT would represent a practical compromise, eg the detector described in the first example may be increased in length to say 1400 ,um giving a scantime of 10.8 ys with consequent increase in R and D*.

It is noted that the passive read-out provided by contacts 11 and 7 may be replaced by an active read-out. In this case the contact 11 is replaced by a diode implanted or diffused in the strip in the vicinity of the bias contact 7. As known this may be discharged to earth through a discharge resistor and the current flowing in this discharge resistor measured by means of a high impedance pre-amplifier. The combination of this pre-amplifier and the discharge resistor replacing the output circuit 13 referred to above. As already set forth above the time constant of this circuit, which is here defined by the bandwidth of the pre-amplifier, is set to match the diffusion of the photocarriers, to optimise resolution.

The detector used in the system of Figure 2 may have a parallel content. Thus the detector shown in Figure 4 comprises an array of eight parallel elements 31 to 38. Each element has the same dimensions and is of the same material as the single element described above. This detector is positioned in the focal plane of the same lens 25 (F-number 3) of the system of Figure 2 above.

Each element 31 to 38 is mounted on a sapphire substrate and is spaced apart from the next element by a gap of 12.5 'ism wide. Gold metal end contacts 41 to 48 and 51 to 58 are formed at the ends of the elements and having a fanned out configuration to facilitate lead-out bonding.

Additional metal contacts 61 to 68 are also provided at one end to provide read-out. These contacts 61 to 68 and the adjacent end contacts 51 to 58 are connected in pairs 61 and 51,to 68 and 58 each across the high input impedance of a corresponding pre-amplifier (not shown). For this parallel array the frame scan flapping mirror 31 is advanced so that the thermal scene is scanned band by band, eight lines at a time, until a frame of information is derived. The line signals may then subsequently be processed to provide a video signal for TV display.

Claims

1. A thermal imaging system comprising: an elongate detector element (3) of photoconductive material; a bias supply (1 5) connected across the detector; a read-out circuit (13) responsive to the detector; an optical assembly (21 to 27) for focussing an image of a thermal scene onto the surface of the detector; and, a scan mechanism (31, 33) for scanning the image along the length of the detector element (3); wherein for appropriate bias and scan velocity, photocarriers generated in the element (3) are caused to drift along the element (3) at an ambipolar drift velocity that is matched to the scan velocity; charaterised by detector material of relatively short photocarrier lifetime, and a spacing between the first bias contact (5) and the read-out region (9) of the detector element (3) such that for the appropriate bias and scan velocity, the time taken to scan the distance between the first bias contact (5) and the read-out region (9) is greater than the photocarrier lifetime.

2. A system as claimed in Claim 1 wherein the time taken to scan the distance between the first bias contact (5) and the read-out region (9) is greater than the photocarrier lifetime by a factor between 1 and 6.

3. A system as claimed in Claim 2 wherein the time taken to scan the distance between the first bias contact (5) and the read-out region (9) is greater than the photocarrier lifetime by a factor between 1 and 3 inclusive.

4. A system as claimed in Claim 2 wherein the time taken to scan the distance between the first bias contact (5) and the read-out region (9) is greater than the photocarrier lifetime by a factor between 3 and 6 inclusive.

5. A system as claimed in any one of the preceding Claims 1 to 4 wherein the system includes a plurality of detector elements (3) arranged parallel to each other in the focal plane of the optical assembly (21 to 27), and each having a bias supply (15) and a read-out circuit (13).

6. A system as claimed in any one of the preceding Claims wherein the or each detector element (3) is of cadmium mercury telluride material and is sensitive to infra-red radiation in the 8 to 14 ,um band of the spectrum.

7. A method of operating a thermal imaging system including an elongate detector element (3) of photoconductive material including the steps of biassing the detector element (3) and scanning an image of a thermal scene over the surfce of the detector element (3) at a velocity that is matched to the ambipolar drift velocity of photocarriers generated in the element (3) characterised by selection of the bias and the appropriate scan velocity such that the distance between the bias contact (5) at one end of the detector element (3) and the read-out region (9) remote therefrom, is scanned in a time greater than the photocarrier lifetime.

8. A method as claimed in Claim 7 wherein the time is greater than the photocarrier lifetime by a factor between 1 and 6.

9. A method as claimed in Claim 8 wherein the time is greater than the photocarrier lifetime by a factor between 1 and 3 inclusive.

10. A method as claimed in Claim 8 wherein the time is greater than the photocarrier lifetime by a factor between 3 and 6 inclusive.

11. A thermal imaging system constructed, adapted and arranged to operate substantially as hereinbefore described and as shown in Figures 1 and 2 or Figures 4 and 2 of the accompanying drawings.