FR2735935A1 - CCD for atmospheric lidar - Google Patents

Abstract

The CCD is used for an image zone (20) formed from P lines and N columns of photosensitive sites, with a memory zone (24) having M lines and N columns of non-photosensitive sites. A reader (26) with N sites receives successive lines from the memory. The charge is transferred to an integration zone of N columns at a first rate, f, and accumulated during an acquisition time, T, greater than P/f and to the memory zone at a second rate, F, whose magnitude is of the order of 1/T. The sites in the image zone are less than 30 microns in size parallel to the surface and those in the memory zone are smaller in one dimension than those in the image zone. The acquisition time is less than 10 ms.

Description

LOAD COUPLING DEVICE SUITABLE FOR ACQUISITION
FAST OF A LIGHT SIGNAL, AND ITS
APPLICATION IN A LIDAR
The present invention relates to a charge coupled device for temporally analyzing a light signal, as well as a lidar incorporating such a device.

 A charge coupled device (CCD) according to the invention finds applications in any distance measurement system by reflection of light pulses (lidar) of average spatial resolution (typically at least a hundred meters). In particular, space lidars measuring light pulses reflected on atmospheric particles with a resolution of the order of a kilometer can incorporate a CCD according to the invention.

 A first type of lidar, called atmospheric lidar, makes it possible to determine the position of the clouds from time-of-flight measurements of the pulses emitted and retroreflected by the surface of the clouds. Such a lidar usually has an avalanche photodiode associated with an electronic multiplication device as its detector.

 Wind lidars are a second type of space lidars which, in addition to distance measurement, carry out for each altitude a local wind measurement taking into account the Doppler effect (taking into account the speed of the satellite which is known and the speed of the air mass we are trying to assess). For this, a wind lidar must therefore resolve the wavelength spectrum of the reflected signal. Due to the spread of the spectrum over several wavelength channels, the number of photons captured can be extremely small. Since the measurement accuracy requires a sufficient signal-to-noise ratio, the measurement noise must be almost zero and maximum detectivity.

The need for multi-channel detection condemns the use of avalanche photodiodes. The detectors used must be multi-element detectors possibly associated with an electronic amplification device of the “micro-channel pancakes” type (MCP, “Multi-Channel Plates”).
MCP associated with multiple anodes exist, but they have the major drawback of having dead zones between the anodes, which reduces the sensitivity, and of requiring as many signal processing chains as there are anodes, which makes the system heavy and complex.

 Applications are known where CCDs are used on board space vehicles to perform distance measurements by retroreflection of light pulses. Particularly known are position measurement systems allowing an appointment in space between a hunting vehicle provided with the laser and the detector and a target vehicle provided with retroreflectors (see FR-A-2 602 057). But the context is quite different from that of atmospheric measurements. First, the spatial resolution sought is much finer (of the order of 100 times).

Second, the reflected light flux is much greater and less diffuse, since the target is cooperative, unlike atmospheric particles.

 An object of the present invention is to take advantage of the advantages of CCD detectors in applications, such as in space lidars, where for an average resolution in distance, a high sensitivity to the incident light flux is sought.

 The invention thus proposes a charge coupled device for temporally analyzing a light signal, comprising an image area constituted by a matrix of P rows and N columns of photosensitive sites to which the light signal is directed, a memory area constituted by a matrix M lines and N columns of non-photosensitive sites receiving the charges which appeared in the image zone during an observation period, and a reading register of N sites successively receiving the charges stored in the lines of the memory zone. At the border between the image area and the memory area, the device presents an integration area comprising a line of N sites. During the observation period, the charges appearing in each column of the image area are transferred along this column to a respective site of the integration zone at a first transfer frequency, and accumulated in said site over an acquisition period substantially greater than P times the inverse of the first transfer frequency. During the observation period, the charges accumulated in each site of the integration zone are transferred along a respective column of the memory zone in the direction of the reading register at a second transfer frequency of the order of l 'inverse of the acquisition period.

The CCD thus performs all the real-time acquisition, integration and storage functions which, in traditional photon counting applications, would have been performed by a set of fast and complex detection chains. CCD therefore considerably simplifies the detection system, and also has the following advantages
- very high quantum efficiency of the ultraviolet at near infrared, especially for thinned CCDs
- filling rate of the sensitive surface equal to 100%
- possibility of high frequency charge transfer to medium surface detectors
- possibility of integrating and storing information in the CCD structure without degradation by maintaining the current for generating internal charges at values that do not degrade performance (by cooling and / or using MPP technology (Multi
Pinned Phase))
- reading noise less than a few electrons for average pixel reading frequencies (20 to 100 kHz), or even reduction of this noise to values less than 1 electron by non-destructive methods of multiple reading ("skipping");
- removal of the electronic amplification device upstream of the sensor.

 The spatial resolution of the detector is proportional to the duration of acquisition. If the number M of lines in the memory area is at least equal to the ratio between the duration of the observation period and the acquisition duration, all of the measurements relating to an observation period can be stored in the CCD, then processed subsequently by a system that does not have the same speed constraints.

 A lidar according to the invention comprises a pulse light source such as a laser, and a detection system, comprising a charge coupled device as defined above, arranged to receive reflected echoes of the light pulses emitted by the source. .

In the case of a wind lidar, the detection system further comprises an optical filter arranged to direct spectral components of different wavelengths of the echoes of the light pulses to different columns of the CCD image area. We can then process up to N different wavelength channels.

Other features and advantages of the present invention will appear in the description below of a preferred but nonlimiting embodiment, with reference to the accompanying drawings, in which
- Figure 1 is a block diagram of a lidar according to the present invention
- Figure 2 is a schematic plan view of a
CCD according to the invention
- Figure 3 shows an example timing diagram of the control signals applied to different CCD sites in Figure 2
- Figures 4 and 5 are diagrams illustrating the transfer of charges in the CCD of Figure 2; and
- Figure 6 schematically shows an orbiting satellite equipped with a lidar according to the invention.

 The lidar shown in FIG. 1 comprises a source 10 of light pulses and a detection system comprising a focal plane optics 12 and a CCD sensor 14.

The source 10 is for example constituted by a laser of the type
Nd-YAG emitting pulses of very short duration (typically 20 to 30 ns). The emission wavelength is for example 355 or 532 nm. The pulses have for example an energy of one joule and a repetition rate of 10 Hz.

The optics 12 receives the light pulses emitted by the source 10 and retroreflected on a distant medium, and directs them towards the image area of the CCD 14. A relay optics 18 focuses the light signal received on the image area of the cCD 14, and a interference filter 16 distributes the spectral components of different wavelengths of the received signal to different columns of the image area of the CCD 14. The filter 16 can be of the Fabry-Perot type or of the Fizeau type.

 The CCD 14 has the structure shown schematically in FIG. 2. Its image area 20 is presented as a matrix of P lines and N columns of photosensitive sites on which the received light signal is focused (P = N = 16 in the example shown ). In the direction of charge transfer (from top to bottom in Figure 2), the image area 20 is followed by an integration area 22 formed by a line of N sites, by a memory area 24 formed by a matrix of M rows and N columns of non-photosensitive sites (M = 32 in the example shown), and by a reading register 26 constituted by a row of N sites. Charge transfers in the image area 20, in the area of integration 22, in the memory area 24 and in the read register 26 are controlled by respective signals + ", delivered by a generator 28 of clock signals. The generator 28 is clocked by a sequencer 30 synchronized on the instants of emission of pulses by the source 10.

The charges delivered in series from the reading register 26 are supplied to a reading and amplification stage 32 which formats the measurements for a digital system 34 for interpreting and processing the measurements.

 The sequencer 30 defines after each emission of a pulse in the source 10 an observation period corresponding to a time window during which the reflected echoes are likely to reach the detection system. During this observation period, the CCD 14 is controlled to measure and store the received optical signal. After the end of the observation period, and before the start of the observation period relating to the next laser pulse, the sequencer 30 has the time to command the reading of the content of the register 26.

 The spatial resolution of the CCD is given by an acquisition duration T corresponding to the inverse of the frequency F of charge transfer in the memory area 24.

In the image area 20, the charges are transferred along the columns at a higher frequency f. The integration zone 22 is used to accumulate column by column the charges which have appeared during each acquisition period
T in the image area 20. The charges are therefore transferred from the image area 20 to the integration area 22 at the frequency f, and from the image area 22 to the memory area 24 at the frequency F.

 FIG. 3 shows the control signals, ', "which can be sent to zones 20, 22, 24 to achieve these different transfer frequencies, in an example with four phases per site, indexed A, B, C, D in the direction of load transfer (ie from top to bottom in Figure 2 or from left to right in Figures 4 and 5) Figure 3 shows the control signals over a period of acquisition T. A high level control signal in FIG. 3 corresponds to the presence of a potential well under the corresponding portion of the site, as illustrated in FIGS. 4 and 5.

 In FIGS. 4 and 5, the reference 36 designates the metal mask deposited on the CCD 14 to mask the sites of the integration zone 22, of the image zone 24 and of the reading register 26 from light. . Note however that the integration zone 22 could be partially or totally exposed to light. In the case represented in FIGS. 4 and 5, the integration zone can thus be seen as a first line of the memory zone but, if it were completely exposed to light, it could be considered as the last line of the zone image or, if it was partially exposed to light, as an intermediate zone straddling the image zone and the memory zone.

 An acquisition duration T begins, between the instants marked by t0 'to t4' in FIG. 3 and corresponding respectively to the five lines of FIG. 4, by a transfer cycle along the columns of the memory area 24. During of this same cycle, the charges present in the potential wells 38 of the integration zone 22 are transferred to the first line of the memory zone 24. During the major part of the acquisition time, between the instants marked by t0 , and t0 in FIG. 3, the charges are not transferred in the image area 20 to the integration area 22, the phases vAS 4> E '4> C' B CE D and A remaining constant. During this time, the photons incident on the image zone generate electrons which trap in the potential wells of the image zone. At the end of the acquisition time, the sequencer 30 controls P transfer cycles in the image area to the integration area. The first of these cycles, between instants t0 and t4, is detailed in the diagram in FIG. 5. At the end of the P transfer cycles, all the charges 42 which appeared in a column of the image area 20 were transferred to the corresponding potential well 38 of the integration zone 22. At the start of the next acquisition period, these charges will be transferred to the first line of the memory zone 24.

 It is observed that, during the P transfer cycles at the end of the acquisition period, certain charges can continue to appear in the image zone without reaching the integration zone 22. These charges will be assigned to the following acquisition duration . So that this does not disturb the measurements, it is expected that the transfer frequency f in the image area is substantially (at least twice) greater than P times the inverse of the acquisition time T.

 It will be understood that various CCD structures and various control modes make it possible to obtain different transfer frequencies on either side of the integration zone 22. For example, by retaining a four-phase structure, it would be possible to control transfers in the image area to the integration area without interruption during the entire acquisition period, the charges being then transferred as they appear, instead of waiting for the end of the acquisition period.

 The numerical values of the dimensioning and sequencing of the CCD 14 depend on the application considered.

FIG. 6 shows a satellite 50 in Earth orbit equipped with a lidar according to FIG. 1, pointed at an angle a relative to the nadir. The altitude domain in which it is desired to measure has a height h, and the satellite 50 is itself located at a height H above this domain. Thus, the observation period following the emission of a laser pulse must at least include a time window starting 2H / (c.cosa) after the emission of the pulse and of duration 2h / (c.cosa). For aiming angles a less than approximately 600 and a dynamic altitude typical of h-40 km, the observation period may therefore have a duration limited to approximately 500 Es.

 The acquisition time T is chosen as a function of the altitude resolution sought (tu10 Rs in the application described here). For an altitude resolution of 1 km, the acquisition time T is of the order of 6 Rs.

 To ensure sufficient mobility of the electrons in the CCD semiconductor material, the sites of the image area 20 preferably have dimensions less than 30 Rm parallel to the surface of the CCD. In the case where P = N = 16, if we consider a typical pixel of 23x23 pin, we can thus produce either an equivalent detector of 368x368 pin, or 16 detection channels of 23x368 pin. Given the small quantity of charges to be transferred, the sites of the memory area 24 preferably have, along the direction of the columns, a smaller dimension than those of the image area. This reduces the amount of charges generated in darkness during transit in the memory area. If the sites have a height of 11.5 pine and a width of 23 pine in the memory area, it only requires a surface 368x368 pxn for M = 32 and N = 16.

 The frequency f of charge transfer in the image area 20 is for example 12 MHz. In this case, if we consider that the frequency f must be at least equal to 2P times the frequency F corresponding to the inverse of the acquisition time, this frequency F would be at most 375 kHz (T = 2.66 rus), which corresponds to a minimum resolution at altitude of around 400 meters in pointing nadir.

 The number M of lines in the memory area 24 is taken at least equal to the ratio between the duration of the observation period and the duration of acquisition T. M must therefore at least be equal to the ratio between the dynamics at altitude h and the altitude resolution of the system. For 30 km of altitude dynamics and for a resolution range from 450 m to 1000 m, we can estimate that we need at least M = 32 and at most M = 100.

 For a reading frequency of 50 kpixel / s, allowing with current CCDs reading noises of 2 to 3 electrons on average, the whole of register 26 can be read in 10 ms if M = 32, N = 16, and in 32 ms if M = 100, N = 16.

The register 26 can therefore be entirely read between two successive observation periods as soon as the source 10 operates at repetition rates smaller than 30 to 100
Hz.

 The CCD has N detection channels, one per column, which provide spatial resolution perpendicular to the direction of the columns. As the interference filter 16 distributes the incident light flux on the columns as a function of the wavelength, the CCD allows a wavelength resolution of the pulse echoes. Since the source 10 is monochromatic, the CCD can be used for Doppler measurements in a wind lidar system.

 To optimize quantum efficiency, the CCD can be thinned and backlit. Given the small size of the CCD, and therefore the resulting low parasitic capacitances, the generator of the clock signals 28 can be installed on the same chip, which makes it possible to limit the number of external control signals. If the CCD is polarized by the MPP (multi-pinned phase) technique, the dark current is less than 20 pAmp / cm2 for a temperature of 250C, or 330 electrons per second at 250C in the memory area. In certain applications, it may be necessary to cool the CCD to further reduce the charges due to the dark current, when a very high sensitivity is sought for a relatively long total reading time.

Claims (10)

 1. Charged coupling device (14) for temporally analyzing a light signal, comprising an image area (20) constituted by a matrix of P rows and N columns of photosensitive sites to which the light signal is directed, a memory area (24 ) constituted by a matrix of M rows and N columns of non-photosensitive sites receiving the charges which appeared in the image area during an observation period, and a reading register (26) of N sites successively receiving the charges stored in the rows of the memory zone, characterized in that, at the border between the image zone and the memory zone, the device has an integration zone (22) comprising a line of N sites, in that, during the observation period, the charges appearing in each column of the image zone are transferred along this column to a respective site in the integration zone at a first transfer frequency (f), and accumulated in said s ite over an acquisition period (T) substantially greater than P times the inverse of the first transfer frequency, and in that, during the observation period, the charges accumulated at each site in the integration zone are transferred along a respective column of the memory area in the direction of the read register at a second transfer frequency (F) of the order of the inverse of the acquisition time (T).  2. charge coupling device according to claim 1, characterized in that the sites of the image area (20) have dimensions less than 30 microns parallel to the surface of the device.  3. charge coupled device according to claim 1 or 2, characterized in that the sites of the memory area (24) have, along the direction of the columns, a smaller dimension than those of the image area (20) .  4. Charge coupling device according to any one of claims 1 to 3, characterized in that the acquisition time (T) is less than 10 microseconds.  5. Lidar, comprising a pulsed light source (10) and a detection system arranged to receive reflected echoes of the light pulses emitted by the source, characterized in that the detection system comprises a charge coupling device (14) comprising an image zone (20) constituted by a matrix of P lines and N columns of photosensitive sites towards which the echoes received are directed, a memory zone (24) constituted by a matrix of M lines and N columns of non-photosensitive sites receiving the charges appeared in the image zone during an observation period following the emission of each light pulse by the source, and a reading register (26) of N sites successively receiving the charges stored in the lines of the memory zone, in that , at the border between the image area and the memory area, the charge coupled device has an integration area (22) comprising a line of N sites, in that, during each observation period, the charges appearing in each column of the image zone are transferred along this column to a respective site of the integration zone at a first transfer frequency (f), and accumulated in said site over an acquisition period (T) substantially greater than P times the inverse of the first transfer frequency, and in that, during each observation period, the charges accumulated in each site in the area of integration are transferred along a respective column of the memory area in the direction of the read register at a second transfer frequency (F) of the order of the inverse of the acquisition time (T).  6. Lidar according to claim 5, characterized in that the detection system comprises an optical filter (16) arranged to direct towards different columns of the image area (20) of the charge coupled device of the wavelength spectral components different from the echoes of the light pulses.  7. Lidar according to claim 5 or 6, characterized in that the sites of the image area (20) have dimensions less than 30 microns parallel to the surface of the charge coupled device (14).  8. Lidar according to any one of claims 5 to 7, characterized in that the sites of the memory area (24) of the charge coupled device (14) have, along the direction of the columns, a smaller dimension than those in the image area (20).  9. Lidar according to any one of claims 5 to 8, characterized in that the acquisition time (T) is less than 10 microseconds.  10. Use of a lidar according to any one of claims 5 to 9 on board a satellite (50), to analyze echoes of the light pulses reflected by atmospheric particles.