Classifications

• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10F—INORGANIC SEMICONDUCTOR DEVICES SENSITIVE TO INFRARED RADIATION, LIGHT, ELECTROMAGNETIC RADIATION OF SHORTER WAVELENGTH OR CORPUSCULAR RADIATION
  • H10F39/00—Integrated devices, or assemblies of multiple devices, comprising at least one element covered by group H10F30/00, e.g. radiation detectors comprising photodiode arrays
  • H10F39/10—Integrated devices
  • H10F39/12—Image sensors
  • H10F39/15—Charge-coupled device [CCD] image sensors
  • H10F39/151—Geometry or disposition of pixel elements, address lines or gate electrodes
  • H10F39/1515—Optical shielding
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10F—INORGANIC SEMICONDUCTOR DEVICES SENSITIVE TO INFRARED RADIATION, LIGHT, ELECTROMAGNETIC RADIATION OF SHORTER WAVELENGTH OR CORPUSCULAR RADIATION
  • H10F39/00—Integrated devices, or assemblies of multiple devices, comprising at least one element covered by group H10F30/00, e.g. radiation detectors comprising photodiode arrays
  • H10F39/10—Integrated devices
  • H10F39/12—Image sensors
  • H10F39/15—Charge-coupled device [CCD] image sensors
  • H10F39/153—Two-dimensional or three-dimensional array CCD image sensors
  • H10F39/1536—Frame transfer

Definitions

• the present invention relates to a charge-coupled device (CCD) solid-state image sensor having a built in test structure which utilizes an optical injection to measure the charge transfer efficiency for a CCD solid-state image sensor.
• CCD charge-coupled device
• CCD image sensors Many high-end imaging applications today utilize large format, area CCD image sensors. These area arrays are composed of a two-dimensional array of pixels, often called the vertical or parallel registers, that are usually transferred row by row into a single row, often called the horizontal or serial register that is used to clock out the signal. Some sensors may have more than one horizontal register.
• the benefits of this architecture are the high sensitivity, high charge capacity and low dark currents resulting in very large dynamic ranges.
• An important measure of performance for these large format imager sensors is charge transfer efficiency (CTE), which measures how completely charge is transferred along a CCD register. Brodersen, et al., in "Experimental
• CTE is typically measured by the inclusion of a "fill and spill" electrical injection circuit (M.F. Tompsett, IEEE Transactions on Electron Devices, ED-22, No. 6, Jun. 1975, pp. 305-309 and W.F. Kosonocky and J.E. Carnes, RCA Review, 36, p. 566, Sept. 1975, incorporated herein by reference) incorporated on the input end of the CCD shift register (usually the horizontal).
• the difficulty with electrical injection structures is that they require adjustment for each individual die, thus making automation of testing more difficult.
• these small gates are sensitive to electrostatic discharge (ESD); loss of an otherwise functional device can result from ESD failure of the test structure.
• the only charge read out from the horizontal pixel that follows the pixel containing the optical injection signal must come from charge that was not transferred so that an accurate measurement of CTE can be made.
• the input stimulus is supplied by illuminating the sensor with light. Varying light intensity or exposure produces a transfer curve of transfer efficiency as a function of signal. For each different exposure, there will be a resultant signal level.
• the resulting horizontal profile can be used to calculate the transfer efficiency in the case of few transfers (leading edge) and many transfers (trailing edge). If the sensor is uniformly illuminated or if it is read out such that the same charge is transferred from each pixel of the optical injection column or columns, horizontal profiles may be averaged to improve the accuracy of the CTE calculation. Software routines within a camera can use this information to compensate for such inefficiencies.
• Figure 7 is a transfer curve of horizontal CTE vs. signal level.
• the fullframe CCD image sensor shown in Figure 1 is used as an exemplary image sensor 100 in the following discussion. Not all aspects of the structure shown in Figure 1 are known conventionally.
• Vertical CCD shift registers are formed, which serves as both the integrating photoactive region and for parallel (line by line) readout of the pixels.
• Each vertical (parallel) register includes a rectangular array of pixels.
• the photoactive area or image area, 103 is usually surrounded by a dark reference area 102 comprising pixels covered with an opaque material or light shield 1021, such as metal, which blocks the incident radiation of interest.
• a horizontal register, 200 accepts each row or line from the vertical registers one at a time, and shifts charge packets from the pixels to a single output node V out in a serial fashion.
• the output node V out converts each charge packet into a voltage, which can be processed and digitized.
• pixels Contained within the column grouping forming the optical injection structure 104 (see Figure 2) at the leading and trailing edges of the device beyond the dark reference regions 102 are special pixels that enable in-situ monitoring of CTE. As shown in Figure 2, these pixels include optical injection columns 110 and 112, which are photoactive and are bounded on each side by scavenging columns 114. In order to remove optical and diffusion crosstalk components and spurious charges from the periphery 1102 ( Figure 4), scavenging columns 114 are added on both sides of the optical injection columns 110 and
• the scavenging columns 114 may be covered with an opaque material such as metal and transfer any collected charge in the opposite direction of normal vertical charge transfer to a drain (as shown by the arrows in Figure 2). More importantly, since the scavenging columns are not connected to the horizontal register 200, no spurious signal is transferred into the horizontal pixels 214 of the horizontal register. As shown in Figure 4, a drain 1104 at the top of the array is electrically connected to the scavenging columns 114 to remove any charge collected in these scavenging pixels.
• the input stimulus is supplied by illuminating the sensor with light. Electrons generated by this light in the optical injection columns 110 and 112 are "injected” or transferred from each row into pixels 210 and 212, respectively of the horizontal register 200.
• the resulting horizontal profile shown in Figure 5, can be used to calculate the transfer efficiency in the case of few transfers (leading edge) and many transfers (trailing edge). Horizontal charge transfer efficiency is composed of several components; among these are charge transfer from the first phase of the horizontal register over the gate onto the output structure and transfer along the horizontal register.
• the optical injection structure 104 at the leading edge which includes the optical injection column 110, primarily measures the CTE over the first gate onto the output structure, since there are so few transfers from the optical injection column 110 to the output.
• the invention provides an optical structure for injecting charge into the horizontal register of an area-array CCD image sensor for characterization and calibration of the horizontal register.
• This structure can be used to measure the charge-transfer efficiency (CTE) of the horizontal register versus signal level.
• CTE charge-transfer efficiency
• Figure 5 shows a plot of the signal at the output amplifier vs. column number for a typical line of the imager.
• the signal 1300 from the "dummy" or overclock region contains no charge. This is the zero reference level for the image.
• the signal 1302 from the first optical injection column 110 is followed by the trailing signal 1314 in the trailing horizontal pixel 214.
• the box 1303 represents signal lost by CTI (which shows up as 1314 in the trailing pixel) and signal lost by diffusion from the optical injection columns 110 to the scavenging columns 114.
• the signal 1306 is the dark current signal from the dark reference columns under the light shield 1021.
• the dark reference column adjacent to the first photoactive column in the image area 103 has an additional signal 1310 which is partly composed of the signal 1316 lost due to diffusion of photo-generated electrons under the first photoactive column of the image area 103. All other photoactive columns in the image area 103 lose charge due to diffusion, but most of them also gain diffusion charge from adjacent columns.
• the first and last photoactive columns have less signal because they don't have a photoactive column on one side to contribute diffusion charge.
• the first and last “few” columns should have less signal, depending on the wavelength of the incident light, but this application simplifies, to first order for the sake of ease of understanding.
• This size of the signals 1310 and 1311 depends on the wavelength of light. The signal is smaller for incident blue light and grows larger as the incident wavelength grows larger (redder). Incident light with a longer wavelength (red) will create a larger diffusion signal 1310 (or 1311) since it has a longer absorption depth. Charge is generated further below the pixel depletion (or collection) region and more will by diffuse to adjacent pixels.
• Each row of the image area 103 contributes signal 1308, which is transferred to pixels 203 of the horizontal register.
• the sensor is illuminated with a spatially uniform light source.
• the signal in the last photoactive column of the image area 103 also experiences a signal loss 1316 due to diffusion of charge to the dark reference region 102.
• This lost signal shows up as a part of the additional signal 1311 in the first trailing column of the dark reference region 102.
• Optical crosstalk also contributes to 1311.
• the signal 1311 also has a component due to CTI, since this pixel trails a photoactive column.
• Figure 6 illustrates an exploded detail cross section of the circle shown as item 1100 in Figure 4. More specifically, Figure 6 illustrates the dark reference region 102 and a few of the pixels 1200 that are adjacent to an edge of the dark reference region 102.
• Light 1204 that is incident at the edge of the dark reference light shield 1021 may be "piped" by the optical waveguiding of the layers to pixels 1200.
• electrons 1202 generated below the depletion region also may diffuse to pixels 1200.
• This extra signal in pixels 1200 is due to optical and diffusive crosstalk and to CTI.
• This signal 1311 from the first trailing dark reference column of the dark reference region 102 is transferred into the horizontal register.
• this trailing signal 1311 depends on the wavelength of the incident light.
• the signal 1306 known to be charge transferred into the horizontal pixels 202 from the dark reference region 102 and can be separated from signal 1310. Further, because of the drain 1104 connection to the scavenging columns 114, the signal 1300 due to charge packets in horizontal pixels 214 does not contain any charge from the scavenging column (or columns) 114 in the vertical register since these columns are not connected to the horizontal register. The signal 1300 gives the background or zero reference for the optical injection column. If desired, the scavenging column on the leading side of the optical injection column can be omitted.
• the signal 1304 is due to charge transferred from the trailing optical injection column 112.
• a key point of this invention is that the signal 1315 in the adjacent, trailing pixel in the horizontal register is only from charge left behind due to transfer inefficiency, since no signal from the scavenging column (or columns) 114 is transferred into the horizontal pixels 214. Further, the signal 1315 is independent of the wavelength of incident light used in the test, since the adjacent column containing the charge due to diffusion is not connected to the horizontal register.
• the signal 1300 is from the horizontal overclock, which is the repeated transfer of empty pixels after the signal from all the columns has been transferred out.
• N is the number of transfers from pixel 212 to the output.
• the leading optical injection column 110 Assume that the horizontal overclock signal 1300 is 2000 electrons, the optical injection signal 1302 in pixel 210 (from optical injection column 110) is 12000 electrons, the trailing signal 1314 is 80 electrons, and the optical injection column number is 20 for a two-phase horizontal clock.

Landscapes

• Solid State Image Pick-Up Elements (AREA)
• Transforming Light Signals Into Electric Signals (AREA)

Priority Applications (3)

Applications Claiming Priority (2)

Publications (1)

Family

ID=22624794

Family Applications (1)

Country Status (5)

Cited By (1)

Families Citing this family (21)

Citations (5)

Family Cites Families (6)

• 2000
  • 2000-12-14 US US09/736,936 patent/US6803960B2/en not_active Expired - Lifetime
  • 2000-12-20 WO PCT/US2000/034954 patent/WO2001047022A1/en not_active Ceased
  • 2000-12-20 DE DE60039419T patent/DE60039419D1/de not_active Expired - Lifetime
  • 2000-12-20 JP JP2001547658A patent/JP5309284B2/ja not_active Expired - Lifetime
  • 2000-12-20 EP EP00988278A patent/EP1159760B1/en not_active Expired - Lifetime

Patent Citations (5)

Non-Patent Citations (3)

Also Published As

Similar Documents

Legal Events