MAS-CCD: New technique for measuring low-level charge content based on the multiple amplifier architecture
Miqueas E. Gamero, Guillermo Fernández Moroni, Fernando Chierchie, Agustin J. Lapi, Blas J. Irigoyen Gimenez, Juan Estrada, Javier Tiffenberg · 2026
Read MAS-CCD: New technique for measuring low-level charge content based on the multiple amplifier architecture by Miqueas E. Gamero, Guillermo Fernández Moroni, Fernando Chierchie, Agustin J. Lapi, Blas J. Irigoyen Gimenez, Juan Estrada, Javier Tiffenberg (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-13220.txt.
Introduction
Gamero and six colleagues present a method for gauging spurious charge in multi-amplifier sensing charge-coupled devices, a key noise source for faint-object astronomy and exoplanet searches. Clocking the gates produces spurious charge that is hard to measure with current methods, and tuning it trades against full-well limit. Their method uses covariance analysis across the device's output amplifiers, which read the same charge packet at different times. It enables fast precise measurement where conventional approaches struggle.
Catalog record
| CREATOR | Miqueas E. Gamero, Guillermo Fernández Moroni, Fernando Chierchie, Agustin J. Lapi, Blas J. Irigoyen Gimenez, Juan Estrada, Javier Tiffenberg |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56452 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.13220
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
MAS-CCD: New technique for measuring low-level charge content based on the multiple amplifier architecture Miqueas E. Gamero, Guillermo Fernández Moroni, Fernando Chierchie, Agustin J. Lapi, Blas J. Irigoyen Gimenez, Juan Estrada, Javier Tiffenberg Abstract Low-noise detectors are a key technology for the next generation of astronomical instruments aimed at spectroscopy of faint objects and the search for exoplanets. In this context, the multiple-amplifier sensing charge-coupled device (MAS-CCD) emerges as a promising technology for future scientific instruments. A critical parameter affecting the performance of these devices is spurious charge, produced by the clocking of the gates. Its measurement is typically challenging with existing methods. In practice, the optimization of this parameter often relies on empirical procedures that require significant time and careful consideration of the trade-off with full-well capacity. In this work, we present a new technique to estimate spurious charge based on covariance analysis of the output amplifiers of the MAS-CCD, which measures the same charge packet in different amplifiers at different times. The method enables fast and precise measurements of spurious charge under operating conditions where conventional approaches are difficult to apply. We develop the theoretical framework of the method and validate the model through simulations. The results demonstrate the feasibility of this approach and suggest that it could serve as a basis for reliable large-scale characterization of sensor performance. I Introduction The Multi Amplifier Sensing Charge Coupled Device (MAS CCD or MAS-CCD) offers a powerful architecture to reduce readout noise without increasing the sensor’s readout time. The MAS-CCD is based on many floating gate output stages on the same serial register 6 ; 4 . Figure 1(a) shows a diagram of its architecture. As the pixels are transferred through the serial register, the different output stages measure the pixel’s content. Then, the information from the different amplifiers for the same active area pixel is combined into a final pixel value. Since the samples from the different amplifiers are independent, the final pixel value has readout noise smaller by a factor equal to the square root of the number of amplifiers. Its ability to reduce readout noise with a negligible impact on the readout time has encouraged its use in the next generation of astronomical instruments 8 ; 10 . Two examples are its use in a new terrestrial instrument to make an extensive survey of the spectrum of distant galaxies for universe expansion studies 3 , and its use to measure the spectrum of exoplanet atmospheres to find Earth-like planets from space-based observatories 11 . Once the readout noise of the sensor is reduced, other intrinsic sources of noise define the new uncertainty limitation in the pixel charge measurements. One of the dominant sources in CCDs is the spurious charge (SC), also referred to as clock-induced charge (CIC), produced by the clocking of the gates to move the charge in the pixels of the array 7 ; 2 . In particular, for astronomical applications, since the swings of the clocks are large to be able to transfer large amounts of charge (several thousand carriers), the SC could have a significant impact on the measurement error of the charge collected by the pixels 13 ; 10 . In particular, the SC produced in the serial register has been measured as the most significant contribution to errors 5 ; 14 . (a) Sketch of the MAS-CCDs architecture. (b) Graphical representation of the pixel offsets and random variables definitions for the calculation of the covariance values using pixels from different amplifiers when they are in charge phase (blue) and in noise phase (red). Figure 1 : Sketch of simplified MAS CCD output stage and its output images. Measuring the SC using traditional techniques is challenging due to the convolution of the Poisson distribution of the SC with the Gaussian distribution of the readout noise. Since the Gaussian component includes an arbitrary offset introduced by the amplifier electronics, it is difficult to disentangle the mean of the Poisson-distributed SC from the mean of the Gaussian noise. An alternative approach is to attempt to infer the presence of SC through the overall width, or standard deviation, of the distribution. However, this method is also imprecise because, for low mean values of SC, the total width is dominated by the readout noise, making the SC contribution difficult to isolate and quantify accurately. These challenges can be addressed by leveraging the intrinsic features of the MAS-CCD architecture. Since the charge signal is correlated across all output channels—being derived from the same pixel—while the noise contributions from the amplifiers are uncorrelated, it is possible to separate the two random process using correlation analysis. By computing the correlation function among amplifier outputs, one can isolate the signal-related (correlated) components from the noise (uncorrelated) components. This approach provides a promising method to efficiently decouple the contribution of SC from readout noise, enabling more accurate characterization of SC effects even when their mean levels are low. Section II describes the main architectural components of the MAS-CCD and introduces mathematical models for the intrinsic error sources in the device. Section III takes an initial approach to estimating the SC from the covariance of two pixels carrying the same information but originating in different channels. Section IV presents the variance of the proposed estimator. Section V explains the general framework for estimating the signal with an arbitrary number of channels and presents an optimal way of combining their information to improve the estimation. Section VI describes how to handle a possible source of correlated noise among channels. For an extension of this method to dark-current estimation, refer to Section VII . Finally, Section VIII presents the implementation of the technique along with simulations that validate its functioning. Appendix A includes the full mathematical derivations used across the different sections. II MAS-CCD operation and intrinsic noise source modeling A simplified layout of a 3-amplifier MAS-CCD is presented in Fig. 1(a) . During the readout operation, each row of the active matrix is transferred to the serial register, where it is sequentially read out through the output amplifiers. The prescan area is used to facilitate the transport of charge packets to the output stage. For simplicity, in the present analysis, these few pixels are not considered. Once a row has been read out, the next row is transferred to the serial register and undergoes the same readout process. Thus, although the image is displayed as a 2D matrix, the readout scheme can be interpreted as a 1D stream of data that is later rearranged into a 2D image. Since each amplifier records data independently, the readout system generates distinct images corresponding to each amplifier. Figure 1(b) depicts three distinct output images obtained from each amplifier. The spatial shift between pixels with the same information is due to the extra transfers, N_{p} , required for the charge packets to reach the next amplifier. Although the amplifiers are read out simultaneously, each one samples a different pixel at a given time. Consequently, the same pixel in the active region is measured by different amplifiers at different times. The resulting pixel information for amplifier i , where 1\leq i\leq N_{a} , can be defined as p_{i}[n] , where n is the index of the pixel in