Friday, February 19, 2016

0.25e- Pixel Noise in 2007

It was brought to my attention that in 2007, IEEE Transactions on Nuclear Science published Max Plank Institute paper "A Novel Way of Single Optical Photon Detection: Beating the 1/f Noise Limit With Ultra High Resolution DEPFET-RNDR Devices" by Stefan Wolfel, Sven Herrmann, Peter Lechner, Gerhard Lutz, Matteo Porro, Rainer Richter, Lothar Struder, and Johannes Treis. The paper is also available for free download here. The described pixel structure allows a multiple reading and averaging of the photoelectric signal, so that 1/f noise can be averaged:

"In this work we demonstrate theoretically and experimentally the capability to reduce the readout noise of an optical and X-ray photon detector based on the semiconductor DEPFET device below a level of only 0.3e- ENC (equivalent noise charge). The readout method used is called "repetitive non destructive readout" (RNDR) and was realized by placing two single DEPFET-devices next to each other and by coupling their charge storing region by an additional gate. By transferring the stored charge from one DEPFET to the other and vice versa the same charge can be measured non-destructively and arbitrarily often. Taking the average value of a large number n of these measurements, the noise is reduced by 1/radicn. The main advantage of such a detector is to greatly reduce the contribution of the 1/f noise to the readout noise. The theoretically and experimentally achievable resolution for different operating parameters (leakage current, readout noise, number and duration of readouts) was investigated by Monte-Carlo simulations and verified on a real RNDR minimatrix (pixelarray). Single optical photon detection with high quantum efficiency and, even more fascinating, the possibility to distinguish between different numbers of photons, e.g., 100 from 101 are demonstrated in measurements."

Thanks to LP for the link!


  1. New to you Vlad, but referenced in our DSRN paper:

    J. Ma, D. Starkey, A. Rao, K. Odame, and E.R. Fossum, Characterization of quanta image sensor pump-gate jots with deep sub-electron read noise, IEEE J. Electron Devices Society, vol. 3(6), pp. 472-480, November 2015.

    You should note it is 320 reads at -55C and a fairly large pixel.

    I invited that group to submit a paper to the SENSORS journal Special Issue on Photon-Counting Image Sensors so maybe we will get an update soon. I get the sense they may have abandoned this particular line of work and are instead focused in the x-ray energy range.

    Caeleste also achieved DSRN by multiple (many) reads at lower temperatures as reported at the 2015 IISW.

  2. Being able to distinguish single electrons can be of advantage even for the x-ray energy range, as the resulting histograms deliver a very robust method for in-situ calibration of the electronics chain gain function (and therefore the photon energy) by simple counting of the (electron-) bumps.
    Gain stability and linearity (after calibration) can be more important than pure noise performance for many telescopes.


All comments are moderated to avoid spam and personal attacks.