Digitization - Princeton Instruments Activity Tracker User Manual

76
Software Binning
CCD Array
EEV CCD-37
512 x 512
PID 582 x 782
PID 1030 x 1300

Digitization

During readout, an analog signal representing the charge of each pixel (or binned group of
pixels) is digitized. The number of bits per pixel is based on both the hardware and the
settings programmed into the camera through the software. Depending on the MicroMAX
system, single, dual (100 kHz/1 MHz), or multiple digitization rates may be available.
Dual and multiple digitization provide optimum signal-to-noise ratios at all readout
speeds. Because the readout noise of CCD arrays increases with the readout rate, it is
sometimes necessary to trade off readout speed for high dynamic range. In the most
common ST-133 configurations, there will be a 1 MHz conversion speed for the fastest
possible data collection and a 100 kHz conversion speed for use where noise
performance is the paramount concern. Switching between the conversion speeds is
completely under software control for total experiment automation.
MicroMAX System User Manual
One limitation of hardware binning is that the shift register pixels and the output
node are typically only 2-3 times the size of imaging pixels as shown in Table 10.
Consequently, if the total charge binned together exceeds the capacity of the shift
register or output node, the data will be corrupted.
This restriction strongly limits the number of pixels that may be binned in cases
where there is a small signal superimposed on a large background, such as signals
with a large fluorescence. Ideally, one would like to bin many pixels to increase
the S/N ratio of the weak peaks but this cannot be done because the fluorescence
would quickly saturate the CCD.
Imaging/Storage
Cells Well Capacity
100 x 10
3
18 x 10
3
34 x 10
Table 10. Well Capacity for some CCD Arrays
The solution is to perform the binning in software. Limited hardware binning may
be used when reading out the CCD. Software binning allows you to perform
additional binning during the data acquisition process, producing a result that
represents many more photons than was possible using hardware binning.
Software averaging can improve the S/N ratio by as much as the square root of
the number of scans. Unfortunately, with a high number of scans, i.e., above 100,
camera 1/f noise may reduce the actual S/N ratio to slightly below this theoretical
value. Also, if the light source used is photon-flicker limited rather than photon
shot-noise limited, this theoretical signal improvement cannot be fully realized.
Again, background subtraction from the raw data is necessary.
This technique is also useful in high light level experiments, where the camera is again
photon shot-noise limited. Summing multiple pixels in software corresponds to
collecting more photons, and results in a better S/N ratio in the measurement.
Readout Register
Well Capacity
3
electrons 200 x 10
electrons 40 x 10
electrons 34 x 10
Output Node
Well Capacity
3
electrons 400 x 10
3
electrons 40 x 10
3
electrons 65 x 10
Version 6.B
3
electrons
3
electrons
3
electrons