Full text loading...
Light from a 280 nm LED is totally internally reflected through a quartz prism, evanescently illuminating the array while the sensor is placed directly above (a). The pixel schematic (b) shows the differential pixel (Ref. 16). As shown in the layout (c), one photodiode is covered with metal, providing a dark current reference. The chip (d) is , and the schematic (e) shows the operation of the sensor, with the rectifier generating the bias voltages and currents for the chip from the sinusoidal power supply. The output from the 64 pixel array is amplified, low-pass filtered to reduce noise, and converted to a digital representation. The data output is modulated and wirelessly transmitted via an on-chip inductor.
The sensor is powered by an 8 MHz sine wave from a function generatior in complete darkness, with each sample number representing 32.8 ms. (a) shows the effects of dark current mismatch and non-linear amplification on three different pixels. The outputs are linearized by using the DDC of a reference pixel, shown here (b) over repeated cycles of integration (the upper trace signifies pixel reset after amplifier saturation). The differential output is fitted to a polynomial (c), to which all other pixel outputs are referenced (d). The relative times to integrate the same change in voltage are plotted, and a linear fit determines the relative photocurrent (e). The DDC referenced to pixel [1,1] is shown in (f).
The array itself, as imaged by a microarray scanner is shown in (f). A sample pixel’s data (a) shows the raw data output from the three regions of imaging: negative control, positive control, and dark current. Representative traces from each region are shown in (b). For each pixel, the relative signal intensity, for each region is shown (c). The net signal to noise ratio is (f). The signals for the different regions are spatially mapped onto the chip (e).
Article metrics loading...