CCD switching noise?

[Bernius1]

How do CCD's/digital chips control noise? If you're looking to detect a microvolt signal, I'm sure the switching spikes & junction capacitance will affect signal. What about an analog (sine wave) scan, or using a much higher frequency to scan than the source frequency? (Like switching power supplies, or class-d amps) And if I want to test a prototype junction, will one junction, fed at switching freq.,exhibit the same noise influences, so I can deal with it early ??

[Ron H]

No-vice, either no one knows the answers, or they don't understand the questions, or both. I know I don't understand your questions. It remains to be seen whether I have any answers. Can you rephrase the questions, maybe with some hint of what you are trying to do?<p>Ron

[Bernius1]

Ron, thanks for the correction; I'll clarify. Any visual image to be captured, whether light, IR, X-ray, or UV, must be scanned off a chip/tube with a bilateral matrix, and the signal is multiplexed. In a tube, the amplitude of the electron beam varies, and this variance becomes the signal. On a CCD, the same thing happens, but as each receptor (pixel) is turned on/off, there's a spike of noise produced. This distorts the signal. Add to that the junction capacitance at high freq.'s, and S/N ratio becomes a problem. So I thought either an analog clock/scan signal (Sine), or use a higher scan freq. than display freq.(TV @ 300 frame/sec.) so that the noise averages out. (???) I'm looking to run a single prototype pixel @ 1 Mhz, and before investing time & $, I want to minimize pitfalls. Does this make more sense?

[haklesup]

The answer to your question is deeply rooted in the methodologies used to design IC chips. The noise you imagine should not exist in a proper design. <p>Much effort goes into simulating the electrical signals that are present in the circuit. The models that go into Spice and VHDL are based on the layout as well as the process used to make the device. IC designers work closely with the fab to obtain the correct model variables for a given process and often have to characterize test structures to get them on their own.<p>On a CCD the circuits that create the alternating electric field that pumps the charge from pixel to pixel down the column form the row lines. At the end of each column is a differential amplifier that senses each bundle of charge. There may be additional circuitry to buffer that signal to the pins but the point is that they are specifically designed at the outset to avoid built in noise. <p>Where one portion of circuitry may produce noise that can couple into another section of the IC (as in many mixed signal devices) the sections are separated by structures known as guard rings or trenches and are powered by seperate external (to the device)power supplies. For example those little 56 pin parts that control the audio in palm computers have 5 seperate supplies inside. (Core Digital, Digital I/O buffers, Analog, Speaker, Headphone)<p>With only one pixel it stops being a pixel and becomes the whole pix so I am not sure what you are scanning anymore. What you seem to be saying is that you want to rapidly sample this light sensing device and were wondering about the signal you would use to do that. That all depends on what exactly you are trying to drive and what it requires to operate optimally.<p>At high frequencies, semiconductor devices tend to use more current and thus generate more heat. Thermally generated charge carriers may become your main source of noise at very fast operating rates.<p>Well, I made a lot of guesses at what you wanted and my knowledge only goes a little deeper but I hope it helps. (I don't design chips I just take them apart)