Dark current and dark count in photon collection devices

[Emperor42]

Does anyone know of a simple relation between dark current and dark count in photon collection devices?

 

[Khashishi]

For a CCD, the dark charge is just the dark current times the integration time. There is an amplifier which converts the charge to a voltage that will be read by an ADC. The gain on the amplifier will affect the dark counts. Many cameras have multiple settings for system gain, perhaps 1 count per electron, 1 count per 2 electrons, etc. The total noise is some combination of the dark noise, the read noise, and the shot noise.

 

[Drakkith]

Does anyone know of a simple relation between dark current and dark count in photon collection devices?

Khashishi had a good explanation. I'd just add that temperature matters and that each pixel in a CCD or other detector has it's own relation between integration time, temperature, and dark current/dark count. If you take a dark frame (a picture with the shutter closed so you just capture an image of the dark current and bias) you'll see that some pixels are well above or below the average dark count. The pixels that are well above it are known as "hot pixels" and the ones well below the average are "cold pixels" if memory serves.

The best way to estimate the dark count is to look at the specifications of the device and find out how many electrons per second per degree are generated on average and what the gain of the detector is. Then you should be able to calculate the dark count.

 

[ZapperZ]

Does anyone know of a simple relation between dark current and dark count in photon collection devices?

This is very device-dependent. Even within the same type of device, such as a phototube, the dark counts often vary noticeably. I've made antimonide-based phototube and even within the same processing run, the QE and dark counts of each phototube in the same batch often vary.

Zz.

 

[Drakkith]

This is very device-dependent. Even within the same type of device, such as a phototube, the dark counts often vary noticeably. I've made antimonide-based phototube and even within the same processing run, the QE and dark counts of each phototube in the same batch often vary.

Zz.

Do you happen to have a link explaining why the same processing run produces drastically different QE and dark counts in these types of devices (and others like CMOS and CCD)? I'm afraid my one class in semiconductor physics didn't get that in-depth.

 

[Vanadium 50]

Do you happen to have a link explaining why the same processing run produces drastically different QE

Too much eye of newt and not enough wool of bat.

As you might guess, this is not well understood, and there are economic incentives for companies like Hamamatsu not to share any tricks they might have learned. But they don't have it entirely under control either, as two different tubes from the same batch may behave very differently.

 

[ZapperZ]

Do you happen to have a link explaining why the same processing run produces drastically different QE and dark counts in these types of devices (and others like CMOS and CCD)? I'm afraid my one class in semiconductor physics didn't get that in-depth.

There really isn't any published indication on why such a thing happens. There are way too many variables involved: surface roughness, uneven heating, variation in vapor pressure, etc...etc.

And this is not confined only to phototubes. I make photocathodes for accelerators, and even with identical recipes and conditions, we get a range of QE for the photocathodes that we produce. One would imagine that photocathodes having different QE will also influence the conditions that dark current is produced, resulting in similar type of variation.

Zz.