Photon Counting

In photon counting mode the basic idea is to search every bias corrected image for a local maximum; any local maximum above some user set event threshold is called an event. In addition to its pulse height, every event is given a time tag and a position tag. Also, since the physics of X-ray CCDs dictates that charge from a single X-ray photon can spread into more than one pixel, a pixel island or neighborhood is generally extracted about the local maximum to collect all the charge.

Events are stored in an event list which is a FITS binary table with the following columns:

1.

TIME: The chosen unit of time for the ACIS 2C data is in seconds since 1994.0. (For flight instrument ACIS data collected at XRCF the unit is in seconds from 1997.0.) Since there is always ambiguity about the exact arrival time of every photon, which is equal to the integration time (one doesn't know if a photon arrived at the beginning or end of the exposure), the time assigned to photons within a single frame, t_photon is the start time of the frame, T_start plus half the integration time, $\Delta t$.

$$t_{photon} = T_{start} + 0.5 \Delta t$$ (3.10)

In staggered frame mode (see above), the time tag is assigned to each photon within a staggered window based on which window it appeared in. Note that because the entire active area of the chip was open to X-rays it is possible that some events are misassigned in time, but the great majority of X-rays arrive within the focussed HRMA image spot. The time tag for staggered mode data is:

$$t_{photon} = T_{start} + \frac{\Delta t}{N} (n + 0.5)$$ (3.11)

where N is the total number of windows in each raw FITS image (see table above) and n is which window the event was recorded in.

2.

READX: The location of the event in the serial transfer direction. Quadrants B and D are reversed from `normal' and overclock pixels are counted.

3.

READY: The location of the event in the parallel transfer direction. In full frame mode, this is the same as CHIPY.

4.

CHIPX: The physical location of the event in the X direction. Nodes have been flipped and overclock pixels removed.

5.

CHIPY: The pseudo physical location of the event in the Y direction. In full frame mode this is the true Y location; however, in Staggered Frames mode, this is the location within the subframe. CHIPY goes from 1 to the size of the window.

6.

LABX: This is the same as CHIPX. `LAB' is a left-over reference to subassembly applications.

7.

LABY: This is similar to CHIPY, except it is centered on the CCD. So LABY goes from 512-(r/2) to 512+(r/2), where r is the number of rows per window.

8.

FRAME: Frames are counted sequentially for all raw FITS images and windows. So in full frame mode and continuous clocking, the FRAME number is the number of raw FITS files in the data set. In Staggered mode multiple images (equivalent to CCD frames but with fewer rows) appear in a single FITS file. In this case the raw FITS image number is equal to the FRAME number divided by number of windows per frame, N:

$${\rm raw\ FITS\ image\ number} = \left\lfloor \frac{FRAME}{N} \right\rfloor +1$$ (3.12)

9.

pulse height(s): The pulse height channel column varies in size depending on the data extraction mode; however all modes share some common bases. The pulse height(s) are stored in a vector (a vector of 1 is still a vector). The first element of the vector is always the bias corrected pulse height of the local maximum. The other pixel values (if present) are arranged in a specific order as detailed below.

10.

NODE: Simply the CCD output node from which the data were extracted, 0 through 3.

The number of pixels, in the vicinity of a local maximum, which are processed to extract information about the X-ray events varies. The different formats are described below.