Previous: The CCD--IPCS Detector
Up: The CCD--IPCS Detector
Next: CCD--IPCS granularity and Dithering
Previous Page: The CCD--IPCS Detector
Next Page: CCD--IPCS granularity and Dithering
The photon--counting detector available on the blue arm of the ISIS spectrograph is the CCD--IPCS. It is fully described in The IPCS-II Users' Manual by J.S.B. Dick.
The CCD--IPCS provides true photon counting with no readout noise, but has an important limit on the maximum photon count rate for a linear response. The present version was rebuilt in June 1990. It has an S20 photocathode, a 4--stage EMI intensifier image tube with phosphors at the last stage, and as back end detector the rebuilt version has a thinned GEC CCD. This latter is slightly smaller than the original (an RCA CCD) so that slightly less spectrum is accommodated. The back end CCD has 22 micron square pixels which can be subdivided into 2, 4 or 8 sub--pixels in spectral or spatial directions. However, there is an internal demagnification, so that `effective' sizes at the photocathode are 3.8 x larger. The maximum available area is 256 x 320 pixels. In the spectral direction, with `x8' subdivision there are 2560 pixels each with an effective size of 10.5 microns; with no subdivision, there are 320 pixels of 84 microns each. In the spatial direction, no subdivision (`x1') yields 1.2 arcsec/pixel, and `x8' gives 0.16 arcsec/pixel.
The CCD-IPCS resolution obtained depends very slightly on wavelength (worse to the blue); a resolution (FWHM) of 32 microns (i.e., 3.0 pixels at `x8') is obtained at 4300Å.
Observers familiar with IPCS-I detectors, used at the AAT and on the INT, will find many things in common between the systems. The principal difference is that the CCD--IPCS has a GEC CCD as back end detector, whereas IPCS-I has a Plumbicon TV tube. Also, IPCS-II has a linear response up to somewhat higher photon count rates than does IPCS-I (see below).
The maximum count rate per CCD pixel for a deviation from linearity less than 10% is dependent upon the frame size (higher maximum rate for a smaller frame), and upon the distribution in intensity of the source (higher maximum rate for emission line than continuum sources). In tests with a prototype CCD--IPCS in 1987 with full format (250x320 CCD pixels) and frame time of 17ms, about 5% loss was found at a measured rate of 1ct/s/CCDpixel. In a `spectroscopic' format (64x320 at `x8' spectral resolution), with a frame time of 5ms, line ratios showed no sign of nonlinearity at count rates up to 6ct/s/CCDpixel. This corresponds to a rate of 0.75ct/s/channel at a spectral resolution `x8' and spatial resolution `x1'.
There are two real--time display screens showing the detection of photon events. On the telescope the display is located next to the high--tension (EHT) controls so that the detector can be monitored as the EHT is brought up. The second display is in the control room, and here the screen must be monitored carefully each time the IPCS shutter is opened.
The IPCS is easily damaged by bright light, and great care must always be taken using it. For any new setup, new on--sky source or new calibration lamp, it is safest to include a large amount of neutral density ( eg, ND4.2) in the beam and reduce it gradually. For comparison lamps, this can be achieved by putting ND3.0 in COMPFILTA and ND1.2 in COMPFILTB. For sky objects, you can do it by putting ND in the main beam filter with the command MAINFILTND ND3.0.
There is a red `Panic Button' and you should keep your finger over this as you open the IPCS shutter each time on a new source of light. Operation of this button will close the IPCS shutter immediately, but will not affect the EHT voltage. Two typical symptoms of a dangerously-high rate are: (a) the dots of light on the real--time CRT display merge to form a solid `splodge' of light (which is not flickering), or (b) if the IPCS shutter is closed, the glow on the real--time display takes more than 2.5secs to fade.