PROCEDURE: 8s

TITLE: Deep external pinhole

CATEGORY: Calibration

PURPOSE: Internal flatfields and diffusing screens in front of the
instrument have lower f/ratios than distant sources seen through the HST
OTA (or simulator like RAS/HOMS).  In order to determine whether this is
an important effect we will take deep images (S/N ~ 100 per pixel) with
the STUFF double-pinhole assembly.  This will illuminate a spot ~8.5mm
in diameter (~91000 pixels) on the detector.  The resultant images will
be divided by deep STUFF diffusor screen flats obtained with the same lamp,
and the resultant ratios will be examined for small scale variations (<~
20 pixels), particularly around the projected position of the repeller
wire.  

FREQUENCY: This should be run with two different filters, to check for
wavelength effects.

DETECTOR: SBC

PREREQUISITES: (1) It would be best if procedure 4s is run first.  It
produces images with the same light source.  These can be used to tune
the lamp current and attenuation to get maximum allowable countrate.
(2) Procedure 4s must be run before analyzing the data. 

HARDWARE REQUIREMENTS: STUFF + pinhole assembley.  Kr and Xe lamps.

SOFTWARE REQUIREMENTS: 

COMMAND MODE: PSTOL

PROCEDURE/EXPOSURES:  
	1. Filter in place (F115LP for Kr lamp; F140LP for Xe lamp).
	2. Turn-on STUFF + line lamp, tune attenuation to get maximum
	   count-rate yet remain under global limit of 2.5e5 c/s (backed
	   off of 3e5 c/s to be safe...) using prior experience under
	   procedure 7s.
	3. Take 120s (tweak) test exposure.  
	   Determine global count-rate
	4. Tweak STUFF attenuation if needed.
	5. Second 120s tweak exposure if needed
        6. Put calibrated PMT in beam, 
	   take reading and enter into log,
	   then remove from beam.
	7. long exposures for example: 
	   2 * 2000s should be sufficient for mean S/N = 100 per pixel
           if global count rate = 2.3e5 c/s.
	8. Same as step 6. above (it would also be good to do this
	   between long exposures)
	9. Dark 2000s

VARIANTS:	a- Kr1236 + F115LP
		b- Xe1469 + F140LP

TOTAL EXPOSURES: ~5

TOTAL TIME: setup + ~2*tweak + 6240s + ~4*overhead = ~2.37 hr
	(setup = 10min, tweak = 10min, overhead = 2min)

ANALYSIS: 
	(1) Combine long exposures, subtract scaled dark.
	(2) Determine total count-rate (dark subtracted).  Compare to
	    model expectation for mean PMT reading: calculate SBC
	    throughput.  Compare with results of procedure 7s.
	(3) Divide result by (dark subtracted) high S/N PtNe flats with
	    diffusor screen (F115LP for 1236 Kr; F140LP for 1469 Xe).
	(4) Normalize result to mean within center of pinhole image.
	    note structural differences larger than 1%.
	(5) For individual (dark subtracted) long exposures compare
            value of global countrate/PMT reading with the values for
	    the corresponding images taken under procedure 7s.  Note any
            deviations in this ratio larger than 1%.