NEAR MSI/NIS Calibration Update #1

To: MSI/NIS Science Team
From: Mike Malin (malin@msss.com)
Subject: NEAR Calibration Update #1
Date: 26 December 1994

Introduction

This WWW page is intended to provide occasional updates to the NEAR MSI/NIS team on my interactions with APL on the calibration of the MSI and NIS. It does not reflect everything that is going on, nor is it an official release by the team or Project. Rather, it is a way for me to keep you all informed and to give you some data to look at if you're interested. It is also a way for me to solicit input.

Status as of 26 December 1994

As of Friday, 23 December, just under 400 images had been taken and placed out in the Science Data Center server. I spent much of the day trying to get into the machine and retrieve data. Since the JHU campus router did not like my ftp attempts, I telnetted into the APL computer, ftp'd to a computer in Arizona, transferred the frames there, then ftp'd to that location from the MSSS server to read them over to San Diego (the latter step because we have a nearly impermeable "firewall" owing to our link to JPL and the DSN). I spent much of the weekend trying to look through the data (see image section, below). Scott is trying to organize the data in a form that will provide more information about the test conditions, but as of Monday 27 December this hasn't happened.

Question about Image Identification

An issue has arisen concerning how we number our images (and, presumably, other data). In the WWW page "Image ID Question", I outline the proposed scheme, and at least one alternative. I believe the team should give the numbering scheme some thought, since this is something we're going to have to use and live with for quite a while. I believe the Project will be open to an alternative proposal, if it is appropriate and made in a timely fashion.

Calibration Image Numbering Scheme

The numbering scheme for the recent CCD detector calibration acquisitions is described in the WWW page "NEAR Calibration Image Numbering Scheme".

Sample Images

When I looked at the calibration data on the APL computers, there were 53 images taken at room temperature, 101 images taken at -20 deg C, 127 at -30 deg, and 95 at -40 deg. These were originally in their IPEX format, but Scott converted them to FITS. I have selected eight example images taken from a sequence to study shutter variability/stability at the "high" temperature range for you to see roughly what we've been looking at over the holidays. Each image is 248 lines (this includes 2 dummy lines before and after the 244 active lines) by 580 samples (this includes inactive, isolation, and dark reference pixels). I've converted the 16-bit data to 8-bit by using a linear conversion, mapping 0-4096 to 0-255, and then converted the resulting images to GIF format.

fuhs741.354
GIF = 43 KBytes

fuhs742.354
GIF = 57 KBytes

fuhs743.354.gif
GIF = 84 KBytes

fuhs744.354.gif
GIF = 101 KBytes

fuhs745.354.gif
GIF = 112 KBytes

fuhs746.354.gif
GIF = 125 KBytes

fuhs747.354.gif
GIF = 131 KBytes

fuhs748.354.gif
GIF = 137 KBytes

Some notes and thoughts about these images:

Note that, since the CCD is "electronically shuttered," one would not expect to see any non-linearity in this measurement (e.g., there are no "moving" parts that could vary with time, and CCDs are very linear charge collectors).

2) The surface of the CCD is clearly extremely dirty. Scott indicates that there was a decal attached to the CCD--I suspect what we are seeing is some of the adhesive left behind on the cover glass. There are also quite a few other bits and pieces of dirt on the detector. Scott indicated that he had discussed the problem with the test engineers, who were reluctant to stop and clean the cover glass for fear that they might remove its anti-reflection coating. After looking at these and other images, I told Scott I felt these data were unacceptable for piece-part-level calibration. I further indicated to Scott that, in its present shape, the detector is, I feel, unflyable. I strongly suggested that the detector be cleaned as soon as possible, so that if a problem were encountered with the cleaning, there would be time to address that problem. Scott believes that the engineers are willing to clean the cover glass, but will only repeat some of the tests already conducted. Since the tests did not take as long as expected, Scott had already gotten their agreement to conduct some tests with a second set of 8 spectral filters (the first tests only used 8 filters). He now thinks he may be able to get them to repeat most of the tests, using this second set of filters.

It should be remembered that piece-part tests like these are useful mostly if something goes wrong in flight and we need to distinguish between system and detector performance (i.e., if you have detector performance separate from system performance, you can isolate whether a problem is in, for example, the electronics or the detector).

3) The bandwidth over the internet from APL is very limited and slow. I was typically getting only a few KBytes/sec when I was transferring the data, as compared to 10's KBytes/sec to most locations, 300-400 KBytes/sec to JPL, and ~900 KBytes/sec to Goddard. I foresee this will be a MAJOR problem for us if the troughput remains this low during the mission.

Final Thoughts

The results of these tests to date strongly reinforces my feeling that we want to be able to look at the data in "near-realtime" to be able to provide feedback on the testing to the Project. As a first test, I think it has been effective in showing where lines of communication need to be strengthened, where more interaction is needed, and where our input may be helpful.