The GLO Instrument has a set of imaging spectrographs that simultaneously observe the entire wavelength range from 1150Å to 9000Å at a resolution of ~5Å in the UV and visible to ~10Å in the near IR. The spectrographs consist of five modules, each with its own CCD detector. The spectrographs are designated with the letters A - E, starting at the long wavelength end. Spectrograph A used an unintensified CCD and has been found to be too insensitive to be useful for airglow observations. The CCDs in spectrographs B - E are coupled to image intensifier tubes and have two gratings each, so that two spectra are imaged on each detector. The field of view of each of the set of 9 co-aligned spectrographs is 0.1° X 8°.

Spectrograph Wavelength Ranges at present
Plate Scale

The resolution is determined by the width of the entrance slit. Typically the slits are two pixels wide. The slit image in the image plane is not rectangular but broadened slightly by an event spread function in the image intensifier.

In later flights, we have improved the capability of the experiment by adding star tracking. This allows us to make absolute absorption measurements of the neutral atmosphere by tracking stars into the limb. Stars are used as calibration sources. By selecting a star of the proper type, emission can be recorded simultaneously across the 9 segments of the spectrum, solving the cross-over problem from segment to segment. We have also implemented limb tracking to remove the limit cycle of the shuttle from the data as it was being recorded. When deployed on the ISS the pointing will be referenced to an inertial platform improving the use of these techniques.

Upgrade of the GLO Spectrographs


The potential sensitivity of the GLO spectrographs has never been reached because of the electronic technology of the late 80's. The primary problem was read noise caused by the multiplexing technique required to read-out eight CCDs quickly. A single analogue board was used some distance from the CCDs. The analogue board used a 12 bit A/D converter a fast device at that time but resulted in a sensitivity threshold of about 60 electrons, not low enough to allow photon counting. The sensitivity limit was about 100 Rayleighs regardless of the integration time. The calculated performance of the spectrographs is about 4 Rs.

The new CCD detectors will have their scan and program electronics integrated onto the back of the CCDs along with an analogue chain. The A/D converter will have a 16-bit range. Communication with the CCD will use TCPIP ethernet protocol.

The question of a bare or intensified CCDs will be addressed as part of the phase A study. The intensified CCD has an advantage when photon-counting sensitivity is required. Reports on new CCDs suggest they are capable of single event sensitivity. That with their higher quantum efficiency may recommend application at the longer wavelengths. With photon counting we can expect a few Rayleighs sensitivity with fast exposures and tenths of Rayleighs with integrated exposures.

The improved CCD readout capabilities will allow us to implement spectrograph A and extend the wavelength to the CCD cutoff near 10900Å. With bare CCDs in channels A and B we should see an improvement in resolution to nearly the 2 pixel width.


Although we have continued to fly simple monochromatic imagers on the GLO, the data has not received much attention for lack of funding. The imagers had some obvious shortcomings due to the electronics points mentioned above. The limited dynamic range of the A/D converter required too much real-time interaction to keep the exposure within the right range. The new CCDs will solve that problem. We will also implement auto gain control.


In the last few flight of GLO we have used a commercial CCTV camera for two functions. Scene viewing was available through the Shuttle video channel. A frame grabber was used for tracking purposes and for images to accompany the science data. This procedure will have to be reviewed with respect to the new data stream capabilities.

o Last Updated: 23 August 2001

This page is maintained by:
Terry Forrester