GENERAL

 Concern about minimal or no calibration of flight instrument - with incomplete calibration unit, difficult to scan over potential conversion area; also difficult or impossible to get good calibration at 45 degrees azimuth and large polar angles

GAMMAS

* Low energy gammas (20, 50, 100 MeV) - difficult to get a clean beam at a multi-GeV accelerator

* Wide angles - Important because the simulations are likely to be less certain

* Moller scattering - Can distort efficiency measurements in a tagged brems beam if not handled carefully

* Energies above SLAC range - need to find a clean 100-300 GeV photon beam - Fermilab? CERN would be a lot less convenient.

* Broad-spectrum tagging vs. "monoenergetic" tagging - The GLAST beam tests were both done with a beam that used a broad tagger that measured the energy of the scattered electron, as opposed to the SAS-2/COS-B/EGRET calibrations that all used more monoenergetic beams. The energy tagger approach allows a lot more energies to be covered at once, but it reduces the ability to collimate the beam. At least some consideration might be given to going back to a monoenergetic beam. Also, using the broad-spectrum beam does not give the same flexibility to manage the beam intensity at a given photon energy.

* Beam intensity - The issue of multiple photons/pulse needs to be kept at the top of the list. Correcting for those was a major pain in EGRET. The drawback is that a low-intensity beam requires more calibration time.

* Use minimal trigger requirements, record all parameters that can be used for cuts, similar to balloon flight. (Need toA verify that level 2 trigger requirements actually have the desired effect, but they shouldn't be used in normal data acquisition.

BACKGROUND

* Protons and other potential background - hopefully can be done in same beam line as 300 GeV photons, or at least at the same facility

* Electrons 100-300 GeV - significant background concern because of leakage in ACD