PET Coincidence Logic

The Proton/Electron Telescope (PET) counts particle events in rate scalers and puts pulse-height analyzed (PHA) events into telemetry buffers based on coincidences and anticoincidences of detector triggerings. The flexible design of the instrument's logic circuits has allowed us to change the definitions of the various rates and buffers by command several times in the course of the mission. This page and those linked to it in the table at bottom show a graphical representation of the various definitions we have used, and illustrate the consequences of each set of definitions.

The PHA event buffers are CAL, P, E, RNG, EWG, and PEN. When a particle (or noise, or calibration) event occurs, the detector and PHA thresholds triggered (and the calibration flag) are tested against the logic for these six buffers, in the order listed; the event is read out into the first empty buffer, if any, whose logic is satisfied by the event. This means, for example, that if the coincidence logic is set up so that a given event satisfies the conditions for two buffers, it can be read out into either depending on whether the higher-priority one is already full or not. I will ignore CAL and EWG because they are not subject to confusion with the other rates, the former being identified separately and the latter being "walled off" by the requirement that the P1 detector not be triggered (in all command states to date, P1 is required for P, E, RNG, and PEN). The event rates that are accumulated are, in order, LIVETIME, PEN, RNG, PHI, EHI, PLO, ELO, and EWG; unlike the event buffers, only the first rate scaler whose coincidence conditions are satisfied is incremented. I'm not too clear exactly why LIVETIME is included in this set, but I'll only need to discuss PEN, RNG, and the P and E subcategories (PHI, EHI, PLO, and ELO) in what follows.

The vast majority of events that I have seen in the telemetry have been "orderly", in that all detectors from P1 through PX are triggered (X = 1 to 8), with or without high or low guards, so I will provide a table for each operating mode that shows which event buffers' equations are satisfied by these "orderly" events. I will lump P and E buffers, and therefore PHI, EHI, PLO, and ELO rates, together, since their coincidence logic is affected all together by most changes of command state (one minor exception will be noted). The P and E buffers (and PHI/PLO and EHI/ELO rate pairs) are distinguished by whether or not, respectively, the P1A high threshold is triggered, and the *HI and *LO rates are separated by whether or not, respectively, the P3 detector is triggered.

On each page linked below I will include a set of diagrams like the one above for P & E, RNG, and PEN; these cross-sections of PET show in color the requirements for the detector and guard coincidence logic. The two curved detectors are P1 (above) and P2 (below); the stack of five large-diameter cylindrical detectors composes P3, with its narrow annular guard detectors outside it, and the smaller-diameter cylindrical detector are (top to bottom) P4 through P8, with their broader annular guard detectors around them. Green denotes a detector that is required to be triggered, red indicates one required not to be triggered, and blue shows which detectors are not constrained. The guards are all ganged together in the coincidence logic (that is, it doesn't matter which one is triggered), so they change color together; as a special case for the guards, yellow indicates that the illustrated coincidence condition accepts events with the low guard threshold triggered but not the high threshold, while red indicates that either high or low guard threshold triggers cause an event to be rejected. In detail, then, the above diagram shows a state where P1, P2, and P3 are required to be triggered, and P4 and the high guard threshold are required not to be triggered; I will denote this as P1.P2.P3.P4*.AH* (for some reason, standard PET documents refer to the low and high guard thresholds as AL and AH respectively--"A" for "annular"?). Depending on whether or not the P1A high threshold is triggered, this diagram corresponds to the PHI or EHI rate, under certain operating modes described in the linked pages (AH and AL requirements vary).

Time period

Brief description


Original design, pre-launch


As launched; disable PENE buffer, read out PEN events in RNGE buffer


Accept P and E events with low guard triggers


Test usefulness of events with any (high or low) guard triggers


Same as 92337-92357


"P1 singles"; accept any event that triggers P1, to study low-energy electrons


Accept P and E events with any guard triggers


Accept RNG events with any guard triggers, PEN with no (high or low) guard


Accept P, E, RNG, or PEN events with low but not high guard triggers

Back to top of Data PagesBack to top of Data Pages

new 6 August 1999, revised 7 January 2000

Valid HTML 3.2!