Complement to D0 note 706 ------------------------- For each INDIVIDUAL TRIGGER TOWER signal (= one EM signal + one HD signal), we construct a certain number of energy quantities: e.g. EM Et(=transverse component) and HD Et, also TOT Px and TOT Py. These quantities are built by PROM lookup. The lookup will (a) scale the input signal to the appropriate component AND (b) make corrections using the fast vertex position from level 0 AND (c) possibly make a low energy cut (=force to zero all energies below a given Et for physics cut AND below a given incoming signal level for electronics noise suppression) AND (d) make some offset and scale adjustments internal to our system (you can probably forget about this last point). This lookup is unique; I mean only one value is constructed for each quantity type. We also feature a second lookup ability. That is: immediately after the first lookup of EM Et and HD Et we can construct an extra quantity on the EM and HD signals, using the PROM resources spared by the first lookup. We used to think (i.e. 5 years ago) about constructing the EM E and HD E signals, but it brings no help to triggering. The other possible use of this second lookup COULD BE another EM Et and HD Et lookup with a different low energy cut. From that point on, the information takes TWO DISTINCT PATHS. ~~~~~~~~~~~~~~~~~~ The simplest path to follow is the arithmetic sums of all this scalar quantities over the whole detector. global EM Et = sum of Trigger Tower EM Et global HD Et = sum of Trigger Tower HD Et global TOT Et = global EM Et + global HD Et global EM L2 = sum of Trigger Tower EM second lookup global HD L2 = sum of Trigger Tower HD second lookup global TOT L2 = global EM L2 + global HD L2 global Px = signed sum of Trigger Tower Px global Py = signed sum of Trigger Tower Py final P stage = comparison of Px and Py against a SQRT(Px**2+Py**2) threshold Any number (min 0, max probably 16 or 32) of GLOBAL thresholds can be set for any of these global energy quantities (with the exception of global Px and Py where it is useless) and the result of each comparison goes to the andor network of the trigger framework. The other path starts by comparing some of these INDIVIDUAL TRIGGER TOWER QUANTITIES to some LOCAL THRESHOLDS: for each trigger tower, TWO decisions are made: Is [EM Et greater than threshold] AND [HD Et lower than threshold (i.e hadronic veto)]? ~~~ and Is [TOT Et (=EM Et+HD Et) greater than threshold] ? Note that each decision of the first type requires two threshold values (one EM Et and on HD Et threshold), and each decision of the second type only requires one threshold value. Since all these thresholds are individually programmed for each trigger tower (and also to avoid confusion with global thresholds), we chose the name "reference set" to qualify a list of thresholds of one kind specified for all the individual trigger towers. The multiplicity is FIXED at four; there are four INDEPENDENT decisions of each type made for each trigger tower. There are thus FOUR reference sets of each of these THREE kinds to make decisions of TWO types. After each individual trigger tower decision is made, we count all the positive decisions over the whole detector: global EM Et Trigger Tower count = count of all Trigger Towers satisfying EM Et >= threshold AND HD Et NOT >= threshold global TOT Et Trigger Tower count = count of all Trigger Towers satisfying TOT Et >= threshold There are four independent counts of each type (this follows the reference set multiplicity fixed at four). For each such count, once again, any number (min 0, max probably 8) of GLOBAL thresholds can be set, and the result of each comparison goes to the andor network of the trigger framework.