LSM Input for the Central Calorimeter's L1 Calorimeter Trigger --------------------------------------------------------------- Revised 3-DEC-1990 We are cooking the PROM's for the production CTFE Rev. B cards. There are 2048 PROM's in the CC's worth of Cal Trig. This is about 4.2 MBytes of PROM data. Because it is a big job to change these 2048 PROM's, we want to make this first pass of CTFE PROM's as close a possible to a good final version. The LSM program generates and manages all of the lookup PROM data but it needs to know a few input parameters before it can start its job. Some of these input parameters are purely technical, some have to do with physics and optimum triggering, and some have to do with what choices we want to make about how we operate the Cal Trig and what options to the Cal Trig system we want to choose to use. 1. Level 0 Bins Geometric Coverage ---------------------------------- 9-OCT-1990 a note from Rich says that they have reached 2.31 GHz ; so lets go for the 2.4 GHz number. 3.00 E10 cm/sec --------------------- = 12.5 cm 12.5 cm / 2 = 6.25 cm 2.4 E9 Hz 2. Lookup Types --------------- Transverse Energy is the Lookup Type for the "1st lookup" in the energy PROM's. The "2nd lookup" will be a one to one page with no cuts i.e. data out equals address in. 3. Lookup Pages --------------- 7 pages will be used for the "first lookup" to calculate Et's in the energy PROM's and Px and Py in the momentum PROM's. One of these pages is centered at Z=0. A single page is used for the "second" energy lookup. For both the Energy and the Momentum PROM's The 8th page will be a one to one page (i.e. address in equals data out, with no cut for noise or for physics). For the cosmic ray test, we will use the one to one pages. 4. Mapping of Level 0 Bin vs Lookup Page ---------------------------------------- We will use a 6,4,4,3,4,4,6 map of L0 Fast Vertex Bins to Lookup PROM pages. Specifically -15 through -10, -9 through -6, -5 through -2, -1 through 1, 2 through 5, 6 through 9, and 10 through 15. 5. Lookup Page Nominal Center ----------------------------- We will use the geometric center of each page for all vertex corrections. 6. Global ADC Scales -------------------- For now all Eta's use 1/4 GeV per bit. We will always keep this for the Eta Index range -8 through -1 and 1 through 8 but we may change this for Trigger Towers a bigger absolute eta's. 7. Global Energy Scales ----------------------- 1/4 GeV per bit in the energy circuits and 1/2 GeV per bit in momentum circuits. 8. Electronics Noise Suppression Coefficient -------------------------------------------- 29-November-1990: implement a bilateral cut with signed arithmetic. 30-November-1990: decide on using a +/- 2.5 sigmas for all Energy and Momentum Lookups Bo showed that 1) the EM noise is completely eliminated by the Physics cut of item 16 below. 2) the HD noise is also completely suppressed. Safety Margin: If the actual noise were 50 % larger than calculated, only 1 in 50 event would have an hadronic Et off by +/- 3 GeV. 3) The noise is fairly well suppressed in the Momentum calculation. Only 3 in 1000 events show 1.5 GeV of Missing Pt due to only noise. Safety Margin: 15 in 1000 would have about 4 GeV of MPt if the noise is 50% higher than calculated. 9. Trigger Tower Geometry ------------------------- 5-NOV-1990: For the Total Energy signal, use geometric middle of EM and HD 7-NOV-1990: Sylvain's Geometry program Corrected by Steven (+Philippe) use file D0$STP:CAL_STPFILE.DAT All of Sylvain's numbers were off (mainly because it was using the wrong STP file) . But the ratio sin(theta)= r/z was very close in the central region. The Resistor networks already built are thus USABLE. 10. Trigger Tower Electronics Noise ----------------------------------- We have no accurate measurements. Bo used a program by B.Pope's student that computes electronics and U noise in readout channels and electronics noise in trigger towers. Bo verified that it uses the proper cell capacitance, the proper cable length. The program refers to D0 note 459 for its noise calculation. 28-November-90: decide to use "3.0" as the ratio of (square root of) bandwidth between main readout and trigger towers. Of the measurements that we have none say that it is bigger than this calculated noise. TRGMISC:TRG_TWR_NOISE.28-NOV-90 holds the noise values and the generating program. 11. Input Energy Error ---------------------- All known effects have been taken out in the summing resistors of the BLS cards. The scaling of E to Et has been done with accuracy on the CTFE cards. These numbers are all set to zero. 12. Analog Input Scaling Factor ------------------------------- The attenuation values used in the Terminator-Attenuators for the CTFE cards in the eta range -8 through -1 and 1 through 8 are in the file: CTFE_DEV:CTFE_ATTENUATORS_V10.TXT These resistors do not achieve the "perfect" factor, because they were computed from incorrect tower coordinates. But for eta [-4,+4] they are close enough for the cosmic ray test. They will probably be rebuilt along with the resistors for the second delivery. 13. Downloaded Bytes -------------------- This data comes from another source and is not needed to generate the data for the lookup PROM's. 14. ADC Zero Energy Response ---------------------------- After some discussions the decision is to use zero GeV into the CTFE card gives an ADC output code of 08, i.e. you can see about negative 2 GeV of Et Energy. Items #6 and #14 together imply: The conversion from ADC counts to Et energy is accomplished by subtracting 8 from the ADC count and multiplying the result by 1/4 GeV per bit. This is what is in the Data Block that goes to L2. 15. Lookup Energy Scale Shift ----------------------------- For now there are no energy scale shifts for either the energy PROM's or for the Momentum PROM's at any eta's. 16. Lookup Transverse Energy Cut -------------------------------- 0.5 GeV of Et for both EM and HD at all eta's. 0.5 GeV of Et for both Px and Py at all eta's. Note that this cut does not contribute, because it is always smaller than 1 LSB count. 17. Final Fitting (Tweak) ------------------------- We have no measurements on which to base any final adjustment, so for now they are all zero. 18. Lookup Zero Energy Response ------------------------------- Because we do have to handle "negative" numbers in both energy and momentum circuits; the zero energy response of the lookup PROM's cannot be zero. The larger the offset, the lower the full scale energy can be. The minimal offset is chosen to allow a correct lookup of the lowest negative energy read by the ADC. An exception is the "8th page", the one to one page that has no cuts. In this case; zero energy into the CTFE card will result in the code of 08 coming out of the PROM's (both energy and momentum). 19. PROM Output Minimum Energy Cut ---------------------------------- This is compiled by LSM, taking into account both Electronics noise contribution (item 8 and 10) and Physics cut (item 16) 20. PROM Slope -------------- This is compiled by LSM, taking into account geometry (item 1, 5, 9) and corrections (item 17).