Tentative trigger alignment procedure:
Initial settings:
Module |
trg_alignment |
tack_delay |
---|---|---|
CAL FM 109 |
0x80f00 |
0x2d |
CAL FM 101 |
0x90f00 |
0x2c |
CAL FM 119 |
0x80f00 |
0x2e |
TKR FM 16 |
0x900005 |
0x0 |
TKR FM 8 |
0x900005 |
0x0 |
namely nominal SLAC settings (CAL treq delay = 0, TKR treq delay = 5, CAL tack delay ~ 45, TKR tack delay = 0). Also the delay_ext_trig is set to 0x0, at the beginning.
With these setting the trigger panel in RunControl (run 700000799) shows that, within the trigger window:
- TKR peaks at 10 ticks.
- CAL peaks at 7 ticks.
Also note that the CAL_LO and EXT trigger rates are almost identical, while the TKR is lower by ~ 40%.
Then next try is the following:
Module |
trg_alignment |
tack_delay |
---|---|---|
CAL FM 109 |
0x80f00 |
0x2d |
CAL FM 101 |
0x90f00 |
0x2c |
CAL FM 119 |
0x80f00 |
0x2e |
TKR FM 16 |
0x900002 |
0x0 |
TKR FM 8 |
0x900002 |
0x0 |
with delay_ext_trig set to 0x5. With these settings:
- TKR peaks at 3 ticks.
- CAL also peaks at 3 ticks.
the two primitives being perfectly aligned. The TKR trigger rate is still ~30% lower that CAL_LO and EXT, though. Following figure refers to a run taken immediately after with a different ext_delay; sorry for the confusion, but it shows how the primitives are aligned.
Need to understand why:
- The TKR trigger rate is lower than the others (maybe worth mentioning that the beam is 5 GeV electrons, 90 Hz average rate, 500 Hz istantaneous).
- The (apparently) optimal setting for the TKR treq_delay is 2 and *not* 5 as expected from SLAC. Does the particle energy makes a difference?
So here is the proposal for the data taking during the beam test:
- CAL treq_delay = 0x0 (nominal)
- TKR treq_delay = 0x2
- CAL tack_delay ~ 0x2d-0x2e (nominal)
- TKR tack_delay = 0x0 (nominal)
- delay_ext_trig = 0x8
With these settings the trigger primitives within the trigger window look like this:
How the trigger delay should be set?
The proposed trigger delay settings are fine in case all the trigger primitives are enabled (allowed to open the trigger window), but in the BeamTest only external trigger can open the window so in roughtly half of the event that have an internal trigger we will not see this trigger primitive in the summary word. This can be a problem if people want to study trigger efficiency.
So a second option for trigger settings (to be used in case of ext trigger only) is to delay the internal treq for TKR and CAL by 6 ticks to make the peak in the center of trigger window:
- CAL treq_delay = 0x6
- TKR treq_delay = 0x8
- CAL tack_delay ~ 0x2d-0x2e (nominal)
- TKR tack_delay = 0x0 (nominal)
- delay_ext_trig = 0x8