OneShot Tests

This is a first round offline examination of the one-shot test data taken during the STR8 tests on May/6/05.  The setup was for two towers A,B (0,4) both enabled for charge injection with DAC=63 (~5 MIPs ?) on every 1/64 channels.  We enabled only periodic triggers with a time separation of dT, and only repeat once, i.e. effectively a doublet of triggers separated by dT. The dT are typically 550-950 ticks (25-50ms) to see in particular what the 2nd event is seeing in the wake of the Qinj from the first event. These doublets are then repeated at much slower intervals (<100Hz) to accumulate statistics. The summary of the run info can be found from the trigger run data summary under STR8 2nd table "One-shot Tests". One caveat is that for all the oneshot tests, we forgot to set the tower 4 (B) TEM busy enable, so that one should be cautious in interpreting the TEM data from tower 4. Empirically, we found the TEM error resulted from this was very consistent on the 2nd event for dT<800 ticks, and gone for higher dT.         

The most informative first look is in fact simply just dumping some pairs of events from these runs (note that the event dumps were limited to up to 16 TKR hit strips per layer, so that not all hits are displayed for the busy events):

Run 135002088   OneShot=OFF   dT=700 ticks (0x2bc ticks, 35ms)     

Observations:

• First event (e.g. event 0) show very uniform charge injection on 1/64 channels which are 'terminated' at TOT~0x80 (6.4ms), but some layers tend to systematically drag on somewhat longer up to the 0xff max.
• There is not TKR triggers as seen from both TEM diagnostics or GEM for the first event, as the charge injection was activated only at the CLOSING of the GEM window so that the trigger data latching for the first event was before the Qinj actually took place.
• The 2nd event (e.g. event 0, easily distinguishable from the GEM delta-event time) at 700 ticks later still have many TKR digis left over from the first event, and curiously most concentrated on strips neighboring the charge injection strips but not the charge injection strips themselves. It looks like the charge on the injection strips are controlled and systematically swung down after pulsing, but the charge were already spilled/crosstalked to the neighboring strips with significant pulse height to have lasted this far. There are significant variations from layer to layer on the extent of the 'spill' in the shadow of the first event.
• TEM diagnostic data and GEM condition summary consistently report the TKR trigger being on from the shadow. The TKR arrival time =0x1f at the GEM also indicate correctly that it was already on when periodic trigger opened the window. (Periodic=0x20 in condition summary, TKR=2).

Run 135002084   OneShot=OFF   dT=950 ticks (0x3b6 ticks, 48ms)  

Observations:

• First event still similar to 2088, but the effect of the Qinj shadow are mostly faded on the 2nd event and GEM TKR mostly no longer firing although a few individual layers are on.

Run 135002087   OneShot=ON STRETCH_OR=12   dT=700 ticks (35ms)

Observations:

• The behavior of the TKR digis are still very much similar to run 2088 at the same dT, but the TKR trigger at TEM diagnostic level and GEM are cleanly killed by the oneshot on the 2nd event even though the shadow activities in the TKR digis are still abundant.