You now know that our detector can quickly measure the energy deposited by a charged particle passing through, using scintillation counters and photomultiplier tubes, and that it can also track the particle, albeit more slowly, using drift chambers. You also know what the two types of neutrino interaction look like in our detector. This is all you need to understand what NuTeV's trigger does and how it works.
The job of the trigger is to continually "look at" the signals from our 600 or so phototubes, and recognize "interesting" events. When the trigger sees an interesting event, it alerts our data acquisition computer, which then collects and records the data from photomultipliers and TDCs (time to digital converters, which measure the drift times for the drift chambers, as well as other important timing information). Interesting events include charged current and neutral current neutrino interactions, "straight-through" muons that pass through the entire length of the detector, and a few other things. The looking is done with fast pulse electronics modules, including amplifiers, discriminators, and logic modules.
You probably already know that an amplifier makes a weak signal stronger. It is often necessary or desirable to amplify phototube signals before further processing. Discriminators are devices that accept one analog input, from a phototube for instance, and output a standard digital "logic pulse" each time the input signal exceeds some preset threshold. Logic modules accept multiple digital inputs and perform operations such as "and", "or", "not", or "majority", producing an output pulse whenever the result of the logic operation is "true". A trigger is an assemblage of these modules connected in such a way as to act as a sort of "smart funnel", with hundreds of analog input signals arriving continually, and one digital output signal which is "true" if and only if certain conditions are met. It is this signal that activates the data acquision computer. (Actually NuTeV's trigger is really sixteen triggers sharing many of the same components - that is, it can recognize sixteen different sorts of interesting events, and thus has sixteen outputs.)
All of this must happen quickly; the same signals that go into the trigger are the signals we want to digitize and save if the trigger decides the event is interesting. To accomplish this, the signals go through a delay box on their way to the analog to digital converter (ADC). A delay box is nothing more than a box containing long cables - 200 nanoseconds long, in our case. The timing of the trigger is such that when it sees an interesting event, its signal reaches the ADC just in time to tell it to capture the phototube signals and send them to the computer. My first introduction to fast pulse electronics was also my first job at NuTeV, building a part of our trigger. It impressed me mightily that all cable lengths were measured not in feet or meters, but in nanoseconds. I find it impossible to imagine a billionth of a second, but it helps to remember that light travels about one foot in that much time!
Let's look at one specific example, the "charged current trigger". Recall that in a charged current interaction, a neutrino enters the detector, interacts somewhere in the target producing a muon, and the muon proceeds through the target into the toroid where its momentum is measured. The neutrino is neutral and produces no signal in the scintillation counters, and the muon is charged and therefore produces a signal. The trigger thus looks for no signal from the two most upstream counters, a signal from at least two of the four most downstream counters in the target, and a signal from the first third of the toroid. Whenever all these conditions are met simultaneously, the trigger puts out its "true" pulse. There is one subtlety here worth mentioning. Although the neutrino and muon are both travelling at nearly the speed of light, it still takes about 100 nanoseconds to travel the length of the detector - not a negligible time for our purposes. To compensate for this, the cables from the detector to the trigger room are cut at progressively shorter lengths, so that all the signals from one event do in fact arrive nearly simultaneously.
Once the data acquisition process is triggered, the phototube pulse height information from the ADCs and the timing information from the TDCs are read into the computer and written to a disk or tape file. The data in this file can then be accessed for analysis and the event reconstructed. A sampling of the events can be reconstructed and displayed on the fly from the control room, so we can get an idea that things are working properly, but most serious data analysis takes place offline, on large collections of events. If you would like to see an event I just captured (12/1/95), click here. Since we do not yet have our neutrino beam, this event shows a cosmic ray muon passing through a portion of our target. Speaking of the neutrino beam, I guess I should take you upstream next, and tell you how we will be using the Tevatron to produce a beam of high energy neutrinos for this experiment.
