We would like to make a "heat map" of the supersector response to particles passing through a particle location within each tile.  There will surely be some variation, and some experts (Les, Gerard) have estimated it could be as high as a factor of 2, depending whether the particle crosses near to the fiber, at the corner, or in the middle of the tile.  Note that this is not "actionable" information.  I.e. we will not change construction of the detector (obviously) or operate it differently.  It will rather be used in the simulation and may feed into our systematic uncertainties if we do ADC weighting.

There are several ways to produce such a heat map.

Cosmics
In principle, the best approach would be to use true MIPs-- i.e. cosmics-- with some external trigger to pinpoint where the cosmic crossed the detector.  This can be a super-slow process, though Les and Prashanth almost had a test-stand going at BNL, using two scintillator rods with relative timing read out from each end of each rod.  Still, it would be really slow, because you want to look at a real ADC distribution (at least 1000 events) for each position in the detector.  We do not discount this, and may follow up on this approach down the road.

Triggered particles from a source passing through the detector
One can also use a strong radioactive source.  In particular, Sr90, whose decay chain emits a beta with high enough energy to punch completely through the supersector and be detected on the far side.  We have a 200-muCi source and external triggers that give clean signals.  We used a detector loaned to us by Wlodek Guryn (simply called the "Wlodek" detector here at OSU), which is about 3"x3", read out on two sides by PMTs.  The problem here, as above, is rate.  In a single measurement:
When there is no radioactive source under the detector, Wlodek counts 130 times in 5 minutes
When we have the source below the SS and Wlodek above the SS, Wlodek counts 160 times in 5 minutes

This means that the additional rate (above cosmics) through Wlodek is about 0.1 Hz.  Too slow, and too high background, to use for a full map.  We may revisit this with a much smaller trigger detector that Joey Adams has built.

Increased dark current due to the source near the detector
One may also look at the increase of the dark current in a given tile, due to the presence of the (collimated) radioactive source right at the surface.  Our SiPMs, which have been used for a high-luminosity pp run at STAR, have dark currents on the order of 1 muA.  When we place the source at the surface of a tile, we find that the dark current increases by about 200 nA.  This is certainly observable and we found it to be repeatable.  This test would be a lot easier, since it does not require a trigger and is fast.  The downside is that one is not looking at "punch-through" particles (though they are not MIPs in any event) and might be sensitive to the first several millimeters of the detector.  Still, we are basically looking for differences when the light is generated in one position versus another position.  Furthermore, we can easily get a feeling for this by flipping the detector over and re-run the heat map.

We plan to produce a heat map in this manner, in the next week or so.

It makes sense to ask whether 200 nA is something we should reasonably expect.  This depends on SiPM gain, photons per particle, source activity, etc.  Here is a quick "Fermi calculation" that indicates that a few hundred nA is indeed reasonable: