On June 7 afternoon about 2:30-4:30pm, at the end of a physics store, CAD provided low luminosity collisions in the STAR IR upon our request. There was one low luminosity calibration run happened on April 10. This request near the end of the current AuAu 200 GeV run is to further check the HFT alignment over a long period. The request and some information about the early low luminosity run taken on April 10 are summarized in this following document.
http://www.star.bnl.gov/protected/heavy/dongx/hft/Run14/Trigger/HFTSoft_LowLumi_06022014.pdf

The low luminosity runs started on ~2:30pm June 7 when the CAD lowered the STAR ZDC rate to about 1kHz by mis-steering beams. Here is the plot of the ZDC coincidence rate vs. time.

Carl Gagliardi nicely summarized the observed correlation between the ZDC rate with the vpd-10 rate, and how to better coordinate with the CAD to keep the reasonable vpd-10 rate.
http://www.star.bnl.gov/HyperNews-star/protected/get/starops/5384/1.html

We looked at the PXL hit response offline with the standard offline chain + PXL/IST data chain. Three runs (15158046, 15158047, 15158048) were used in the following plots compared to the runs taken on April 10. These three runs were taken under the ZDCx ~ 2-3kHz, while the vpd-10 rates were typically 30-40Hz. 

	run 15158046-048 (ZDCx ~ 2-3kHz)	run 15100093 (ZDCx ~ 5kHz)

It clearly shows the UPC electron and pileup hits in the low luminosity runs taken on June 7 are under much better control compared to those taken on April 10. The original plan is to have the PXL1-TPC correlation plot offsets at <200 nPXL1Hits. The vpd-10 rate is a bit lower than what we requested (three runs were running at the average vpd-10 recording rates of 28, 37, 33Hz). I feel with the similar setup, we can still tolerate more collisions to keep the vpd-10 rate to be around 60-100 Hz, so we can get ~300k events in about 1.5 hours.

The observed ZDC rate was higher than we expected, because CAD was simply mis-steering beams and then pulled many collisions happening at high Z region. Here are the online L3 Vz and ZDC Vz distributions.

I think this is generally fine as long as those collisions are far away from the PXL detector. 

So, to summarize:

1) The PXL hit densities from the low luminosity runs (test) taken on June 7 satisfy our request for the HFT internal alignment calibration.

2) We collected ~80k vpd-10 events with (PXL+IST+SSD) in these three runs within about 40 mins. Still need more statistics for the calibration purpose. The request is to have 300K for field on and 300K for zero-field runs.

3) We may increase the collision rate a little bit to have a reasonable vpd-10 rate (targeting at 60Hz on average). To get there, we may ask CAD to increase the ZDC rate to about 3-4KHz according to Carl's observation assuming the beam conditions are similar next time.