TOF Results


	Frank L. H. Wolfs

	for the Rochester Time-of-Flight Group

	PHOBOS Collaboration Meeting
	June 26, 2003

Outline


	What is new with the TOF?

		Summary of conclusions of PR03. Note: solutions will be discussed in DaveÕs presentation.

		Final analysis of Au+Au data: how well did we do, despite all the limitations we encountered in PR01 (relevant for PR04). The Au+Au study is the main focus of Erik JohnsonÕs thesis, to be defended on August 1, 2003.

		Status of analysis of d + Au and p + p.

		A brief update on the Universal Logic Module.

		Publications É.

Conclusions of PR03


	The main struggle during PR03 was the cross talk within and between twisted-pair cables, which was discovered after observing ÒweirdÓ correlations between T0 counters installed on opposite sides of the IR.

	Andrei Sukhanov was able to show that these correlations were due to cross talk between the two ribbon cables that carry the individual T0 signals to the CH.

	During the pp run, correlations in the T0 ORs were discovered, despite the fact that these signals arrived in the CH over Twinax cables. These correlations were due to the twisted-pair cables between the patch panel in the CH and the TDCs.

Conclusions of PR03


	In our study of different solutions for the twisted-pair cable problem, we additional cross talk problems related to the discriminators being used.
	A study carried out by the CLAS collaboration showed significant cross talk in the LeCroy 4413 discriminators, which are used for the T0 detectors.
	Andrei verified their basic observations, except that the magnitude of the effect was smaller.
	No cross talk was observed in the TOF discriminators.

Au + Au analysis: PR01


	The new tracking procedures, using the extended field maps, were applied to the Au + Au data collected in PR01.

	The width of the correlated peak is 0.75 cm (1s), which is about the width of a single scintillator.

	Hits within 2 cm of the extrapolated (horizontal) track position are considered in the analysis.

Au + Au analysis: PR01



	The background is reduced by correlating the vertical position of the TOF hits with the extrapolated vertical track position.

	This requirement increases the signal to noise ratio by about a factor of 2.

Au + Au analysis: PR01




	The energy loss in silicon (dE/dx) is highly correlated with the velocity obtained from the time-of-flight information.

	Monte-Carlo simulations are used to define this correlation, and specify the actual cuts applied.

Au + Au analysis: PR01


	Those hits that satisfy the previously described cuts are considered valid hits.

	The measured time-of-flight is converted to velocity (this allows all scintillators to be considered at the same time).

	The measured 1/v resolution for high-momentum tracks (p > 3.5 GeV/c) corresponds to a time resolution of about 200 ps (1s).

Au + Au analysis: PR01


	The ultimate resolution achieved in PR01 is limited by variations in cable delay over the run period.

	Although calibrations were carried out on a run-by-run basis, the observed drifts over the 30 - 60 minute period of a run.

	It is still remarkable that we end up with a resolution of 200 ps.

Au + Au analysis: PR01



	Based on the resolution we achieved in PR01, we can separate protons from Kaons and pions up to about 2 GeV/c.

	With a careful deconvolution of the velocity distributions, we may be able to push the limit a bit higher (up to 2.5 GeV/c).

Au + Au analysis: PR01


	The time-of-flight data were used to extend the range over which particle ratios were studied for Au+Au.

	The acceptance for protons and anti protons depends strongly on p_t and rapidity.

	Note: the blue data points are hits on wall TC while the black data points are hits on wall TB.

Au + Au analysis: PR01


	The proton ratio in the low pt region (pt < 1 GeV/c) agrees well with the ration reported based on the spectrometer alone:

		TOF: 0.72±0.01(stat.)±0.03(syst.)

		Si: 0.73±0.02(stat.)±0.03(syst.)

	The measured ratios are, within the error bars, independent of centrality and transverse momentum

d + Au: PR03


	The first step in the analysis of the PR03 data was the decission to give up on using the individual T0 counters to generate the collision time, but rely on the T0 OR information.
	The T0 OR information is digitized using both FB TDCs and VME TDCs.
	Clear correlations between the T0 ORs appear in the FB data, but not in the VME data.
	We should rely on the VME T0 ORs to determine the collision time.

d + Au: PR03


	Impact of relying on the VME T0 ORs:

		VME T0 ORs are not included in the InfoTrees.

		VME T0 ORs are not included in the calibration procedures (although the raw data contain the calibration information).

		Need to modify the T0 event structure since the OR is now used to determine parameters such as the collision time (and this information was not included in the T0 event).

d + Au: PR03


	Status:

		All required modifications to the software have been made.

		The final repass of the d + Au data will provide the required calibration information for the VME T0 Ors.

		Current resolution is about 300 ps, but this is based on an analysis of a series of runs ( ÉÉ ) with a single set of calibration parameters for the T0 ORs.

The Universal Logic Module (ULM)


	The first prototype of the Universal Logic Module has been completed, and two additional units have been produced.

	Wojtek received a DOE SBIR Phase 1 grant to develop a digitizing front end for the XLM-72 (this is a collaborative effort between SkuTek instrumentation and Jan Toke, and Wojtek will be on leave effective 9/1).

	Wojtek and I have submitted an NSF SBIR Phase 1 grant to develop a stand-alone, low-cost signal processing and data acquisition system, based on the ULM, aimed at the educational market.

Universal Logic Module:
Three parts need to be programmed!

Universal Logic Module:
Programming Status


	FPGA Firmware:
		Programmable delay on all inputs, 25É400 ns.
		Pre scaling on all NIM outputs by 1É255.
		Pulse area and baseline determined for every pulse.
		Programmable upper & lower threshold on all ADC inputs.
		Either self-triggered or externally triggered.
		Complete 12 microsecond history of inputs, outputs, and ADC.
	PHOBOS trigger firmware.
		One VHDL module implements PHOBOS-specific Òlogic equationsÓ.
		A few pages of well commented VHDL code.
		Will need to be customized Òas neededÓ.
	Embedded micro processor:
		Still below the limit of the free Keil compiler (90% of the 4kB limit).
		Communication with FPGA implemented.
		Setting the ADC gain and input offset implemented.
		USB command protocol implemented.
		Histogramming and board monitoring needs the full compiler.

Universal Logic Module:
PC GUI



	No memory leaks, no dangling pointers.

	The 1D graphics toolbox tested.

	The 2D graphics available, not tested yet.

	The USB interface and communication tested.

TOF-related Publications


	By the end of July we plan to have drafts of the following papers:
		The T0 counters (to be submitted to NIM): contains detailed performance studies of the T0 counters and their performance in Au+Au in PR01.
		The time-equalizer module (to be submitted to NIM): contains design concepts and stability studies.

	By the end of August we plan to have drafts of the following papers:
		The TOF array (to be submitted to NIM): contains detailed performance studies of the TOF counters and their performance in Au + Au in PR01.
		Particle ratios as function of centrality and p_t for Au+Au at 200 GeV (to be submitted to PRC)??