Run 6 EEMC Tower Timing Scan

02/26/2006 jcw

Plan:In 2005 we established the relative timing between the crates, and this should not have changed for this year. This year we simply need to sweep the TCD phase delay to optimize all boxes simultaneously.

2:30 am -- Begin taking data. Signal:background is 1:1 for blue and 3:1 for yellow. Blue backgrounds should be in time with collisions, so we procede with the scan. Backgrounds improved over the course of the data taking to 3:1 for both beams.

We were getting error rates < 0.5%, few header corruption warnings for TDC channel 6 per 100k events.

STAR Shift Log

Table 1 -- First crack at a timing scan

Run number	TCD Phase	N events	Comments
7057020	12	100k
7057021	15	100k
7057022	15	100k	oops
7057023	18	100k
7057024	21	100k
7057025	24	100k
7057026	12	10k	test
7057027	12	50k	test
7057028	27	100k
7057029	32	100k
7057030	37	100k
7057031	42	100k
7057032	52	100k
7057033	09	100k
7057034	06	100k
7057035	03	100k
7057036	00	100k

02/27/2006 jcw

Plan: Perform TCD phase scan (1 unit TCD phase = 1 ns) from 0 to +50 in 5 ns steps. Change crate delays by 10 ns from -10 to -40 ns.

The TCD phase for the towers was set at 12 last year.

Table 2a -- Second crack at a timing scan, TCD sweep.

Run number	TCD Phase	N events
7058031	0	100k
7058032	5	100k
7058033	10	100k
7058034	15	100k
7058035	20	100k
7058036	25	100k
7058037	30	100k
7058038	35	100k
7058039	40	100k
7058040	45	100k
7058041	50	100k

Table 2b -- Second crack at a timing scan, box delay sweep. TCD phase set to 0.

Run number	Intended delay	crate 1 FIFO/delay	crate 2 FIFO/delay	crate 3 FIFO/delay	crate 4 FIFO/delay	crate 5 FIFO/delay	crate 6 FIFO/delay
last year	0	0x13/0xf	0x13/0xa	0x13/0x5a	0x13/0x50	0x13/0x2d	0x13/0x23
7058054	-10	0x13/0x5	0x13/0x0	0x13/0x50	0x13/0x4a	0x13/0x24	0x13/0x19
7058056	-20	0x14/0x5f	0x14/0x5a	0x13/0x46	0x13/0x40	0x13/0x1a	0x13/0xf
7058058	-30	0x14/0x55	0x14/0x50	0x13/0x3c	0x13/0x36	0x13/0x10	0x13/0x5
7058059	-40	0x14/0x4b	0x14/0x46	0x13/0x32	0x13/0x2c	0x13/0x6	0x14/0x5f

03/01/2006 jcw -- Analysis of timing scans.

Cuts:

ADC > ped + 10 channels
N events > 0 (ie no cut)

Figure 1 -- Timing curves, relative to 2005 settings

It appears that crate 1 had (somehow) lost its configuration prior to the start of the run. Once I started reconfiguring boxes to change the box delays (i.e. the delta_T < 0 data points) crate 1 came back. THe panitkin plots were not accumulating enough events to be useful as a diagnostic tool (i.e. pedestals were actually showing up, we just weren't getting data above ped).

I'm somewhat concerned about the points at 0, and will retake them this evening during the MAPMT scans.

03/06/06

Analysis of tower timing scans taken on 3/3. See log entry. Table 3 shows the box delays configured during the timing scan sweep. The intention was to place all boxes at a point 19ns earlier than last year's timing. Unfortunately, for crates 1 and 2 I counted backwards from 100ns instead of 107ns after changing the fifo setting. This means that crates 1 and 2 were configured 26ns earlier than last year.

Crate 4 was unconfigured until the middle of the 2nd data point.
ADC > ped+15
N events > 0

Table 3 -- Box delays set during the timing scans.

crate	delay	fifo
1	0x60	0x14
2	0x5b	0x14
3	0x47	0x13
4	0x3d	0x13
5	0x1a	0x13
6	0x10	0x13

Figure 4 -- Timing scans for crates 1-3. X axis shows the TCD phase delay minus the change in box delays from last year. The blue line indicates the current TCD phase delay.

Figure 5 -- Timing scans for crates 4-6. X axis shows the TCD phase delay minus the change in box delays from last year. The blue line indicates the current TCD phase delay.