Checking the Tagger T-counter high voltage
	D. Sober, K. Loukachine
	19 May 2000

At the start of the RadPhi experiment, we have set up the tagger high voltages
for T1-19 by the following procedure:

1. Start DAQ with configuration TSCAL.  This configuration was created by 
Vardan with only the scaler crates running, and is necessary in order that
the scalers are not gated off by the deadtime gates.

2. Start the program psd (Photon Status Display) on any clonxx. This program
causes the 10-second readings of the T counter scalers to be written to the
ascii file
	/usr/local/clas/parms/status/tg_dps.txt
each time the scalers are read.
The graphic display also shows the T-counter rates (select Ungated scalers
only) and was useful for finding the range of voltages needed.

3. Since no downstream devices were running, we changed half the counters
while normalizing to the sum of the other half.  The scalers read the 
left-right coincidence for each counter, so the HV for the left and
right PMT's was varied separately.

There were 4 sets of runs:
	Tleft 1-10 varied, others at nominal
	Tleft 11-19 varied, others at nominal
	Tright 1-10 varied, others at nominal
	Tright 11-19 varied, others at nominal
In each set, the HV was set to offsets of -200, -150, -100, -50, 0, +50,
100 and +150V from the nominal for that counter.

4. The nominal values were read from the gui before starting, and were as
follows:
	Left		Right
T1	1600		1650
T2	1650		1700 (*1600)
T3	1550		1550
T4	1550		1600
T5	1500		1650
T6	1500		1700
T7	1550 (*1650)	1650
T8	1650		1600
T9	1550		1550 (*1450)
T10	1650		1550
T11	2000 (*1800)	1600
T12	1550		1600
T13	1700		1550
T14	1550		1550
T15	1450		1800
T16	1650		1550
T17	1500		1650
T18	1500		1800 (*1600)
T19	1550		1600
The values with * were set by eye to be the new nominal values, to assure
that the knee was within the range of HV variation.

5. The fortran program /home/clasrun/radphi/tagread was used to read the
current contents of /usr/local/clas/parms/status/tg_dps.txt into memory, and 
sort the results for each counter by high voltage.
e.g. when varying T1-10 left,
wait for the first HV setting to stabilize, then
>tagread
 Enter file name for full data set:   full1-10left.dat
 (program reads tg_dps.txt)
 Enter comment for this run (optional)
 Enter voltage offset for this run (with or without decimal point)
 Enter file name for sorted data set: sort1-10left.dat
 Get another run? [Y}  --- wait until voltage stable, then <CR>
 ...
 Get another run? [Y]  --- after last voltage recorded, enter "n"

The file sort1-10left.dat contains unnormalized data like
T 1
    -200.0         1
    -150.0        52
    -100.0       130
     -50.0        72
       0.0     20503
      50.0     14985
     100.0     38276
     150.0     65455

T 2
	...
sorted in order of increasing voltage offset (for those who like connecting
their data points with lines.)

6. To normalize the data, we used the fortran program 
/home/clasrun/radphi/tagnorm .
>tagnorm
 File of sorted data to read: sort1-10left.dat
 Counters to analyze:      first, last = 1,10
 Counters to normalize to: first, last = 11,19
 Output file names have form nameroot.T01, ...
 Root of output filenames (include .):  tleft.  <-- dot required

The program divides the number of counts for each of T1,...T10 by the
sum of T11-T19 for that voltage run, and writes the values of HV offset,
normalized counts, and unnormalized counts to files tleft.T01,...,
tleft.T10 , one for each counter.

Run this program 4 times:
	sort1-10left.dat  -> tleft.T01,...,tleft.T10
	sort11-19left.dat -> tleft.T11,...,tleft.T19
	sort1-10right.dat  -> tright.T01,...,tright.T10
	sort11-19right.dat -> tright.T11,...,tright.T19

7. Plot column 2 (normalized counts) vs column 1 (HV offset) for each PMT 
using paw.

8. Inspect the plots and set the new HV approximately 50V above the knee.
Plots are avaiable in /home/clasrun/radphi/Tscalers.ps .

9. Table of new HV values
	Left		Right
T1	1600->1700	1650->1600
T2	1650->1650	1700->1500
T3	1550->1600	1550->1500
T4	1550		1600
T5	1500		1650->1550
T6	1500		1700->1600
T7	1550->1700	1650
T8	1650->1600	1600->1550
T9	1550		1550->1450
T10	1650		1550->1500
T11	2000->1750	1600->1550
T12	1550		1600->1550
T13	1700->1650	1550->1500
T14	1550->1500	1550
T15	1450->1500	1800
T16	1650->1600	1550
T17	1500		1650->1550
T18	1500		1800->1550
T19	1550->1500	1600->1500