Details on NMR lock

-check if the dipole current is above 22A. The integral setup can't work for lower currents (22A $\sim$ .043T $\sim$ 517MeV), due to the NMR probe limitation.

-check if the probe is at a central position, corresponding to Zd $\sim$ 1604mm. It should be at this position if no special motion was ordered since the previous integral measurement. If not, enter 1604 RETURN in the "set Zd position" input field, wait for the end of the motion (look at cursor and position readback labelled "out") and wait for the NMR lock for up to 1 minute. Note: due to the software, the position readback may be updated $\sim$ 10 seconds after the real probe motion. Zd $\sim$ 0mm is for the upstream position, $\sim$ 1604mm for the central position and $\sim$ 3208mm for the downstream position. If the probe does not obey, call Arun. If Arun is not available, reboot the ioc (see "integral ioc reboot" below)

-check if the NMR probe selected is the good one. The system has 4 probes to cover the field/energy range [0.043T/0.517GeV,1.05T/12.63GeV]. The software selects automatically the probe from an estimated field and the following range table (from file "magnet.dir"):

mini	maxi	probe
T	T
0.043	0.1254	# 1
0.1250	0.250	# 2
0.177	0.470	# 3
0.338	1.06	# 4

The estimated field is computed from the "set" current and the coefficients given above. Note that there is an overlap between the range of each probe. In the standard mode ("auto selection"), two selection algorithms are used:

-initial selection: select the probe in which the estimated field will be the most "centred"

-routine check: change the probe only if the estimated field is outside the selected probe range.

Initial selection is executed at boot time, at each integral start and when you leave the "manual selection" optional mode.

In some cases in the past this algorithm did not select the best probe. If it occurs again, force the probe selection by opening arc_nmr.adl medm screen,then switch to manual selection mode, click on the probe you want to select and wait for NMR lock. Don't forget to return to the "auto selection" mode before leaving the integral measurement. Report about the problem in the e-logbook.

-expert users can also use nonstandard search modes and DAC values as proposed by the arc_nmr.adl screen ("Dac Value", "Dac+-5%" and "All Probes Range"). Note that "Dac Value" needs a scope. They are asked to restore the standard mode ("SeLected Probe Range") at the end. The DAC to field relationship used by the software is linearly interpolated in the table (from file "magnet.dir"):

dac=0	dac=1000	dac=2000	dac=3000	dac=4095	Probe
T	T	T	T	T
.040362	.05388	.07443	.104184	.132404	# 1
.080713	.107771	.148845	.208340	.264790	# 2
.161430	.215539	.297680	.416680	.529570	# 3
.322870	.431112	.595471	.833455	1.059210	# 4

Note: as the 9 $^{th}$ dipole has a uniform field, there is no gradient coil around the probes and the lock is in general easier to obtain than with the HRS dipoles. Nevertheless some difficulties may occur at low field due to the hysteresis gradient. The NMR system is the same as for the HRS (Metrolab) but the EPICS implementation is different. The field polarity is "+". Some lock problems may apear in the [60A,70A] range. In this case select the "Dac+-5%" option, compute the approximate field as explained above, estimate the corresponding DAC value from the table above (depending on field value and probe #), set this estimated value in the arc_nmr.adl screen, correct this setting by using the "+", "++", "-", and "-" buttons up to have the "OUT IN T" value close to the field you computed, and then wait for the lock. It should work, you have 100s to get the lock at the begining and at the end of the integral sequence! The best is to get the lock by this way before starting the integral sequence.