Hello.
I have a few comments on the ABS tuning measurements done on
saturday and sunday:
The settings of the ABS target sequence include several
parameters. These settings are measured for different BLAST
field values, repeated over time, checked for consistency and
repeatability and logged into the ABScaServer. Any questions
about the performance of the ABS can be answered by looking at
the history. Before we jump to conclusions about the previous
ABS target sequence settings you have to make sure that you do
everything in exactly the same way as we did previously,
otherwise you cannot compare the results. The difference between
the recent scans and the previous scans are mainly (if not
entirely) due to a different hysteresis suppression sequence.
The recent scans are all taken with a different hysteresis
suppression sequence. I had a look at all new *.cycle files and
all cycle clockwise in the B-I diagram.
A good reason to stick to the old hysteresis sequence (setting
-15 before one sets the actual value, that is cycling counter
clockwise in the B-I diagram) is that we can compare the new
data with previous data: We have several scans for MFT and SFT
with good signals taken with the previous cell and better
statistics. Those scans give an important reference for the
scans that are presently taken with bad or impossible diagnostics.
That's also the reason why I did not switch to higher frequencies
on the MFT yesterday.
One important point is: Tools to measure the routine checks
on the ABS should be developed by the target group for obvious
reasons. Any work on the ABS must be coordinated with the target
group BEFORE it is done. That is contact Genya or me. Otherwise
you create a big confusion for everybody and risk valuable time
for data taking.
Hauke
On Sunday, July 6, 2003, at 01:18 AM, Electronic Log Book wrote:
> Operators: tancredi tancredi
>
> This shift was devoted to ABS studies, collecting data on MFT and SFT
> scans
> as described below.
>
> At startup, HK tuned the ABS rf transitions. He reported problems with
> ligit
> "bursts" and poor resolution in the compression tube ion gauge. Again,
> the
> CT is the only tool against which you can tune/check the SFT,WFT
> transitions.
>
> We investigated the issue of the CT ion gauge first: unfortunately
> instead of an
> analogue read-out we have to digitize an analogue version of the
> digital gauge
> controller read-out. Thus the gauge controller intrinsic ADC
> "resolution" may
> affect the final radout value. We essentially toggled between two
> settings: HIGH
> scale and a gauge calibration factor for D2, vs LOW scale and g.c.f.
> for N2 (this
> is the default we have been using). In the first case the pressure
> reading is
> multiplied by a factor 2.86 which we hoped could increase our
> sensitivity. The
> manual is unclear about the HIGH and LOW scale but the HIGH seemed more
> appopriate for pressures above 1e-7 (which occurs when the scale is
> set to D2)
>
>
> We then turned to the MFT, as the MFT leaves a signal in both ligit
> and the CT.
> The MFT can be tuned for 3-4 and 1-4 transitions. Each has different
> gradient
> and static magnetic fields. Data was taken with WFT=SFT=off, gradient
> fields set
> to -1.0 and -3.0 A (as recommented) and as a function of the static
> field in the 8.5
> to 3.5 A range. Before the static field was scanned, the magnet
> hysterysis was
> compensated for. The beam was off, blast and holding field were on and
> in "data
> taking" conditions. Each scan was repeated 3 times.
>
> The MFT static field is monitored by a gauss probe, and is cycled
> between -15
> and 15 A. At those extremes the gauss probe reads -13 and 57 Gauss
> respectively, which gives a clear indication of the direction of the
> BLAST field.
> Therefore we decided to use the following recipe for taking out the
> hysterisys:
> ramp the MFT to +15, -15, (+15, -15) and then go back to +15 and come
> down
> monotonically from the high side.
>
> Unfortunately this is done differently during data taking but I
> express my doubts
> that you can cycle to the -15 point (low field) and then set the field
> for any given
> transition. Also, since this was not critical for this measureement,
> we used a 2 s
> setlling time instead of 0.4.
>
> We then executed scans as described above, for both the MFT 3-4 and
> 1-4 (using
> the different settings for the grad field) and for two sets of gauge
> settings.
> At each value of the static field current we took 10 data points (1
> second apart)
> Results are attached in pdf, and shown as a function of time (left)
> and MFT current (right).
>
> In these plots the time scale corresponds to 100 s = 7.5 A, 200 s =
> 6.5 A , 300 s = 5.5 A. The 3-4 and 1-4 resonances are clearly visible
> in ligit. The ligit burps are
> also visible but they clearly do not pose a problem. Note that at each
> current
> we do 10 measurements (which is the cause of the vertical bars on the
> right side
> plots).
>
> The CT tube signal is more problematic: it is not always there and I
> would
> say in either scale for the gauge. We nevertherless chose "HIGH and d2
> scale
> factor" as our preferred setting since it seemed cleaner during the
> first MFT 3-4
> scans. Also, the CT-bckg gauge (ideally not sensitive to the sx
> effects) seems
> responsive to the incoming spin state, although this may only be an
> artefact.
>
> Two main conclusion form these studies:
>
> 1) The CT is still not in a position to help (with reasonable
> confidence) the tuning
> of the other ABS RF units. Therefore this tuning is ambiguous. The
> relative
> efficiency of the WFT and SFT units is critical for the combined V-T
> analysis
> we plan on D2.
> I suggest changing cell since with this cell we have no CT signal.
>
> 2) Contrary to what was set by hauke for data taking, on july 05 the
> 1-4
> transition does not occur at a static field current of 7.6. He
> pointed me to
> some data from april 03. Please see for yourself: the cycle
> programs I used
> are ~/pro2003/ABS/mft34.cycle and ~/pro2003/ABS/mft14.cycle. The
> data is
> plotted with the respective .kumac files (in paw, just hit enter
> and then type
> the name of the kumac. It is plain to see what input files are
> expected)
>
> The difference regarding point 2) may have to do with hysterysis curve
> in the BLAST field. But I wish to hold to the considerations made in
> this email. It would
> be interesting to compare with ion pol data for the 1-4/MFT-off.
>
> We then turned to the SFT: to do so we leave the MFT in the (better
> known)
> 3-4 transition and we cylce the SFT static field. The SFT 2-6 and 3-5
> transitions
> are expect for the same value of the grad field. A plot is attached.
> In this plot
> the scan is wider so 100 s = 9.5 A, 200s, 8.5 A and so on. The nominal
> transitions
> for the SFT 2-6 and 3-5 are at a current of 4.6 and 9.9 respectively
>
> During this scan ligit stays constant, as desired. Again burps are not
> an issue.
> Again data was taken 3 times. The CT "sees" the small signal from a
> transition
> although not reproducibly. Unfortunately we have a transition in both
> the
> 9.5-9.0 and in 4-3.5 range.
> I do not know how to explain this: the 3-5 transition should not occur
> since state
> 3 is already taken out by the MFT. The consistency of this data also
> should be
> checked against the ion pol data. For the SFT scan we used
> ~pro2003/ABS/CYCLES/sft26_mft34.cycle
>
> The hysterysis procedure was the same as of the MFT, but it may have
> to be
> different in case of SFT.
>
>
> In summary: spent time developing a tool for routine checks of the ABS
> transitions (thanks to VZ for giving me the blueprint and leave
> early). This
> required more time than expected and taking over the target and
> preventing
> data taking. On the other hand we wanted to wait long time for the Wch
> due to the fact the Wch had no He this morning.
>
> Such a check of the ABS can now be done quickly. It does not take much
> more
> then simply typing ABSsoft blabla.cycle. It is a few minutes per
> transition unit.
> Data is presented in ascii to the out-file defined by the "log"
> command in .cycle.
> We should plan on making this a standard routine. In particular these
> checks
> may be useful at some point to compare the health of the transition
> units and
> monitor "long term" stabilty effects.
>
> Found that the MFT 1-4 does not occur where planned, the SFT tuning
> has no
> clear signature. Therefore the target sequences for data taking seem
> junk. Discussed this with hauke at end of shift. However plan not to
> do anything until
> these scans can be confirmed (or not) tomorrow. In particular, they
> should be
> repeated tomorrow with the "previous" hysterysis cycle.
>
> More worried about the CT, after all.
>
>
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:29 EST