Hi,
A few comments follow my email last night.
The fact that pure TOF rate depends on beam current can not absolutely
exclude the possibility that WC performance is not affected than beam
current. Because the 2nd level trigger is dependent on WC. so a less
efficient WC would cut down TOF rate even though the analysis seemingly
does not rely on tracking information.
However, again looking at 2nd level trigger rate, there did not appear to
be a 20% change in counts/C at the change of current from 200mA to 100mA.
Moreover, looking at nsed, it is not clear that WC is much noiser at
200mA. looking at data, it does not appear that # of WC hits in each event
were different at 200mA than ata 100mA. other indicators such as # of hits
per track and so on also do not suggest that WC performance is any
different at different current. So I found it hard to believe that WC is
the culprit.
Turning to the target, we observe beam current dependence in rate even in
unpol runs. So I found it hard to say that ABS performed differenty in
2005.
It seems to me to be something at the storage cell. How certain are we
that the cell temperature is independent of the current. a surge of cell T
from 100K to 160K would produce a ~20% drop in rate. That
seems to be the easiest way to "explain" things we saw.
Unfortunately, cell T epics changed several times during the last couple
years quite arbitrarily. this means for quite a big number of runs, cell T
is not properly recored in ntuples/dst's because the program went query a
deprecated epics channel. I also recall hot debates on whether certain
temperature readouts truely reflected actual cell temperature but I do not
know the conclusion.
Chi
On Mon, 5 Sep 2005, Renee H Fatemi wrote:
>
> Hi Vitaliy,
>
> I looked at % of events removed for each cut in the 2004/2005 (e,e'p)n
> analysis. I found that 2.2/2.4% of events survived exactly the same series
> of cuts (no MM or kinematic corrections). Yet the rate of ep_coin/kC for
> each year is 19.36/16.52 -> about a 15% reduction. From this the
> difference seems to be in the charge. We are looking at different channels
> but I would be interested to see if your analysis give the same results,
> ie similar reduction in #events from 2004-2005 but still a different rate.
>
> One other difference is that the rate of ep_coin in 2004 was equal for the
> R/L sector but in 2005 the electron L side had about 20% for events than
> eRight.
>
> Just another clue...
> Renee
>
> P.S. I chose runs 15392-15398 from 2005 which are after the charge
> calibration changed, so this shouldn't be an explaination/issue.
>
>
> On Fri, 2 Sep 2005, vziskin wrote:
>
> > Dear all,
> > I include two graphs similar to the ones in my thesis for the 2005
> > data. The first is the rate per coulomb for each run and the second is
> > the histogram of the rate. The mean rate is about 0.76 \pm 0.1.
> > Compare that to 1.05 \pm 0.11 from 2004. Correspondingly the mean rate
> > in 2005 is about 72 % of 2005. Of course the mean rate is not average.
> > It is if the rate is not current dependent (like in 2004). So I expect
> > the average current to be even lower. One thing to note is in the first
> > plot the rate is higher at the beginning and the end of the run period.
> > These are the times when the injection current was kept lower. So,
> > where is the missing rate.
> >
> > Best regards,
> > Vitaliy
> >
>
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:32 EST