Star-fst-l mailing list

[videbaek <videbaek AT bnl.gov>]

Hi Ziyue

Thanks for the expanded sets of plots. It seems to me that there is not 
much systematic variation with
run time/run-index. We are left with having a large set of APV with a 
main (max) peak in tb-1 and another group that has almost equal amount 
in 0 and 1. As I understand there is no way to get rid of that phase 
differece (< 1 tb).
From the 9 tb we learn that the primary response is in fact properly 
times and 2) that there are likely some triggers that give riseto the 
pre-hits, but that is not something FST has to take care of.

If you create by 9 tb plots either for tomorrow or Sunday, I will 
strongly suggest to go back to 3 tb.
And think about if need be if the ADC value is important if the ADC 
deduced can be created from some weighting of
tb 0 and tb 1.

my thoughts on this
Flermming




On 2023-06-23 17:16, Ziyue Zhang wrote:
> Hello Gerard,
> I'll try to make the plots in the early morning and pass it around and
> see if people can gain some opinions on whether we shall switch back
> to 3 tb "right away".
> So far I don't remember hearing people push this in the meeting, but
> you are very welcome to join if you have time.
> Ziyue
>
> On Fri, Jun 23, 2023 at 4:45 PM Visser, Gerard <gvisser AT indiana.edu>
> wrote:
>
> > hi Ziyue,
> > Great, thanks! I think your plots will certainly give everyone
> > all they need to contemplate if there's a problem. My own opinion at
> > the moment is that things are probably working fine.
> >
> > Until when do you & Zhenyu think that we continue 9-tb running.
> > [If you expect this needs discussion in 10AM meeting tomorrow, I
> > could join. But for now I figure that's not the plan.] If the
> > deadtime doesn't hurt too much, keeping this going to monday makes
> > sense to me.
> > I will be completely unavailable from tuesday ~3PM til
> > wednesday ~8AM.
> >
> > Gerard
> >
> > -------------------------
> >
> > From: Ziyue Zhang <zzhan70 AT uic.edu>
> > Sent: Friday, June 23, 2023 4:29 PM
> > To: Visser, Gerard <gvisser AT indiana.edu>
> > Cc: Landgraf, Jeffery M. <jml AT bnl.gov>; Videbaek, Flemming
> > <videbaek AT bnl.gov>; L Star-fst <star-fst-l AT lists.bnl.gov>
> > Subject: Re: [Star-fst-l] FST with 9 Time Bins Diagnostic
> >
> > Hello Gerard,
> > I've already had the 9 time bin plots , but after discussion with
> > Zhenyu, I need to make some adjustments:
> > 1. I did not exclude bin 0-1; we decided to exclude bin 0-1 and 8,
> > as the obvious reason you suggested
> > 2. Calculate the following fractions: bin 3, bin 4 and bin 5's
> > fraction among bin 2-7. This should correspond to the fraction of
> > bin 0,1,2 's (3 tb setup) fraction in the entire signal
> > 3. Plus additional value: mean of the tb distribution; total entries
> > of the distribution (both exclude bin 0,1,8)
> >
> > Jeff has helped me set up the working machinery on the online farm.
> > I'll get the latest plots before tomorrow's 10 am meeting. In the
> > attachment is a sample plot for the 3 tb running period, with
> > additional 15 runs (100->115) compared to last plots I showed.
> >
> > Ziyue
> >
> > On Fri, Jun 23, 2023 at 2:50 PM Visser, Gerard
> > <gvisser AT indiana.edu> wrote:
> >
> > Hi Ziyue,
> > I understand there were problems with the way you had
> > generated your nice plots before... Is there a feasible correct way
> > to make such plots? I think it would be very interesting to have
> > this to see the stability or instability of which is the max timebin
> > in the current 9-timebin running.
> > Because of the sorta-known wrong trigger timing issue, doing
> > your plots both ways where you consider all 9 timebins and where you
> > simply exclude the first two timebins from your analysis and only
> > look at the last 7, may be good.  [BTW note that if wrong-trigger
> > timing issue is variable, that is a source of variation in which is
> > max timebin of three, at least a source of variation between 1st and
> > 2nd, and potentially contributes to the thing we've been worrying
> > about.]
> >
> > I suppose we should try to decide soon when we have had enough
> > of the 9-tb data and will switch back.
> > Switching back is best done during any planned 45 minute
> > no-collisions opportunity, when there will be someone experienced in
> > the control room to correctly set the busy parameter back.
> > It's debatable but 6 timebin (or 5 but the deadtime is same as
> > with 6) is also possible. For such I would need to prepare firmware,
> > but it's feasible.
> >
> > Sincerely,
> >
> > Gerard

--
Flemming Videbaek
senior scientist, emeritus
videbaek @ bnl.gov
Brookhaven National Lab
Physics Department
Bldg 510D
Upton, NY 11973

phone: 631-344-4106
cell     :  631-681-1596