[Halla_pi0] (no subject)

[D. W. Higinbotham]

To be clear, we have done checks and we are not loosing half the data.

What does happen is if you make the chi2 parameters too tight you 
loose real tracks.  This is why online we usually just select coinc. 
events with HRS and a proton in BigBite and make sure we get a track 
for 90+% of the events.

With Jin Ge's help, we did this optimization by hand for the 
3He(e,e'N) experiments in about 15 minutes.  If you don't get to 90+% 
tracking, something is wrong.  It could have been that we didn't have 
the chambers setup correctly for pi0 as it was one of the first 
experiments, but I sincerely doubt this.

What is true is that we intentionally blind the chambers to pion and 
electrons by setting the voltages and thresholds such that we are only 
efficent for protons.  Thus you must make sure you are only looking a 
proton tracks.

Best Wishes,
Douglas


On Mon, 26 Jul 2010, richard lindgren wrote:

> Hi
>
> After we got off the phone, Khem did have some results to discuss but
> preferred to wait until next time to discuss them. Cole and I
> emphasized it was important to discuss anything he had worked on no
> matter how trivial he thought it might be.
>
> We discussed his previous inefficiency results again  and tried to
> correlate the lower inefficiency numbers (70%) results with the new
> plots he had put up on the web -  in particular the number of hits per
> plane plot. There seems to be some crucial number that when you
> surpass it the algorithm or whatever breaks down.
>
> I think Khem is going to try to understand why this is happening.
>  Cole suggested he also look to see what the tracking routine did
> when he had two or more simultaneous proton tracks. He should also be
> able to check this.
>
> I will look into our original background calculations for protons,
> pions and gammas that Pavel made for us. I want to see if the
> calculated proton rates are comparable to what we see in the
> scintillators. I remember doing this for neutrons and Pavel was high
> by a factor of 2 using a hydrogen target. Since this tracking
> efficiency is so rate dependent, we should try to understand the
> origin of the accidentals.
>
>
> I don't understand why Khem has to vary this "resolution parameter"
> for every run in order to find the tracks efficiently. This has to be
> discussed further.
>
> For much of the low Q2 data the tracking looks stable and about
> 85-90%.  This is where we ran with low beam current.1-3 ua.
> Khem is anxious to start looking at the cross section while continuing
> to troubleshoot the tracking.
>
> I am surprised that nobody else has looked into the tracking
> efficiencies. Sure if your measuring asymmetries it is canceling out
> but maybe your losing 30-40 % of  the data!!!!
>
> Richard
>
> _______________________________________________
> Halla_pi0 mailing list
> Halla_pi0 at jlab.org
> https://mailman.jlab.org/mailman/listinfo/halla_pi0
>