Hi Tim,
 Thanks for your comments. I was looking for you to discuss the physical geometry
to see what constraints there are on the slot/hole module coplanarity. Although the 
slot module statistics are low the differences looked  statistically significant. And those
comparisons had layers 1-2-3 in common. The projections back to the wire are roughly 
2.5 meters, so it wouldn't take much.
Your comments about the opening angle are spot-on. On my list for today is to fit
modules 1-2-3 and project into the four quadrants of layers 4-5-6.
Norman

-----Original Message-----
From: Nelson, Timothy Knight 
Sent: Friday, July 24, 2015 10:10 AM
To: Graf, Norman A.
Cc: Nathan Baltzell; [log in to unmask]; hps-software
Subject: Re: [Hps-analysis] July 24 DAWG Meeting

Interesting, Norman.  I'm not sure I understand the non-coplanarity comment on the last slide.  Hole/slot sensors can't be in different planes by very much, a couple hundred microns at most dominated by non-flatness of the half modules.  Also, anything that find shifts that are bigger than about 100 microns in x-y (I see >mm lots of places) or a few hundred microns in z from the geometry needs to be looked at very carefully, since that flies in the face of all other data, even if you are using the v1 detector.  Shifts even that large in v2 would indicate problems in the survey measurements or mistakes in implementing them in the v2 detector (neither is completely impossible).

What's being done for multiple scattering errors here?  If you want to get meaningful results (e.g. unbiased residuals with a layer left out) you really need to run GBL or change the way multiple scattering is dealt with in seedtracker. Otherwise, relative misalignments in layers 1-2, which are close together (by ~an order of magnitude relative to the full SVT length) and are assigned small errors in seedtracker, will dominate the track parameters and bias what you see everywhere else.  Obviously, the former (using GBL) is much more desirable.

Finally, once GBL is working, probably the first thing that needs to be done (both here and with field on data) is use the data to figure out the actual opening angle of the detector, since uncertainty in that is now expected to be the dominant alignment uncertainty by far.  Without having that right, trying to point to the wire is fraught with difficulties and you can really only look at internal consistency.

Cheers,
Tim

> On Jul 24, 2015, at 9:05 AM, Graf, Norman A. <[log in to unmask]> wrote:
> 
> Hello All,
> Apologies for the SNAFU this morning. I've posted my slides to the meeting page on confluence.
> Bottom line is that the data seems to be of sufficient quality and 
> quantity that we should be able to derive some alignment parameters 
> for the SVT independently of the field-on data. I'll give an update next week. But in the meantime, please let me know if you have any questions or comments.
> Enjoy the weekend,
> Norman
> ######################################################################
> ##
> Use REPLY-ALL to reply to list
> 
> To unsubscribe from the HPS-SOFTWARE list, click the following link:
> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1

########################################################################
Use REPLY-ALL to reply to list

To unsubscribe from the HPS-SOFTWARE list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1