[RASMB] measuring cell positions

[John Philo]

Regarding the comment that the interference optics appears to give the
average position for the sample and reference sides, it seems to me the
physics says the fringes can only exist where the light goes through both
sides, so their disappearance should measure the position of whichever side
is closest to the center of the channel rather than the average of the two.
Further, if the averaging idea was correct then the radial calibration
procedure for interference would not work, since the counterbalance is
designed to put the edge at 5.85 and 6.15 cm on only one side.

However, in theory it should be possible to image the two sides separately
by deliberately moving the delay setting (laser timing) so that only one
side or the other is actually illuminated. You won't see any fringes under
such conditions, of course, but nonetheless the superimposed image of the
cell should tell you the position of that channel (and probably with more
precision than the absorbance optics).

A good test for whether that idea would work would be to try to see the two
different edge positions (sample vs. reference) for the radial calibration
slits within the counterbalance. (I can't try it myself since I only have an
XL-A!).

John Philo
Alliance Protein Laboratories

-----Original Message-----
From: rasmb-admin at rasmb-email.bbri.org
[mailto:rasmb-admin at rasmb-email.bbri.org]On Behalf Of Borries Demeler
Sent: Friday, November 16, 2001 1:52 PM
To: RASMB
Subject: [RASMB] IF/vs. intensity scans


----------------------------------------------------------------------------
------
The older archived RASMB emails can be found at:
http://rasmb-email.bbri.org/rasmb_archives
and current archives at
http://rasmb-email.bbri.org/pipermail/rasmb/
Search All the Archives at:
http://rasmb-email.bbri.org/rasmb_search.html
----------------------------------------------------------------------------
------

To follow the suggestions given earlier by others on this list, we ran
the empty cells in both intensity and interference mode.

The scans were scaled and plotted in the same plot. A magnification of
one of the dark spaces between two channels is shown in a plot at:

http://www.biochem.uthscsa.edu/demeler/plot.gif

Several things become pretty clear from looking at this data:

1. Intensity scanning is the best way to determine the boundaries of a
channel.

2. either the alignment of the centerpieces is poor or the centerpieces
   are poorly cut, because there is a noticeable difference between the
   positions of sample channel (black) and reference channel (red).

3. The extra peaks seen in the absorbance scans are optical artifacts
   presumably resulting from the overlapping of the boundaries of
   reference and sample channels.

4. a similar effect is visible when looking at interference (shown in the
   bottom of the plot (URL above). In that case the boundaries for both
   channels seems to be "averaged" out.

I gather that it will be difficult to reproducibly determine the bottom
position of each centerpiece channel - arghhh.  As Tom Laue pointed out,
this position is probably even sensitive to cell housing, which rotor
and which rotor hole is used, although I haven't confirmed this yet.

Any suggestions on how to keep the position constant from run to run
or how to correct for these errors?

Thanks, -Borries
_______________________________________________
RASMB mailing list
RASMB at rasmb-email.bbri.org
http://rasmb-email.bbri.org/mailman/listinfo/rasmb