From: Jack Correia <jcorreia@umsmed.edu>
  To  : rasmb@bbri.eri.harvard.edu
  Date: Fri, 14 Jul 1995 10:37:47 -0500

Re: Cleaning optics

     Since the first few months of the XLA being in the lab we have tried to 
clean the optics once a month.  Our procedure does not use tooth paste.  
First I blow off the slit with canned air 5-6 bursts at a slight angle 
blowing in from the outside to avoid blowing it off.  This alone often 
improves the light intensity due I guess to a build up of oil?  To clean the 
lamp I remove it from the machine, loosen an allen screw, expose the flash 
lamp face and rub with isopropanol and a kim wipe.  You are trying to remove 
a burn spot at the point the light leaves the lamp.  Do this until you 
appear to succeed, and reassemble.  Be careful to mark the housing so you 
tighten the top to the exact position it came from.

     The diagnostic for doing this was noisey data to the eye.  This 
obviously can be buffer and/or wavelength dependent so we established a more 
quantitative criteria, ie the one described in the service manual - part 
number 679045.  [I recommend buying it!]  Run a wavelength scan at fix 
radius on a windowless cell at 3K and check the intensities.  If they are 
down significantly (<50%) than clean the optics.  Also run a radial scan at 
280 nm on a windowless cell at 3K and check % variability at 6.0 vs 7.0 cm.  
If >20% adjust lamp position for more even illumination.  Our lamp has an 
intensity of approximately 25000 just below 230 nm.  So that is the criteria 
we look for.  Note that this apparently varies dramatically between lamp.  A 
newly installed lamp at LSU has intensities of 60000!!  This will also make 
you realize that light intensity is more important than wavelength when 
making choices about what wavelength to scan at!  Our radial scans vary by 
12-17% decreasing from low to high position.  I have always wanted this 
trend reversed but have not been able to do so.  We adjusted it one day to 
8% by rotating the lamp to get constant illumination or a business card in 
an empty lamp hole.  However, we now had so much light that the meniscus now 
often lacked a sharp peak but rather approached a constant value slowly, 
apparently due to stray light.  So we also did stray light test with 8.3 
gm/l of NaBr or 9 gm/l of KCl.  See manual!  We ended up reducing the light 
intensity and getting back to 15% variability.  Note this means the LSU, 
60000 lamp is probably not good, too much light, and should be "tuned" down 
to a more reasonable intensity level.  After cleaning we always rerun both a 
radial and a wavelength scan to be sure we made it better and not worse!!

     So as I wrote this message my tech came in to tell me the sed vel scans 
are very noisey at 260 nm (a PO4 buffer so not likely).  So we are doing 
scans to prepare for a lamp cleaning as I "speak".  This happen quickly so 
it could obviosly be something else.   

 -------------------------------
| Dr. Jack Correia              |
| Univ. of Miss Medical Center  | 
| 2500 N. State St.             |
| Jackson, MS, 39216            |
| (601) 984-1522                | 
| fax (601) 984-1501            |
| correia@fiona.umsmed.edu      |
 -------------------------------