From: Borries Demeler/Biophysics <demeler@selway.umt.edu>
  To  : hansen@bioc02.uthscsa.edu
  Date: Mon, 9 May 1994 17:25:23 -0600 (MDT)

Message to John Philo

>We know the intensities by asking for the intensity data in addition to the
>absorbance data.  What I have noticed is that this slope seems to get worse
>as the lamp gets dirty and the intensities in the far uv drop off.
>
>The reason why both Jack and I mentioned this is that this is the behavior
>you would expect if there is a zero offset in the system somewhere, such that
>zero light through sample or reference does not give zero net signal.  Yes,
>in principle the xl-a is supposed to take care of all this and do a careful,
>correct photometric conversion.  But does it?  Who knows?  My queries to Palo
>Alto about EXACTLY how the data is collected and processed have been
>rebuffed.  I strongly suspect that some compromises were made that Beckman
>doesn't want us to know about.

One of the things I remember when I was in palo Alto and talked to 
Les Holladay was what he told me about the design of the monochromator.
There is a half-way mirror built in that feeds part of the light back to a 
feed-back control system that supposedly regulates the intensity of the 
light flashes. I'm not sure if I remember this correctly, but it kinda makes
sense. The only other person besides Les who seems to know what is going
on is Barney Hedges. I suggest you try getting him on the phone.

>It is easy for everyone to say " if your instrument is aligned properly"
>and/or "get Beckman Service to come and fix it".  The reality is that I have

Of course! I know how Beckman reacts when we ask them to fix something,
you and I are wearing exactly the same shoes! At one time they told us
we didn't know anything about spectroscopy and had to learn it, when we
told them that the spikes caused by missed lamp-flashings were unacceptable.
These people can be incredibly annoying.

>been trying to do just that for a year.  If users are so demonstrably willing
>to dismiss the reality of what Jack and I are seeing, and to accept that we
>have done the necessary controls, then is it any wonder Beckman is sticking
>its head in the sand?  You also may be interested to know that our other XL-A
>at Amgen was down for nearly a year while Beckman blamed samples, computers,
>networks, and everything and anything besides the instrument, for the fact
>that it would seldom complete a scan without hanging up.   

Let's not misconstrue the facts and fight among eachother, nobody is trying
to dismiss the problems Jack and you are seeing. I am the first one to 
believe every word you are saying about the shortcomings of the XL-A.
However, I was just making some suggestions on how to rule out one common
problem which definitely is REAL and has been seen on other machines as well.
Consider it an attempt to help you guys out. And, don't you think we are the
very first ones to get on Beckman's case when it comes to improving the machine?

It is exactly this kind of double checking which we are suggesting that takes
the wind out of Beckman's sails. Beckman is sticking their head into the sand
because they are clueless, and unwilling to put any more money into the XL-A,
not because the users haven't been able to determine the cause for the problem.
Let's get real, it is Beckman's job to debug a machine for which they are 
charging more than $140,000, don't you think? And with respect to the AMGEN
machine, they used the same kind of excuses on us, if you find that comforting.

However, Jeff Hansen is not the kind of guy to take no for an answer, so 
after many different tries we have a couple of machines that work reasonably
well. Call it good luck, call it persistence, we actually do Biochemistry here
with the XL-A's. We must have gotten about 4 or 5 different monochromators
before we got one that worked. I don't remember how many data acquisition boards
they put in and how many times they updated the eproms on the board. Also,

Just to show you how much I would like to help you out with your problem, 
I offer to run your sample on our machine and see if the problem can be
reproduced on our XL-A.

Let me know if I can be of any further help.

Regards, -Borries

P.S. FYI, I am working on a baseline subtraction routine, and I will make this
available to everyone as soon as I have a good running version.