Index:
[thread]
[date]
[subject]
[author]
From: John Philo <jphilo@mailway.com>
To : H. Olin Spivey <ospivey@bmb-fs1.biochem.okstate.edu>
Date: Mon, 27 Mar 2000 09:30:42 -0800
RE: Aberrant Intensities
This is a multi-part message in MIME format.
------=_NextPart_000_0000_01BF97CF.1EE4D180
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Olin and the group,
I'm not sure I have any definitive answers, but this issue of lamp
intensities is one where there is a lot of disagreement among labs and a
certain amount of confusion, so I'd like to add a few thoughts and
observations.
First, it is important for everyone to remember that the absolute intensity
numbers are meaningless. The same absolute amount of light might produce
10000 counts on one machine and 40000 on another. If you've been swapping
parts like photomultipliers around then you also can't compare numbers
before/after parts swaps.
While I agree with Eric Anderson's comment that your intensity vs. radius
profile shows a much larger variation than most (30-40% variation seems
common), you haven't actually given us the key piece of information: do you
actually see a significant variation in noise level across the cell?
In my experience a 2-fold drop in intensity at 280 nm, and even a 4-fold
drop at 230, has very little effect on the noise level. Remember too that in
many cases much, if not most, of the 'noise' in a real experiment is due to
radial variations in the windows rather than real photometric noise. For
example, without windows on my current machine the r.m.s. deviation from a
flat line for a single replicate is ~0.002 AU, but with windows it is
~0.005.
To me the two most aberrant things going on in your machine are (1) that it
takes several attempts to get your lamp clean, and most puzzling of all (2)
you report that when the lamp is dirty your sample and reference intensities
diverge (implying a significant absorbance offset even with no windows!).
With regard to (1) I can only echo Eric's comment that if I use a pencil
eraser until I can no longer see the 'bulls eye' on the lamp face then I get
back full intensity in one cleaning. Are you perhaps using some solvent or
cleaning tissue that leaves a residue?
With regard to sample/reference divergence, if I understand you correctly
this is stronger when the intensities are low, which might suggest a stray
light problem of some sort. Have you done the stray light tests?
Unfortunately the lamp positioning which gives the maximum intensity and/or
most uniform radial profile also often gives more stray light, so the best
trade-off is often gets you maybe 80% of the maximum possible intensity but
much, much less stray light.
Also, remember it is important to avoid reflections back up into the
monochromator that will fool the lamp intensity monitor---never, never use
the old shiny model E screw rings, and use a black marker pen to cover shiny
spots that appear on your black screw rings and window holders with use.
The only other thing I can think of that would cause the sample and
reference intensities to be different would be something causing the light
to hit the sides of the cell, partially blocking one beam or the other. One
possibility along those lines would be that your cells are not properly
aligned exactly along a radius. Be aware that on some rotors the scribe
marks are not in the correct positions and will cause such a misalignment.
The other possibility would be a lamp timing problem (possibly a bad magnet
on your rotor?). One way to test for either of these beam blockage
possibilities would simply be to scan an empty hole rather than use a
windowless cell and see whether the sample/reference difference disappears.
Best regards,
John Philo
Alliance Protein Laboratories
www.ap-lab.com
-----Original Message-----
From: H. Olin Spivey [ospivey@bmb-fs1.biochem.okstate.edu]">mailto:ospivey@bmb-fs1.biochem.okstate.edu]
Sent: Saturday, March 25, 2000 2:25 PM
To: RASMB@BBRI.ORG
Subject: Aberrant Intensities
3/25/00
>Colleagues,
>
>For about the last two years, we have had a persistent problem with our
XL-A instrument. Radial intensity scans on blank cells (cells with no
windows spun at 3,000 rpm) decrease in intensity from the beginning to about
the middle of the radial range and then increase again to the end (cell
bottom). Examples of our recent intensity I vs. r and I vs. lambda scans can
be seen at
>
>http://osucau.okstate.edu/plot1.gif
>and
>http://osucau.okstate.edu/plot2.gif
>
>Two separate Beckman service engineers have invested several days of
effort and replaced virtually all components in our instrument to cure this
problem, but to no avail. We agreed to live with this situation since we can
get (after vigorous cleaning of the lamp) the sample and reference beams to
superimpose, and the intensity at the minimum position (about 10,000 at 280
nm) has been acceptable for our needs. However, these signals deteriorate
with use giving both lower intensities and separation of the sample and
reference beam intensities (variable ratios with r). This forces us to
invest many hours of lamp cleaning (one attempt is rarely sufficient) after
every few experiments. This requirement is more frequent for us since the
light intensities near the middle r values are so low to begin with. Current
intensity at 228nm is 19188 and at 542nm is 1775. Intensity at 280nm is ca.
19679 at both 5.9 cm and 7.1 cm, but drops to 7932 at 6.53cm (refer to
graphs above). Our questions are:
>
>1) Is anybody else experiencing the same problem? If so were you able to
alleviate it, and if so how?
>
>2) What could be the cause of this aberrant behavior?
>
>Since this problem is seriously reducing our productivity, we would be
grateful for any suggestions.
With much appreciation,
Frank Hays and Olin Spivey
H. Olin Spivey Phone: (405) 744-6192
Dept. Biochem. & Molec. Biology Fax: (405) 744-7799
246 NRC Email: OSpivey@Biochem.Okstate.Edu
Oklahoma State University
Stillwater, OK 74078-3035
------=_NextPart_000_0000_01BF97CF.1EE4D180
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content=3D"text/html; charset=3Diso-8859-1" =
http-equiv=3DContent-Type>
<META content=3D"MSHTML 5.00.2722.2800" name=3DGENERATOR></HEAD>
<BODY>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>Olin=20
and the group,</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>I'm=20
not sure I have any definitive answers, but this issue of lamp =
intensities is=20
one where there is a lot of disagreement among labs and a certain amount =
of=20
confusion, so I'd like to add a few thoughts and=20
observations.</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT><FONT color=3D#0000ff =
face=3DArial=20
size=3D2><SPAN class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>First,=20
it is important for everyone to remember that the absolute intensity =
numbers are=20
meaningless. The same absolute amount of light might produce 10000 =
counts on one=20
machine and 40000 on another. If you've been swapping parts like=20
photomultipliers around then you also can't compare numbers before/after =
parts=20
swaps.</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>While=20
I agree with Eric Anderson's comment that your intensity vs. radius =
profile=20
shows a much larger variation than most (30-40% variation seems common), =
you=20
haven't actually given us the key piece of information: do you actually =
see a=20
significant variation in noise level across the =
cell?</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>In my=20
experience a 2-fold drop in intensity at 280 nm, and even a 4-fold drop =
at 230,=20
has very little effect on the noise level. Remember too that in many =
cases much,=20
if not most, of the 'noise' in a real experiment is due to radial =
variations in=20
the windows rather than real photometric noise. For example, without =
windows on=20
my current machine the r.m.s. deviation from a flat line for a single =
replicate=20
is ~0.002 AU, but with windows it is ~0.005.</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>To me=20
the two most aberrant things going on in your machine are (1) that it =
takes=20
several attempts to get your lamp clean, and most puzzling of all =
(2) you=20
report that when the lamp is dirty your sample and reference intensities =
diverge=20
(implying a significant absorbance offset even with no=20
windows!).</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>With=20
regard to (1) I can only echo Eric's comment that if I use a pencil =
eraser until=20
I can no longer see the 'bulls eye' on the lamp face then I get back =
full=20
intensity in one cleaning. Are you perhaps using some solvent or =
cleaning tissue=20
that leaves a residue?</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>With=20
regard to sample/reference divergence, if I understand you =
correctly this=20
is stronger when the intensities are low, which might suggest a stray =
light=20
problem of some sort. Have you done the stray light tests? Unfortunately =
the=20
lamp positioning which gives the maximum intensity and/or most uniform =
radial=20
profile also often gives more stray light, so the =
best trade-off=20
is often gets you maybe 80% of the maximum possible =
intensity but=20
much, much less stray light. </SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>Also,=20
remember it is important to avoid reflections back up into the =
monochromator=20
that will fool the lamp intensity monitor---never, <U>never</U> use the =
old=20
shiny model E screw rings, and use a black marker pen to cover shiny =
spots that=20
appear on your black screw rings and window holders with use.=20
</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>The=20
only other thing I can think of that would cause the sample and =
reference=20
intensities to be different would be something causing the light to hit =
the=20
sides of the cell, partially blocking one beam or the other. One =
possibility=20
along those lines would be that your cells are not properly aligned =
exactly=20
along a radius. Be aware that on some rotors the scribe marks are =
<U>not</U> in=20
the correct positions and will cause such a misalignment. The other =
possibility=20
would be a lamp timing problem (possibly a bad magnet on your rotor?). =
One way=20
to test for either of these beam blockage possibilities would simply be =
to scan=20
an <U>empty</U> hole rather than use a windowless cell and see whether =
the=20
sample/reference difference disappears.</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>Best=20
regards,</SPAN></FONT></DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN=20
class=3D600342416-27032000></SPAN></FONT> </DIV>
<DIV><FONT color=3D#0000ff face=3DArial size=3D2><SPAN =
class=3D600342416-27032000>John=20
Philo</SPAN></FONT></DIV>
<DIV><SPAN class=3D600342416-27032000></SPAN><FONT face=3DTahoma><SPAN=20
class=3D600342416-27032000></SPAN><FONT size=3D2><FONT =
color=3D#0000ff><FONT=20
face=3DArial>A<SPAN class=3D600342416-27032000>lliance Protein=20
Laboratories</SPAN></FONT></FONT></FONT></FONT></DIV>
<DIV><FONT face=3DTahoma><SPAN class=3D600342416-27032000></SPAN><SPAN=20
class=3D600342416-27032000></SPAN><FONT size=3D2><FONT =
color=3D#0000ff><FONT=20
face=3DArial><A href=3D"http://www.ap-lab.com">w<SPAN=20
class=3D600342416-27032000>ww.ap-lab.com</A></SPAN><BR><SPAN=20
class=3D600342416-27032000> </SPAN></FONT></FONT>-----Original=20
Message-----<BR><B>From:</B> H. Olin Spivey=20
[ospivey@bmb-fs1.biochem.okstate.edu]<BR><B>Sent:</B>">mailto:ospivey@bmb-fs1.biochem.okstate.edu]<BR><B>Sent:</B> Saturday, =
March 25,=20
2000 2:25 PM<BR><B>To:</B> RASMB@BBRI.ORG<BR><B>Subject:</B> Aberrant=20
Intensities<BR><BR></FONT></DIV>
<BLOCKQUOTE></FONT>3/25/00<BR><BR>>Colleagues,<BR>><BR>>For =
about the=20
last two years, we have had a persistent problem with our XL-A =
instrument.=20
Radial intensity scans on blank cells (cells with no windows spun at =
3,000=20
rpm) decrease in intensity from the beginning to about the middle of =
the=20
radial range and then increase again to the end (cell bottom). =
Examples of our=20
recent intensity I vs. r and I vs. lambda scans can be seen=20
=
at<BR>><BR>>http://osucau.okstate.edu/plot1.gif<BR>>and<BR>>h=
ttp://osucau.okstate.edu/plot2.gif<BR>><BR>>Two=20
separate Beckman service engineers have invested several days of =
effort and=20
replaced virtually all components in our instrument to cure this =
problem, but=20
to no avail. We agreed to live with this situation since we can get =
(after=20
vigorous cleaning of the lamp) the sample and reference beams to =
superimpose,=20
and the intensity at the minimum position (about 10,000 at 280 nm) has =
been=20
acceptable for our needs. However, these signals deteriorate with use =
giving=20
both lower intensities and separation of the sample and reference beam =
intensities (variable ratios with r). This forces us to invest many =
hours of=20
lamp cleaning (one attempt is rarely sufficient) after every few =
experiments.=20
This requirement is more frequent for us since the light intensities =
near the=20
middle r values are so low to begin with. Current intensity at 228nm =
is 19188=20
and at 542nm is 1775. Intensity at 280nm is ca. 19679 at both 5.9 cm =
and 7.1=20
cm, but drops to 7932 at 6.53cm (refer to graphs above). Our questions =
are:<BR>><BR>>1) Is anybody else experiencing the same problem? =
If so=20
were you able to alleviate it, and if so how?<BR>><BR>>2) What =
could be=20
the cause of this aberrant behavior?<BR>><BR>>Since this problem =
is=20
seriously reducing our productivity, we would be grateful for any=20
suggestions.<BR><BR>With much appreciation,<BR>Frank Hays and Olin=20
Spivey<BR>H. Olin Spivey Phone: (405) 744-6192<BR>Dept. Biochem. & =
Molec.=20
Biology Fax: (405) 744-7799<BR>246 NRC Email:=20
OSpivey@Biochem.Okstate.Edu<BR>Oklahoma State =
University<BR>Stillwater, OK=20
74078-3035<BR></BLOCKQUOTE></BODY></HTML>
------=_NextPart_000_0000_01BF97CF.1EE4D180--
Index:
[thread]
[date]
[subject]
[author]