*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****
Hi all,
Lots of good answers already but I think that it
is important to remember that Hg has a lot of 10x
peaks in the visible part of its spectrum. So an
answer with one bandpass filter may not indicate
a general trend.
One also needs to know a bit about the optics
used. Critical illumination (source focused onto
the image plane) will usually be brighter than
Kohler, but whereas critical illumination that
focuses the brightest "plasma ball" of the arc
into the centre of the imaged area can be
relatively uniform over a small field-of-view, it
is not so clear what will happen when one images
an LED into the image plane (It will depend on
the construction of the LED).
(Note: The actual ball of a 100 w Hg source is
usually about 150µm in diam. just near the
electrode. Assuming both the collector lens and
the objective are of about the same, high NA, the
most efficient optics to convey this to the focus
plane will do so at a magnification of 1:1. Of
course the field of view of the objective (in µm)
will vary inversely with its magnification, but
the area that can be properly sampled with a
given CCD/sCMOS will not vary: a 1000x1000 array
of 0.1µm pixels will be about 140µm across its
diagonal). However each manufacturer has made
different compromises in terms of the
magnification of their epi-illumination system
(at very least to also accommodate
low-magnification use) and some may utilize more
of the light leaving the original light source
(arc or fibre) than others.
The idea of measuring the output at the
microscope end of a fibre-optic seems sensible as
long as this is how you will illuminate your
sample on your exact setup. However, such fibers
have an NA (angle at which the light leaves them)
and so not all the light leaving them necessarily
makes it to the image plane. For instance, a
given system may under- or overfill the entrance
pupil of a given objective. As one can never make
the light brighter (in photons/second/µm*2)
simply by focusing it, if the end of the fibre is
much larger than 150 µm (in the example above)
then some of the light leaving it must
inevitably be lost somewhere before it reaches
that part of the focus plane covered by the
1000x1000 image sensor.
My preference would be to measure the light
leaving the objective once the field diaphragm
has been set to repeatable diameter (say 100µm
diam. at the image plane). Of course, you can
only set the diaphragm properly if the scope is
set to Kohler. Once it is set (to standardize the
light path to that point) you can still tweak and
condenser focus to approximate critical
illumination. (i.e., make the image as bright as
possible). The adjustment is inevitably a bit of
a "fudge" because, as the arc is a 3D source
rather than the planar object imagined by the
Kohler Illumination diagram, it "cannot be
focused into a plane".
Of course, there is still a problem: You probably
want to use a hi-NA objective but above NA 0.5
more and more of the high-NA light will reflect
back into the objective from its front surface.
Rays at >NA 1.0 will not escape into the air at
all. Efforts to couple the sensor of your
photometer to the objective will a drop of
immersion oil will only work if there is no
air-gap between the sensor window and the
sensitive element. The options are:
1) To couple a small, strong plano-convex lens
onto the back of the microscope slide with
immersion oil to make the light beam less
divergent or
2) To set up using an NA 0.75 air objective and
hope that the difference in NA isn't too
important or
3) Measure the fluorescent light signal at the
CCD from a thin, uniform layer of fluorescent dye
(It should be thin so that you don't end up
focusing too far inside the layer, where SA and
absorption may be variables you don't want).
So now you see why just measuring the output of the fibre seems easier.
Hope that this isn't too confusing. I am really
theoretically very pro-LED (faster, cooler, just
the light you want etc). Indeed, I think that
Chapter 3 in the Handbook was one of the first
places where LED microscope sources were
discussed in any depth. I would just like to see
a few more variables nailed down.
Cheers,
Jim Pawley
>
>
>Hi Phil,
>
>Current LED light sources can be brighter and
>(should have) more stable light output (and
>"instant" on/off, and less heat output and less
>ozone and no chance of the bulb exploding ...
>"do not look at top of arc lamp with remaining
>eye". Also many LEDs have precise - and
>reproducible - voltage control. Purchase price
>will eventually be made up in total cost of
>ownership.
>
>Brighter light sources enable selection of
>narrower wavelength range, for example, at the
>excitation peak of the desired fluorophore (and
>hopefully minima of unwanted fluorophores),
>leaving more room for emission wavelength range.
>
>George
>
>On 11/5/2013 11:56 AM, Philip Oshel wrote:
>>*****
>>To join, leave or search the confocal microscopy listserv, go to:
>>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>*****
>>
>>All,
>>
>>I had this question put to me by a new faculty
>>member, and don't have a ready answer:
>>"Is there a ballpark percentage for how much
>>less bright an LED vs a standard mercury lamp
>>light?"
>>This is for regular epifluorescence, not confocal.
>>
>>This is in the realm of arm-waving over a
>>picture of beer (a good, dark stout), ignoring
>>brands, how old the Hg bulb is, ex/em cubes,
>>which part of the spectrum is used, and all
>>that. Personally, I'd think the answer is more
>>like, "Doesn't matter, the dimmer system is
>>still too bright to use all the available light
>>and not damage the specimen." But ... ?
>>
>>Phil
>
>
>--
>
>
>
>George McNamara, Ph.D.
>Single Cells Analyst
>L.J.N. Cooper Lab
>University of Texas M.D. Anderson Cancer Center
>Houston, TX 77054
>Tattletales http://works.bepress.com/gmcnamara/26/
--
James and Christine Pawley, 5446 Burley Place (PO
Box 2348), Sechelt, BC, Canada, V0N3A0,
Phone 604-885-0840, email <[log in to unmask]>
NEW! NEW! AND DIFFERENT Cell (when I remember to turn it on!) 1-604-989-6146
|
