Re: 1W 405 source

Posted by Benjamin Smith on
URL: http://confocal-microscopy-list.275.s1.nabble.com/1W-405-source-tp7591310p7591329.html

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Yeah, the LightCrafter and DLi's CEL5500 both use a rectangular microlens
array as their homogenizer.  If you remove it, you do get a bit more power
per unit area, but then you cannot fully illuminate the DMD due to the
other apertures and lenses in the path..

I was considering using one of these as a cheap fiber shaker:
https://www.adafruit.com/product/1201, but instead went with the brute
force approach with the 10W LED.  With a 10x/0.3 NA objective, the 3W laser
put about 25 mW / mm^2 onto the sample plane with about half of the total
field illuminated.  The 10W LED put about 45 mW/mm^2 onto the sample plane
with the whole field being illuminated.  For comparison, a xenon arc lamp
passing through a dichroic filter will put about 8 mW / mm^2 onto the
sample plane.

One annoying feature of the microlens array for microscopy applications is
that since the NA of the focal plane is greatly decreased, an out-of-focus
shadow of the lens array is cast onto the sample plane (because the depth
of focus of the microlens array becomes so large you can see the focal
plane on the image plane).  The only way we found to fix this was to
illuminate the entire microlens array, hence the shift back to the LED.

To be honest, if it was cheaper to get a bare DMD and driver board I would
prefer to go that route and use my own illumination setup, but as of right
now they are still much more expensive than the Light Crafter demo module.

One final side note, EKB technologies sells pre-modified Light Crafters for
pretty much the cost of the Light Crafter itself, including a fiber coupled
and LLG coupled model:
https://www.ekbtechnologies.com/e-store/dlp-lightcrafter-e4500-mkii-fiber-couple

These make both LED and laser coupling much easier.

Cheers,
   Ben Smith

On Wed, Sep 23, 2020 at 6:44 PM Zdenek Svindrych <[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> Post images on http://www.imgur.com and include the link in your posting.
> *****
>
> Hi Ben,
>
> I'm glad LEDs work for you. Absolutely, if 10 mW/mm^2 is enough, LEDs are
> the way to go!
> We've used Luminus Phlatlight LEDs in the past, they are much brighter (in
> terms of W/mm^2/sr) than most. But they are also more pricey.
> I'm surprised how much better the LEDs are now, especially in the 400 nm
> range with 1 W optical output from 1 mm^2 chip...
>
> Importantly, there is no reason why a LED-based fluorescence illumination
> system *with* DMD should be dramatically dimmer than *without* it. The
> optical efficiency of a DMD is 70%. Still, even in commercial LED light
> sources most of the light is wasted, as the chips are too big (but it makes
> the optical design easier).
>
> Lasers make things more complicated, but my favorite arrangement -
> multimode fiber shaker and critical illumination - work most of the time.
> The DMD acts as a very poor blazed grating, so it's tricky to make it work
> efficiently, especially with multiple laser wavelengths. But with a 1 W 405
> nm laser you still have plenty of light for photoconversion, photobleaching
> and photodamage experiments.
>
> Finally, I have never heard of microlens homogenizer used together with a
> DMD. All DLP projectors (I've had the pleasure to take apart) use a light
> pipe homogenizer, maybe except for the tiniest ones. I wasn't lucky enough
> to get my hands on a LightCrafter or similar stuff... For microscopy
> equipment, I don't see a reason for a homogenizer (exactly for that same
> reason - the etendue of the source is so much bigger than what the
> objective lens can use).
>
> Best, zdenek
>
>
> On Tue, Sep 22, 2020 at 4:02 PM Benjamin Smith <
> [hidden email]>
> wrote:
>
> > *****
> > To join, leave or search the confocal microscopy listserv, go to:
> > http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> > Post images on http://www.imgur.com and include the link in your
> posting.
> > *****
> >
> > Attaching DMDs to microscopes is actually pretty common in optogenetics
> and
> > retinal research.  Here is an image of a DMD coupled into a BX-51 arm in
> > one of the labs I work with using one of the high power LEDs:
> > https://drive.google.com/file/d/1oFL0CQynji-hmkIDca002IZMmitwtksd/
> >
> > In this setup we wanted a high power 450 nm output for optogenetics (>10
> > mW/mm^2 at the sample plane) while still having it usable as a
> fluorescent
> > microscope, so we used a 10W 450 nm LED (the one with the CPU cooler
> > attached) for the high power channel combined with a liquid light guide
> > from a xenon arc lamp for longer wavelengths.
> >
> > Originally we tried a 3W 450 nm laser we bought off of eBay for $50
> (which
> > to our surprise really was 3W optical output), but quickly realized that
> > the vast majority of losses in a DMD are at the microlens array used to
> > make the light uniform across the DMD.  Additionally the coherence of the
> > laser caused a strong speckling issue, which we tried to abate with 10m
> of
> > multimode fiber but was still pretty strong. We also had to couple at an
> > odd four degree of freedom offset to try to get as much of the field
> > covered as we could.
> >
> > We then switched to the 10W LED, and since it covers the whole microlens
> > array, the field of illumination was perfectly uniform, and did not
> require
> > any odd 4 degree-of-freedom alignments.
> >
> > One thing to keep in mind is that the DMD array itself is fairly big, so
> as
> > along as the LED die is appreciably smaller than the DMD itself, then the
> > etendue really won't cause a loss in intensity due to coupling, and will
> > only work in your favor to have a more uniform field of illumination.
> > Additionally, since these LEDs are intended to be used with DMDs, some of
> > them even have dies with a 4:3 or 16:9 aspect ratio for even better
> > efficiency.
> >
> > And a quick side note, for anyone who want's to couple a DMD into a
> > microscope, I strongly recommend using the BX-51 arm.  It is easy to pull
> > out the lenses, and a Thorlabs SM1 lens tube fits neatly inside with only
> > minor modification.  You can also fit a TTL200 tube lens right behind the
> > manual shutter, giving you the perfect spacing to the back focal plane of
> > the objective.
> >
> > Cheers,
> >    Ben Smith
> >
> > On Tue, Sep 22, 2020 at 8:57 AM Zdenek Svindrych <[hidden email]>
> > wrote:
> >
> > > *****
> > > To join, leave or search the confocal microscopy listserv, go to:
> > > http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> > > Post images on http://www.imgur.com and include the link in your
> > posting.
> > > *****
> > >
> > > Hi Antonio,
> > >
> > > We used to buy from Dragon lasers (China) many years back, now CNI
> Laser
> > > (again China) seems like a great source (no commercial interest).
> > >
> > > If you want to go super cheap, you can search eBay for laser engraving
> > > modules, some are 405 nm (but mostly 445 nm), you want something like
> > "7W"
> > > at least, as it's the electric input power... And don't stare into the
> > beam
> > > with your remaining eye!
> > >
> > > Andor uses a fiber shaker to homogenize the beam (and critical
> > > illumination, like their Borealis technology) for their Mosaic DMD
> > devices,
> > > but I can't guarantee the amount of light you can couple into a 50 um
> or
> > > 100 um fiber from the eBay laser modules...
> > >
> > > UV LEDs sound cool, but the problem is the coupling efficiency
> (etendue).
> > > If your source is a 1 mm^2 LED chip, you can illuminate 1 mm^2 with a 1
> > NA
> > > objective lens with good efficiency (if you try hard). If you try to
> > focus
> > > the light down to 0.3 mm^2, 90% of the light is lost... And higher
> power
> > > (more LED chips) won't help you get more light trough, only the optics
> > may
> > > be simpler... Of course on top of that another 99% of light is blocked
> if
> > > you only have few DMD pixels "on"...
> > >
> > > I've seen papers where folks bolted a DMD digital projector to a
> > > microscope. Lie this one:
> > > https://drive.google.com/file/d/1IEfgKvZ0JrCUaaUPdnRFhUzT3phJO-D-/
> > >
> > > Best, zdenek
> > >
> > > On Mon, Sep 21, 2020 at 10:11 AM Antonio Jose Pereira <
> > > [hidden email]> wrote:
> > >
> > > > *****
> > > > To join, leave or search the confocal microscopy listserv, go to:
> > > > http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> > > > Post images on http://www.imgur.com and include the link in your
> > > posting.
> > > > *****
> > > >
> > > > Dear all,
> > > >
> > > > I'm looking for a 0.5-1W , 405nm, can have many spatial as well as
> > > > longitudinal modes, 405nm laser. This is to feed a photoconversion
> > module
> > > > (Mosaic), a DMD-based system that wastes most power. That's why we
> need
> > > > such high power source.
> > > >
> > > > I appreciate if you could give suggestions on low-budget options,
> which
> > > > I'm failing to find ...
> > > >
> > > > Thank you so much,
> > > > Antonio
> > > >
> > > >
> > > >
> > > > Antonio Pereira
> > > > CID lab
> > > > i3S/IBMC, Universidade do Porto
> > > > Room 001.S2B, +351 22 607 49 59 Ext. 6127
> > > >
> > >
> > >
> > > --
> > > --
> > > Zdenek Svindrych, Ph.D.
> > > Research Scientist - Microscopy Imaging Specialist
> > > Department of Biochemistry and Cell Biology
> > > Geisel School of Medicine at Dartmouth
> > >
> >
> >
> > --
> > Benjamin E. Smith, Ph. D.
> > Imaging Specialist, Vision Science
> > University of California, Berkeley
> > 195 Life Sciences Addition
> > Berkeley, CA  94720-3200
> > Tel  (510) 642-9712
> > Fax (510) 643-6791
> > e-mail: [hidden email]
> >
> >
> https://vision.berkeley.edu/faculty/core-grants-nei/core-grant-microscopic-imaging/
> >
>
>
> --
> --
> Zdenek Svindrych, Ph.D.
> Research Scientist - Microscopy Imaging Specialist
> Department of Biochemistry and Cell Biology
> Geisel School of Medicine at Dartmouth
>


--
Benjamin E. Smith, Ph. D.
Imaging Specialist, Vision Science
University of California, Berkeley
195 Life Sciences Addition
Berkeley, CA  94720-3200
Tel  (510) 642-9712
Fax (510) 643-6791
e-mail: [hidden email]
https://vision.berkeley.edu/faculty/core-grants-nei/core-grant-microscopic-imaging/