Multi LED epifluorescence illuminator

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: Chris S., Pau, rjlittlefield, ChrisR

Pau
Site Admin
Posts: 5079
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Multi LED epifluorescence illuminator

Post by Pau »

More than one whole year after the first idea I have my LED illuminator working well although not fully finished, so as promised, here you have the setup:

A) the whole setup
Image

B) the parts mounted:
1- Linear heat sink machined to fit the adapter with the LEDs
2- Custom machined aluminum adapter
3- Focusable collimating lens module from a Zeiss lamphouse HBO 50
4- Zeiss fluorocondenser IV
5- Zeiss 2FL dual dual cube with filters and beamsplitters
6- Nikon protective light shield mounted by means of a plastic card
Image

C) parts 1+2+3
Image

D) the sliding fit of the heat sink and adapter. The adapter has a spring loaded ball to fix each LED position
Image

E) heat sink with the LEDs glued with adhesive thermal paste
From left to right: UV365nm 700mA; UV 385nm 700mA; blue 455nm 1500mA; blue 460nm 1000mA; green 525nm 1000mA; white 4000K 3000mA
Image

F) The main DIY parts. The electronics box has inside
- DC 12V power source
- one of each Ledsupply drivers: 700mA, 1000mA, 1400mA and 2100mA, each one has its pot and a small LED bulb
The big knob switches the drivers, each one at once. When one driver feeds two LEDs it has a switch to commute between them
Image

E) The blue LED image at the centering window (not perfectly centered) and the card holding the UV/ blue light shied
Image

I want to thank the great help provided by several forum members about the suitable wavelengths and electronic devices in former posts (*) and very specially to Francisco Pujante aka fpelectronica who made the control box and mounted the LEDs with his high skills in electronics of which I absolutely lack and to Charles Krebs who recommended and helped me to get the LED drivers

Some further comments:
- The setup is not finished, the 385 LED came defective and will be replaced with a LG 380nm 1000mA. Likely I'll cut a final part to mount another white LED because it provides perfect Kölher illumination for the microscope transmitted light. I also plan to design some hood parts to prevent the light laterally escaping from the device (UV and royal blue are dangerous...)
When finished I'll label the control box.

* http://www.photomacrography.net/forum/v ... hp?t=28586
* http://www.photomacrography.net/forum/v ... hp?t=28587
* http://www.photomacrography.net/forum/v ... hp?t=31306
Last edited by Pau on Tue Dec 13, 2016 4:23 pm, edited 1 time in total.
Pau

ChrisR
Site Admin
Posts: 8521
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Very interesting, well done! Looking forward to results...
Chris R

zzffnn
Posts: 1826
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Very nice work, Pau, and it shows in your autofluorescence images.

So for chlorophyll:

1) I am guessing you are using the royal blue 455 nm LED? Which exact model/part number is that LED?

2) Did you manage to avoid excitation filters and dichroic beamsplitters? You mention it before in your post.

3) which exact "LP emission filter" did you use for your Polypodium cambricum image? How long was the exposure time for the 1st image made with royal blue LED?

I may do autofluorescence with my immersion darkfield condenser rig, for chlorophyll only. I have a 50w, ~5000 lumen LED fitted for transmitted darkfield, which has a fair amount of emission between 430 nm to 480 nm (see page 7 of the pdf) and some longer wavelength:

http://www.cree.com/~/media/Files/Cree/ ... XA2540.pdf

http://www.digikey.com/product-detail/e ... ND/3947354

I think I may need to use excitation filter with that white LED, if I don't get a dedicated royal blue LED.

Thank you very much.

Edit:
I would also like to know the spectrum spec of your green 525nm LED. If its wavelength is very narrow, then it may be useful for imaging colorless subject, at high resolution, with green-mono light conversion.
Last edited by zzffnn on Wed Dec 14, 2016 12:55 pm, edited 1 time in total.

Pau
Site Admin
Posts: 5079
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

Thanks for your positive comments
zzffnn wrote:...
So for chlorophyll:
1) I am guessing you are using the royal blue 455 nm LED? Which exact model/part number is that LED?
2) Did you manage to avoid excitation filters and dichroic beamsplitters? You mention it before in your post.
3) which exact "LP emission filter" did you use for your Polypodium cambricum image? How long was the exposure time for the 1st image made with royal blue LED?

I may do autofluorescence with my immersion darkfield condenser rig, for chlorophyll only. I have a 50w, ~5000 lumen LED fitted for transmitted darkfield, which has a fair amount of emission between 430 nm to 480 nm (see page 7 of the pdf) and some longer wavelength:
.....
I think I may need to use excitation filter with that white LED, if I don't get a dedicated royal blue LED.
....
Edit:
I would also like to know the spectrum spec of your green 525nm LED. If its wavelength is very narrow, then it may be useful for imaging colorless subject, at high resolution, with green-mono light conversion.
Hi Fan, lots of questions...will try to answer

1) XLamp Cree XT-E (5W) and Cree XP-E (3W) Both work well, I prefer the XT-E

2) In the meantime I've collected enough filter sets to do it properly with complete filter cubes and some extra filters, so I've not tested without them.
The emission filter is always required, the excitation filter I'm sure could be avoided if the excitation and emission wavelengths are well separated like in your case using royal blue to capture red chlorophyll emission, when they are closer is very convenient as LED emission is not really monochromatic (In some cases like the XT-E it is pretty much but in others like my UV it even emits white light)
The dichroic beamsplitter is very important, good ones provide up to 95% reflection of excitation light and 95% transmission of emission and also improves the efficiency of the other filters. It was a big advancement in fluorescence microscopy.
If you use a BS for white light it will split 50% in each direction, so you will loss 75% of light minimum

3) LP = long pass. It's strong yellow, I have not the data.
You have first rate info about filters at Chroma site
https://www.chroma.com/knowledge-resour ... microscopy
https://www.chroma.com/knowledge-resour ... cteristics
and of course excellent more general info at microscopyu.com

- At my Polypodium picture exposure was 0.8s at 200 ISO

- I have not at hand the spectrum curve of the green LED, but being a Cree XP-E you likely could find it at Cree site, I'm pretty confident that it could do well for general green illumination, although a green IF filter on a powerful white LED will also do it in a more versatile setup.
Last edited by Pau on Thu Dec 15, 2016 10:56 am, edited 1 time in total.
Pau

harisA
Posts: 498
Joined: Sun Jul 03, 2011 9:31 am
Location: Greece

Post by harisA »

Very interesting thread with a lot of information.Thank you for sharing.
As far as concern cree 365 and cree 385 can you provide a link to a datasheet.Searching cree site and ebay was not possible to find any info.

Pau
Site Admin
Posts: 5079
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

Haris, my UV LEDs are not from Cree, but from unkown chinese manufacturer( although mounted on Cree star plates) , bought from:
http://www.ebay.com/itm/3535-365nm-395n ... jbmkOU-LJg

Later I bought a LG (yet to be tested) to substitute the non working 385nm (I got immediate refound)
http://www.ebay.com/itm/231926718291?_t ... EBIDX%3AIT
Pau

Jacek
Posts: 5353
Joined: Sun Oct 02, 2011 7:00 am
Location: Poland

Post by Jacek »

Excellent, I collect jaw off the floor.
Congratulations :D

zzffnn
Posts: 1826
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Pau, thank you so much for your very informative and helpful reply!

Your short exposure times shows that your rig is working really well.

harisA
Posts: 498
Joined: Sun Jul 03, 2011 9:31 am
Location: Greece

Post by harisA »

Pau thank you very much.I've just check a cheap uv torch light i bought recently and it has the same 3W 365nm "cree" chinese led.I'm surprised from your results because these are rather weak leds.I have read about some japanese nichia uv leds with spectacular characteristics.Judging from your images i suspect nichia leds will be equal or superior to a common hbo lamp

johan
Posts: 1005
Joined: Tue Sep 06, 2011 7:39 am
Contact:

Post by johan »

Just echoing what's been said much better by other people already, thank you very much for posting this Pau! I've just bought my first serious scope and this unit is very much where I want to head with it.

harisA, yes, Nichia 365nM LEDs is where it's at. I'm surprised the cheap torch Cree works as well as reported, I had a dismal time testing cheap torch LEDs. Fwiw, MTE made a custom 'one of a kind' UV torch for me with a new generation Nichia NVSU333A chip (3540MW LED!!!). I'm able to get autofluorescence working by shining that torch into a Labophot epi tube. My goal is to take that torch apart and use it inside a unit like yours Pau!!!

Fwiw, Nichia vs. cheap torch tests at http://extreme-macro.co.uk/uv-macro-lighting/
My extreme-macro.co.uk site, a learning site. Your comments and input there would be gratefully appreciated.

Charles Krebs
Posts: 5857
Joined: Tue Aug 01, 2006 8:02 pm
Location: Issaquah, WA USA
Contact:

Post by Charles Krebs »

Pau,

Thanks for posting this. Very useful and helpful information.

As far as UV LEDs are concerned this is a very nice one:
http://www.digikey.com/en/product-highl ... n2-emitter

DABOSSUK
Posts: 10
Joined: Mon Dec 05, 2016 2:35 am

Post by DABOSSUK »

This looks very impressive and something i'm looking into in the future ( after pondering a PhotoMicroscope III ) , do you have any links to images taken with this set-up please? :D

ChrisR
Site Admin
Posts: 8521
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Dabossuk: if you look at Pau's recents post you'll see a few.
Be warned, he's a knowledgeable fellow.
For me as a novice microscopist and dunce biologist, the appropriate time for this would be rather a long way into the future!
Chris R

Pau
Site Admin
Posts: 5079
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

Thank you all very much for your kind comments and suggestions :D

Likely the Nichia NVSU333A will be much better (too powerful I guess!, dangerous I think), there is another model less powered that seems to fit well, but I don't find how to buy it online.

The LedEngin LZ1 UV 365 nm Gen2 will be likely better than the no name one I have, althoug the power seems about the same. I bought this one because I searched for single chip emitters, although many small emitters so close will also do well. Anyway with a different Chroma filter cube I've recently mounted emission is clearly better, twice at least.
zzffnn wrote:...So for chlorophyll:

2) Did you manage to avoid excitation filters and dichroic beamsplitters? You mention it before in your post.
...
I think I may need to use excitation filter with that white LED, if I don't get a dedicated royal blue LED.
I've done the tests for you on a fern leaf with the royal blue Cree XT-E:
- With a board band beamsplitter (all wavelengths) it does not work, too much excitation light despite using adequate emission filters
- With the cube removing the excitation filter it works but with less contrast (some glare) and some green light illuminating the subject. In some cases the effect is very nice with the chloroplasts fluorescing red and the cell wall illuminated green, in fact I plan to further test it to eventually use it.

Some observations:
- Chlorophyll photo bleaches very fast although it stabilizes with some emission much weaker than when illumination starts, this difficults to do stacks, any trick to control it? The effect is smaller with green light but also fluorescence is much weaker.
- "monochrome" LEDs are far of monochromatic enough for epifluorescence, despite being very useful light sources
- Fluorescence filters (and some of the filters I have are very good interference ones) seem to let passing an important fraction of the light they are meant to block
Last edited by Pau on Sun Dec 18, 2016 8:41 am, edited 1 time in total.
Pau

zzffnn
Posts: 1826
Joined: Thu May 22, 2014 1:25 pm
Location: Boston, Massachusetts, USA
Contact:

Post by zzffnn »

Thank you very much, Pau.

When you said "With a board band beamsplitter (all wavelengths) it does not work, too much excitation light despite using adequate emission filters".

You meant the combination that did not work was:
excitation filter + broad band beam splitter + emission filter
correct?

When you said "With the cube removing the excitation filter it works but with less contrast (some glare) and some green light illuminating the subject."

You meant the combination that worked by not optimally was:
No excitation filter + dichroic beam splitter + emission filter
Correct?

I just want to confirm, thank you.

Yes, I know for sure that a Nikon green interference filter pass some yellow and red light. Though theoretically it should not.

Do you know the mechanism of chlorophyll photo bleaching? If it is by oxidation, maybe you can pre-soak the plant tissue with some antioxidants (such as vitamin C + E combination). I am just guessing.

Post Reply