High CRI LEDs - Olympus True Color, Nichia Optisolis and Yuji VTC D50

[viktor j nilsson]

Since we were on the topic of LEDs in another thread, I thought I could post some of the things I have found recently.

As many others, I have been using a Cree XML2 10W single-die led in my Wild M20 microscope, using a BuckPuck 1000mA constant current driver. It's been fine, but I have kept looking for options.

Two factors in particular could be improved: light output and color renditioning. This post is mainly about the latter.

As most high-power leds, the XML2 suffers from a striking dip in the spectrum around 480nm:

CreeXML2.JPG (43.21 KiB) Viewed 7833 times

Olympus has to great fanfare introduced a product names Olympus True Color LED, for which I assume they charge a handsome sum.
https://www.olympus-lifescience.com/es/ ... color-led/

The spectrum does look very nice:

Truecolor.jpg (21.33 KiB) Viewed 7833 times

Just now I found a picture of the Olympus True Color LED, and it looks like this:

TruecolorDIE.jpg (14.89 KiB) Viewed 7833 times

source:
https://www.olympus-lifescience.com/pt/ ... rue-color/

Clearly it is a fairly large COB led.


Since I was sure Olympus couldn't be making these LEDs themselves, I started looking around for high-CRI (Color Rendition Index) leds. Here's the spectra of some of the best 98+ CRI LEDs I have found:

Nichia Optisolis 5000K

Nichia optisolis.png (37.49 KiB) Viewed 7833 times

Yuji VTC D50

Yuji.jpg (39.55 KiB) Viewed 7833 times

Here is a nice write-up about them:
http://budgetlightforum.com/node/59478

Based on the spectrum, it looks a whole lot like Olympus might be using technology from Yuji (even though Optisolis might actually be slightly better in terms of output and CRI).

I am not sure exactly which COB LED they are using, but this one does look extremely close:
https://store.yujiintl.com/products/vtc ... 5426214099


Anyway, I thought I'd collect this information here in case it could help someone else that is looking for a high-CRI led for their microscope.

[MichaelG.]

Thanks for the very concise and useful summary, Viktor

We have discussed this before ... but it’s much more convenient to have it all available in one post.

MichaelG.

[Hobbyst46]

Thanks a lot, Viktor, for this post !
The color rendition of LEDs has been mentioned (as well as Yuji) in a previous post or two, 1-2 years ago on the forum.
At that time, we could only find 5mm LEDs from them. The present info about prices and availability helps.
I like very much the spectrum of the 5600K LEDs.
9W nominal is probably adequate for many microscopes.

[Hobbyst46]

Would be nice to know suppliers of these LEDs, who cater small quantities.
Plus, to know if they are single or multiple die units.

[75RR]

As many others, I have been using a Cree XML2 10W single-die led in my Wild M20 microscope, using a BuckPuck 1000mA constant current driver. It's been fine, but I have kept looking for options.

Two factors in particular could be improved: light output and color renditioning. This post is mainly about the latter

Certainly worth looking at improved colour rendition. Thanks for that.

To go briefly to your first point i.e. light output The Cree XML2 10W has a Max drive current of 3000mA, so your BuckPuck 1000mA constant current driver will be limiting its output somewhat.

There is a BuckPuck 2100mA constant current driver that will help with that.

[viktor j nilsson]

To go briefly to your first point i.e. light output The Cree XML2 10W has a Max drive current of 3000mA, so your BuckPuck 1000mA constant current driver will be limiting its output somewhat.

There is a BuckPuck 2100mA constant current driver that will help with that.

Ah, yes, I'm aware. I didn't know that I was going to start dabbling with DIC and other light-hungry techniques when I picked the 1000mA BuckPuck over the 2100mA BuckBlock.

I also have a ZK-4KX constant current driver that goes up to 4A which I can use when I need more output.

For visual use, the problem with the XM-L2 is actually that it is often too bright on it's lower setting.

[Lutra87]

Hey guys, this retailer sells small quantities of the Nichia Optisolis LEDs:

https://www.virence.com/

I'm considering getting 4 soldered by them onto their VR21P4 MCPCB:



Do you think the location of the LEDs (In those yellow boxes) will affect optics? Sorry if thats a dumb question, just getting started here.

The goal is replacement of light source for old Zeiss Standard.

Thanks!!!!

[MichaelG.]

Do you think the location of the LEDs (In those yellow boxes) will affect optics? Sorry if thats a dumb question, just getting started here.

Sorry, I can’t see the image

MichaelG.

[Lutra87]

Does this URL work?

https://static.wixstatic.com/media/5b25 ... 0_s_2.webp

[Saul]

Hey guys, this retailer sells small quantities of the Nichia Optisolis LEDs...The goal is replacement of light source for old Zeiss Standard...

Data page states:
3v x 0.065a = 0.19w
4 x 0.19w= 0.76w only total per one quadruple pcs
Did I misunderstand something ?

[MichaelG.]

Does this URL work?

https://static.wixstatic.com/media/5b25 ... 0_s_2.webp

It offered me a ‘download’ ... which [presuming you to be a decent chap] I did ... but the iPad could not open that.

A little experiment demonstrated that this shortened url does work:
https://static.wixstatic.com/media/5b25 ... 00_s_2.jpg

Now ... to answer your original question: It probably depends on the optics of your illuminator.
If it’s designed for a compact filament, I think you will be disappointed.

MichaelG.

[Lutra87]

Now ... to answer your original question: It probably depends on the optics of your illuminator.
If it’s designed for a compact filament, I think you will be disappointed.

So right now I don't have an illuminator. My stepfather closed down his lab 20 years ago and has two Zeiss IM35 Inverted Scopes, A Zeiss Standard and a Leitz Dialux EB w/ a fluorescent light source, but none of the scopes have regular white light illuminators. I have an interest in circuitry and optics so i figured I might as well learn how to make the illuminator. Is this foolhardy? Given how much time we all probably have right now due to current circumstance, I figured there is no better way to spend my time. My engineer friend has access to a 3d printer, so I was planning on using CAD to design the lightbox, and the circuits are pretty simple. My knowledge weak spots are with regards to the actual optics.

[MichaelG.]

Now ... to answer your original question: It probably depends on the optics of your illuminator.
If it’s designed for a compact filament, I think you will be disappointed.

So right now I don't have an illuminator. My stepfather closed down his lab 20 years ago and has two Zeiss IM35 Inverted Scopes, A Zeiss Standard and a Leitz Dialux EB w/ a fluorescent light source, but none of the scopes have regular white light illuminators. I have an interest in circuitry and optics so i figured I might as well learn how to make the illuminator. […]
My knowledge weak spots are with regards to the actual optics.

.
Ah !! ... That puts a slightly different complexion on it
... I thought you wanted to replace the lamp in an existing illuminator.

If you are starting from scratch, then you need to make an important design decision before you start:
Is the illumination to be diffuse or a beam ?

Watch this space ... I will post a link to an excellent old book that might help you get started.
[ and I’m sure other members will jump-in to advise ]

MichaelG.

.

Found it quicker than I expected:
https://archive.org/details/EngineeringOptics/mode/2up
This will display it as a flip-book, but if you scroll down the page, there are various download options.
Considering the enormous amount of material that it contains; it’s surprisingly easy reading.

[Lutra87]

Thank you sir! Looks like I have some reading to do

[microb]

If you are starting from scratch, then you need to make an important design decision before you start:
Is the illumination to be diffuse or a beam ?

Could you give a page number in the book/pdf you gave on what you mean by the important design decision? I thought the design would be collimated light with an removable diffuser glass inserted for diffuse. But you're mentioning a design issue. So what two optics paths are you refereeing about?

Thanks,
Ted

[Lutra87]

So the Zeiss Standard I have seems to have the collector lens and frosted glass built into the female part of the light tube. So would I just have to adjust the distance between the LED and the frosted glass/ collector lens until I can get an image of the LED centered on the condenser front focal plane? I was worried that I may have to buy and adjust the optics of the base, but it seems like all the pieces except the actual light source are still there.

[MichaelG.]

If you are starting from scratch, then you need to make an important design decision before you start:
Is the illumination to be diffuse or a beam ?

Could you give a page number in the book/pdf you gave on what you mean by the important design decision? I thought the design would be collimated light with an removable diffuser glass inserted for diffuse. But you're mentioning a design issue. So what two optics paths are you refereeing about?

Thanks,
Ted

Not specifically, Ted ... The reason I mentioned making an important design decision is that some illuminators start with a large area diffuse source, and others with a near approximation to a point source [and then, of course there is also Koehler’s clever trick].
Lutra87 had previously mentioned getting an array of several small area LEDs to use as a source ... Which is ‘neither fish nor fowl’ and would [in my opinion] be almost certain to disappoint.

The design decision is ‘What am I trying to make?’
... Once you have decided that; this book [and many others] will help you understand the relevant optics.

MichaelG.

[MichaelG.]

So the Zeiss Standard I have seems to have the collector lens and frosted glass built into the female part of the light tube. So would I just have to adjust the distance between the LED and the frosted glass/ collector lens until I can get an image of the LED centered on the condenser front focal plane? I was worried that I may have to buy and adjust the optics of the base, but it seems like all the pieces except the actual light source are still there.

That’s great news

... but, as just mentioned; I would be wary of using that array of small LEDs
... Do you know what size/shape of filament was originally used in the ‘scope you have ?

MichaelG

[microb]

of course there is also Koehler’s clever trick

Is Koehler's clever trick the lens assembly shown here between the light source and the field diaphragm?

[MichaelG.]

https://www.microscopyu.com/tutorials/kohler

Koehler makes it possible to produce even lighting of the specimen from a source which is not itself either large, diffuse, homogeneous, or a point. ... i.e. from, for example, a compact filament lamp.

I count that as a clever trick

MichaelG.

.

Edit: See also, the two excellent links posted by 77RR at #5 here:
viewtopic.php?p=78768#p78768

[mneium]

Regarding high CRI COBs, here are some good ones:

https://www.digikey.com/products/en?keywords=SAWS1063a (10w)

https://www.digikey.com/products/en?keywords=SAWS1566a (25w)