Which AO scope is easiest to convert to LED lighting?

[Scoper]

As I mentioned elsewhere, I am planning on purchasing an AO scope. With that in mind, I plan on converting the purchased scope to LED lighting.

So which AO is the easiest to convert?

Thanks for any advice.

[PeteM]

"Easiest" may be a matter of opinion. My take is that the 410 series is very easy, fairly modern, yet affordable-enough to be a very good choice. The rear illumination port is easily accessible and fairly easy to adapt to either a DIY or commercial LED retrofit.

Two versions of the 410 were made, a later one (perhaps starting with Leica branding) with a sort of door to hold the lamp. It might be a tiny bit harder to remove that door and fit a LED housing into the opening.

[PeteM]

Browser hiccup - and dupe.

[Scoper]

So is it possible to upgrade an AO scope from 10/20w halogen to 100w LED?

[PeteM]

You might find LED retrofits with near the same lumen output as a 100-watt tungsten-halogen bulb, but it will be more like 10-20 watts in actual power consumption.

The 20-watt halogen bulb these 410s come with is bright enough for most purposes. Personally, I wouldn't swap out a working one for LED at this point unless I wanted something that was cooler, longer-lasting, or made portable. If you really need 100 watts tungsten-halogen equivalent for something like taking movies or DIC at higher powers, there's always the actual 100-watt tungsten-halogen illuminator in the 420 series.

[Scoper]

Thanks for the responses.

I am attempting to assess the expandability of my next scope purchase… hopefully keeping the number of scopes purchased to a minimum (yeah…I know..go ahead and laugh…I am ;<) ).

[dtsh]

Nearly all of the AO scopes I can think of are relatively easy to adapt to LED. The 410's as mentioned, the 10 and 20 would not be difficult, I've converted several 10's with the 1036 or 1036A illluminator and have developed a drop-in that should deliver around 15-20w, the 50/60/150/160 shouldn't be too difficult; I think only the 110 would need serious modification. Any of the older 160mm TL scopes are pretty easy to adapt depending on your goals.

It really all depends on one's fabrication skills. If you can cut an aluminum plate, tap and drill some holes, and do some basic soldering, the 10 is easy enough, the 20 is probably easier especially with a lathe or a 3D printer. I've been meaning to post the drop-in for the 10, but all of it except the aluminum plate could be applied to nearly anything else with a remote transformer.

[apochronaut]

Why?

[Scoper]

Converting a traditionally lit scope to LED lighting is an upgrade path that most will take in the future…one should take it into account when considering the purchase of any scope from the used marketplace.

When one considers that many scopes on the used market are decades old, updating their features as costs and technologies adjust will be the norm…and the electronics/lighting will be the first to be modified.

[apochronaut]

Why is led an upgrade?

[Scoper]

As LEDs get better and better the time will come where one can replace any traditional lamping with a brighter and lower wattage LED..and the older incandescent and halogen lamps will be phased out.

Currently we see users adapting LEDs to scopes in a DIY effort..this will change as more efficient LEDs in more specialized shapes are manufactured.

[apochronaut]

Of course. However , the potential of engineered led microscope illuminators doesn't include diy led retrofits. The engineering of the AO illuminators is a defacto elerment within their construction that ensures a performance level that can be expected.
There is no evidence that one can improve that performance using leds.
I have asked several times for members to post images captured using led illumination and preferably comparison images between post reno. led ones and those lit by the original illuminator. Not any response.
You have to be careful not to break the original engineering.

[dtsh]

Of course. However , the potential of engineered led microscope illuminators doesn't include diy led retrofits. The engineering of the AO illuminators is a defacto elerment within their construction that ensures a performance level that can be expected.
There is no evidence that one can improve that performance using leds.
I have asked several times for members to post images captured using led illumination and preferably comparison images between post reno. led ones and those lit by the original illuminator. Not any response.
You have to be careful not to break the original engineering.

I'm willing to set that up if you can describe what you'd like me to do; I'll do my best to replicate it and share my results.

I have a stock 17w tungsten 1036A and my ~ 8w LED and a newer drop-in that should be 15-20w that I've been working on; both replace the 17w tungsten lamp and bracket in a 1036A (should drop-in on a 1036 too).

I can use the same stand, just swap illuminators. Available are the typical achromats, plan brightfield (including 1309 and 1311) , plan phase, and a darkfield condenser. I have a diatom test slide, a calibration slide, Bright-Line hemo (with 0.4 coverslip) as well as some (mostly entomological) random samples to choose from. There is a fairly limited selection of filters, but there is an AO green filter, blue filter, and a few other filters of more questionable precision. We have a pretty decent selection of various AO eyepieces, 145, 146, 176, 184, 180, 181, 191, 437 as well as some non-AO including some Olympus 10x 15x UWF 31-15-74, and an assortment of misc others.


The imaging system is a Raspberry PI HQ camera (12MP) that is coupled afocal to a Cat.176 or Cat.184, but there is access to an older Canon T2i if that would be useful.

I have the ability to fabricate in plastic and a variety of metals should that be useful in coupling things together. My electronics skill is limited, but I can solder to a prototyping board and understand just enough about LED circuits to get this far.

My gut says there's an improvement, but I recognize my ignorance of the finer details and would love to learn through direct experimentation how to quantify that in a meaningful manner. If it's easier to get reliable results, I'm happy to give one of my drop-ins to someone with the equipment and knowledge to give it an honest and thorough criticism as my goal is to find an inexpensive and easy to build drop-in illuminator that performs at least as good as the 20w halogen that can be built inexpensively by your typical hobbyist.

[apochronaut]

So, the system I use on a 10 is either the tungsten filament or halogen but pretty exclusively for phase, both dark and bright along with occasional low magnification DF, which includes a 4X .12 planachro. The illumination is set for background homogeneity by adjusting both the centering and focus of the field iris, essentially adjusting similar to Köhler. The condenser is the 1242 with a swing in aux. condenser. Intensity is controlled by a combination of adjustments of the power supply, field iris and condenser iris. I rarely use the neutral density filter but occasionally with the 4X .12 planachro.
I use a medium blue clear filter over the illuminator window at all times, which I consider an integral part of an incandescent illuminating system.
There is a BF/DF set too ( 4X .12 planachro, 10X .30 planapo, 20X .50 planachro, 40X .80 planapo and 100X 1.25 planachro w./iris) but it is normally on a model 20. Attempts to duplicate 100 watt halogen performance for 1000X DF with an led replacement so far have not been great.

I would need to see a 1 : 1 comparison for each objective and contrast type. I would have to think about the best subject for a test. Diatoms immediately come to mind but might be tricky due to refraction. Perhaps sticking to a couple of specific species might work.

[Phill Brown]

Just an image of the Kohler diaphragm would do with low power condenser.
Maybe leave the wide yellow blue fringing rather than Photoshop it out.
It's so much more than shine a bright light on it and all will be well.

[dtsh]

Just an image of the Kohler diaphragm would do with low power condenser.
Maybe leave the wide yellow blue fringing rather than Photoshop it out.
It's so much more than shine a bright light on it and all will be well.

I tried to get the leaves as focused as possible. The only change was swapping out the 17w tungsten 1036A illluminator with a modified 1036A with the 15-20w LED.
Objective is the AO 1019 10x/0.25NA through a Cat.176 10x eyepiece.

Link to the full-size images: https://postimg.cc/gallery/1QcVpN2

[Phill Brown]

Thanks for taking that on.
Test slide looks blue with LED.
Yellow fringing is quite extreme for both, wouldn't know what's going on there if the incandescent filament is aligned properly.
Is a diffuser between the light source and the iris?

[dtsh]

Thanks for taking that on.
Test slide looks blue with LED.
Yellow fringing is quite extreme for both, wouldn't know what's going on there if the incandescent filament is aligned properly.
Is a diffuser between the light source and the iris?

This LED is 5700K 90-CRI, so unsurprising it's more blue; most of what I've seen suggested that daylight was around 6000k so that's what I was shooting for. Is there a preferable color temperature that I should be trying to achieve?
The AO Series 10 use a "modified Koehler" with a frosted lens just after the bulb (same with the 410 and I believe 110)
It's possible things are not as well aligned as they could be, I'm learning as I go and haven't found procedures for testing and adjustment.
I'll fish out a couple of condensers and see if things differ; this one is the phase condenser (open, no annulus), but it does have the bottom lens in place.

If it would make things easier, I do have an AO20 stand which I believe is capable of full Koehler and an AO 370 illuminator.

[Hobbyst46]

I doubt that the blue-violet coloration is related to the condenser or presence/absence of the frosted plate.
Possibly a less than ideal camera alignment. Was the camera parfocal with the viewing eyepieces ?

[Phill Brown]

LED often have a blue spike.
Think you'll find direct sunlight is around 5000k

[Hobbyst46]

I doubt that the blue-violet coloration is related to the condenser or presence/absence of the frosted plate.
Possibly a less than ideal camera alignment. Was the camera parfocal with the viewing eyepieces ?

Taking back my thought about the camera alignment.
The blue image can be easily brought to black, by some gamma correction and tweak of saturation. And the result looks nice without any fringe (along the micrometer lines), nor CA.

5000 K is more neutral white, but IMHO sunlight under the blue sky resembles >6000K...

[dtsh]

Since this is not my usual stand, but rather one I've been experimenting on, I decided to check it over and realign things as well as I can.
I swapped out the phase condenser for a typical 1.25NA condenser (No Cat.# so I believe that would make it a Cat.1088). I have a Cat.1084 aspheric condenser too, if anyone thinks that might be of utility.

https://postimg.cc/gallery/70BjxXc

New images, as I think something was improved.

[apochronaut]

What a skewed comparison. You appear to be using the incandescent light unfiltered? Why use that type of illumination without a filter? That's why microscope manufacturers install them or design a nest for one, either in a designated filter wheel, tray or obvious sculpted depression over the illuminator window. Thats what those two little finger depressions in the housing over the series 10 illuminator window are for. So you can lift out the filter and clean it. You filter incandescent illumination with a blue, clear, filter the value of which you can choose to adjust the field background to suit your requirements.
Tested against an unfiltered incandescent light source, leds perceptually seem better because their excessive blue tint seems more natural and absorbs some of the residual ca that achromats leave. If you properly filter an incandescent source, you get about the same amount of absorption and a more natural light than led illumination.

Sunlight is a mixture of all wavelengths and is much more similar in output to incandescent light than a blue shifted led. Sunlight is no where near an absurd 6000K, it isn't even 5000K. It peaks in the green and is blended light as is incandescent light.. Many leds are excessively monochromatic, certainly the ones at high colour temperature and any I have used seem similar to fluorescent illumination in reducing contrast, although not as bad as fluorescent.

[dtsh]

What a skewed comparison. You appear to be using the incandescent light unfiltered? Why use that type of illumination without a filter?

Because I was asked to image the field iris and not modify the result?
I never claimed to be an expert, in fact I've been asking for recommendations of how to compare them in a meaningful manner. I haven't studied optics beyond what I've needed to learn in order to make projects so mostly just the very basics plus some experience with ronchi and knife edge testing from when I was grinding a parabolic mirror some years ago.

If you have advice on how to test it, I would be quite appreciative.

Here's an image taken through a stock 1036A with a clear blue filter on top of the illuminator as suggested.
https://postimg.cc/8JfkVqFd

[apochronaut]

The results one should be looking for are actual photos of the same sample done under condtions and set ups used in actual microscopy. That requires adjustment of the microscope's many settings to optimize performance. Optimize both as much as you can and then see. Visually looking at raw light output isn't very valuable, especially because leds used in microscopy are already corrected to a target colour balance.

[PeteM]

The noon sun, around 5500K, is one typical reading for daylight sun. Kodak used to call it 5600K when using blue filters to adjust tungsten illuminating to "daylight." Direct sunlight at high altitudes might be closer to 5800K.

The sun's actual color temperature depends on how much of our atmosphere it traverses and varies widely from sunrise to sunset.

https://www.naturalux.com/NaturaLux_Lig ... _Color.htm

As has been said, even high CRI LED sources rarely have the comparatively smooth spectra of either sunlight or tungsten-halogen illumination - but some are getting very close.

Then we have personal preference. For me, for both room and microscope lighting, it's a bit warmer - more like 4000K and sometimes less. A candlelit dinner might be 2000K. Others like it "cool" (which is actually a higher color temperature and hotter black body radiation equivalent) and closer to 6000K.

[Phill Brown]

The logic for viewing the Kohler iris is the effect that it's image is refracted through every point in the subject.
When there is fringing at the edges that will be the same effect at the edges of a subject, diatoms are an excellent choice for observing the transmitted fringing.
I'm not a professional either.
Thanks for sharing your findings, keep calm and carry on.

[apochronaut]

The noon sun, around 5500K, is one typical reading for daylight sun. Kodak used to call it 5600K when using blue filters to adjust tungsten illuminating to "daylight." Direct sunlight at high altitudes might be closer to 5800K.

The sun's actual color temperature depends on how much of our atmosphere it traverses and varies widely from sunrise to sunset.

https://www.naturalux.com/NaturaLux_Lig ... _Color.htm

As has been said, even high CRI LED sources rarely have the comparatively smooth spectra of either sunlight or tungsten-halogen illumination - but some are getting very close.

Then we have personal preference. For me, for both room and microscope lighting, it's a bit warmer - more like 4000K and sometimes less. A candlelit dinner might be 2000K. Others like it "cool" (which is actually a higher color temperature and hotter black body radiation equivalent) and closer to 6000K.

5700 K is the temperature of the sun not the colour. Colour temperature is an artifificial construct used as one descriptor for lamp output relative to photographic concerns originally but the sun's peak spectral output is in the green part of the spectrum. It's overall colour balance is s blend of all colours of the spectrum including the far reaches of the visible spectrum but the colour of daylight is not the colour of the sun and will vary widely over parts, times and conditions of the globe depending on many factors, just a few of which are shadow reflection, absorption and clarity of the air. Few if of us have spent even a second in pure unaltered sunlight. Perhaps those on the space station can catch a glimpse of it.

[PeteM]

Phil, it probably helps everyone in this discussion (and evaluating the pros and cons of LED lamps) to distinguish between "color" and "color temperature."

For a pure spectral color, color is defined as the frequency of the light. With something like fluorescence lamps and fluorescence microscopy and highly selective filters we can sometimes illuminate a subject with mostly a single frequency. More commonly, the colors we or our instruments "see" are a mix of wavelengths that looks like blue, red, etc. As you say, our eyes are most sensitive in the green area near the middle of the visible light spectrum. We don't see very far into either the UV or infrared regions.

Color temperature is physically defined as the entire spectral distribution of a "black body radiator" heated to various temperatures Kelvin. It's not one color, but the entire spectrum of wavelengths produced by a "body" heated to that temperature. Our Sun closes approximates a black body radiator. The average of temperatures emitted at its surface is around 5700-5900K, and we call that white light or (Kodak) "daylight." It's a mix of wavelengths below and above the visible spectrum and peaking right in the middle of it. The daylight reaching our planet is in that same range in the middle of day and is the reference (e.g. Kodak's 5600K) for the light spectrum used for color viewing and reproduction. It really is above 5000K.

While it's partly a theoretical construct (we don't have perfect black body radiators) and partly a measurable one (we can see colors with, say, various suns or even pieces of molten metal), it doesn't refer to a single color but the entire spectrum produced by a body).

I'm not sure what sort of color temperature meter is used to measure the "color temperature" of LED lights but it probably samples the visible spectrum at several points and compares the energy distribution to the Sun (or that theoretical black body radiator). Old time photographic color temperature meters used to sample just two regions. They could get away with this because the main sources were tungsten bulbs or sunlight. It's now pretty cheap and easy to measure the energy at every single wavelength in the visible spectrum, which is what I'd hope "CRI" does.

Unlike a black body radiator, the Sun, or a tungsten-halogen lamp, a LED emits at specific wavelengths and then is tweaked a bit with phosphors. Even when we combine different LED types and add phosphors (as we did for fluorescent "daylight" lamps) it's not a smooth spectrum. This is where the CRI (color rendering index) comes in. A high CRI LED will do a better job of providing a smooth spectrum, more closely matching a black body radiator.

One reason this matters is metamerism. Two light sources can look to be the same color temperature to our eyes (say, "daylight" corrected tungsten and fluorescent or LED lamps), but when a specimen (dress, color chip, etc.) is illuminated produce very different colors. This happens when the spectral energy distributions in the light source have widely different peaks and valleys. Metamerism could be just as much a problem for microscopists as for someone trying to match colors in an auto body shop. Full spectrum sources are needed.

I'd add that while someone like a pathologist or petrologist needs to have standardized color references and photos that are comparable, it's less of a concern for hobbyists. The microworld doesn't care about our standards and might deliver its wow factor to a kid, hobbyist, or Nikon-award-winning-photographer as compellingly at some non-standard illumination.

All of which is to say that a very high CRI LED can do about as good a job as a light source (daylight from a mirror, a tungsten-halogen lamp filtered to be "daylight") that naturally approaches a black body heated to a given color temperature.

A few years ago, LED technology was far behind. Today, it's cheap for low lumen output and expensive at higher levels (e.g. equivalent to 100 watt tungsten halogen). It's still behind in providing a full spectrum (black body equivalent) spectrum. However it's other advantages (less heat, longer life, lower energy cost) now make it a choice for many modern instruments.

As stated earlier, it probably doesn't make sense to replace a perfectly good tungsten-halogen illuminator of 20 watts or more with a low CRI led, which may not even have the same lumen output. For DTSH's AO10 scopes, and for someone with his DIY skills, it seems (to me) to be a great choice. These are wonderful microscopes, better built than the later MicroStars, and suffering only from somewhat dim lamps, frequently with faulty contacts, and most commonly with bulky external power supplies. After his retrofit, he has a wonderfully capable, reliable, and relatively compact microscope.

My opinion? If "Scoper" buys an AO10 or earlier, particularly with a bum lamp, he might want a retrofit - same as DTSH. Probably not for a MicroStar with a good 20-watt lamp, IMO, unless he really needs higher output. In which case, he may want a fairly capable retrofit with high lumen output, carefully matched to the scope (easy for a MicroStar) and a fairly high CRI LED.

[dtsh]

As stated earlier, it probably doesn't make sense to replace a perfectly good tungsten-halogen illuminator of 20 watts or more with a low CRI led, which may not even have the same lumen output. For DTSH's AO10 scopes, and for someone with his DIY skills, it seems (to me) to be a great choice. These are wonderful microscopes, better built than the later MicroStars, and suffering only from somewhat dim lamps, frequently with faulty contacts, and most commonly with bulky external power supplies. After his retrofit, he has a wonderfully capable, reliable, and relatively compact microscope.

My opinion? If "Scoper" buys an AO10 or earlier, particularly with a bum lamp, he might want a retrofit - same as DTSH. Probably not for a MicroStar with a good 20-watt lamp, IMO, unless he really needs higher output. In which case, he may want a fairly capable retrofit with high lumen output, carefully matched to the scope (easy for a MicroStar) and a fairly high CRI LED.

Thanks for the comment in regard to my cobbling skills.
For me a part of the love of the AO10 is that I can get plan phase that performs more than adequately for my needs in a relatively small footprint and typically at a more than reasonable price. I really like to 1031 illluminator, but didn't keep one for myself. Part of my goal is to have something that's easy and inexpensive enough for the average hobbyist to be able to assemble so that an instrument with either a missing or insufficient lamp can again be serviceable for the user's needs; it also helps that I like tinkering.

Here's the datasheet for the LED I'm using.
https://www.mouser.com/datasheet/2/723/ ... 326698.pdf

A few images of the drop-in LED. The only challenging part is cutting and drilling the aluminum plate, but even that can be done with a hand hacksaw and a hand drill with a bit of filing to make things look nice. The soldering skills necessary aren't much, I have a basic 30w non-adjustable iron. Total parts cost is well under $50.

AO10_drop-in1.jpg (99.91 KiB) Viewed 2487 times

AO10_drop-in2.jpg (58.19 KiB) Viewed 2487 times

I'm serious about getting someone to give it an honest evaluation and I will put one in a box and send it to someone who has the knowledge and capability of testing it and giving feedback so it can be improved, they can keep it as thanks for testing and reporting. I need 2 of these myself to fit on instruments and due to component cost vs shipping, I bought spares. If one intends to use it in a scope, all you need is an AO 10 and a 1036A illuminator (should work with a 1036 too) with the lamp parts removed, no permanent modification necessary.

Is there enough interest in this to warrant a proper thread for it so I don't keep hijacking Scoper's thread?