Which AO scope is easiest to convert to LED lighting?

[PeteM]

DTSH - the spectral curves on higher CRI versions of the LED lamp you chose look well-behaved to me.

The 4000K version of the 90 CRI has a bit more energy in the blue-green area - a plus in my mind. It will be brighter as well.

The big spectral difference between tungsten halogen and those high CRI LEDS is the big bump in the near UV (around 450nm) region. Most people probably will barely see it. For those that can, it will add a bluish tinge. It could easily be filtered out. For example, with a $2 bit of welder's helmet polycarbonate plastic in the illumination path would knock down UV at and below 450nm. The "bump" might even be a plus for a few specimens, for those whose eyes or cameras could see it?
.
.

90 CRI LED options.png (123.13 KiB) Viewed 1251 times

Tungsten Halogen spectrum.png (75.79 KiB) Viewed 1251 times

[PeteM]

To add, note the spectral transmission of a human cornea at 53 years of age. That LED bump at 450 nm will be knocked down by more than 50% - even more as we age (right up to cataract surgery, when the bluish tint of a LED lamp might prove newly annoying).
.
.

[Hobbyst46]

@PeteM : Very nicely explained. Just two small comments
- UV extends up to 390nm; wavelengths above that are visible: violet (around 400nm) to blue (around 450nm).
- I once read that the ideal daylight for microscopy, that they used 100 years ago or so, was reflected sunlight (from clouds) rather than direct sun rays. I believe that such reflected sunlight is of >6000 K color temperature. Just a thought.

[apochronaut]

I have a diatom slide on a series 10 stage right in front of me right now. It is crudely made and I know the cover glass is too thick. I didn't mic. it but it came from a dish of old used and largely discarded coverslips. I just wanted to get a quick look at something today. I am now centered on a segment of what looks to be a Licmophora that stretches almost exactly 1/2 way across the 19mm field.
The light comes from a med. blue filtered tungsten filament at 5.5 v. modulated through a very precisely aligned dark phase system. The objective over the slide is a 45 year old 40X .66 planachro dark phase. I am a stickler for alignment and have learned over the years that a tiny little refinement in the symmetry of phase annulus and diaphragm rings can have a noticeable effect on the definition of margins in phase imaging and 3 dimensionality of the image. Phase can have relief.
The image in front of me has a smooth, even, dark grey background covering the field. It is flat and free of aberrations to the edge , dark where it should be dark and bright where it should be bright : a clean brilliant ever so slightly blueish irridescence highlighting the glassy frustule.
There are about 300 striae comprised of neat rows of sharply defined dark punctae stretching along the body of the Licmophora. That's about 600 across the 475 micron field or approx. 1.3 per micron. Each and every one separate and clearly defined, little .75 micron black dots. The objective is an achromat and .66 N.A. The condenser is .90. I also have a 40X dark phase planapo of 1.5 times that N.A . If I was to put it into the series 10 , lower the stage and pop on a series 400 head, it would work perfectly fine with the series 10 phase diaphragm. The image would be breathtaking, the resolution remarkable for a 40X objective.

I struggle to understand in what way spectral data and graphs or a diy led illuminator can improve the imaging that this microscope is capable of?

[PeteM]

. . .
I struggle to understand in what way spectral data and graphs or a diy led illuminator can improve the imaging that this microscope is capable of?

The point was that a LED lamp, with equal output (lumens, properly aligned) and with essentially the same CRI (color rendering index) could do equally well. Understanding the spectral graphs and lumen output figures are a guide to getting things right.

Were it a broken illumination system, running too hot for the specimens (say pond critters drying up under a cover slip), or too dim (say, to take photos or movies with fast shutter speeds) LED offers a relatively easy upgrade path. It's especially useful for the many beautifully made microscopes with five to twenty-watt tungsten (not tungsten-halogen) lamps and a desire to take movies. Some of the old Zeiss Standard microscopes come to mind.

If the owner wanted to have a long-lived lamp, a light source that didn't drastically change color temperature with dimming, a bulb that could simply be dimmed without an array of neutral density filters to last the full 200 or so hours it's spec'd for, a portable microscope that could run on a battery, or get rid of a bulky external power supply -- those are other possible pluses.

We'd agree that tungsten and especially tungsten-halogen illumination can still provide superb images - with nice even spectral distributions and superior color rendition. Something like four of my favorite six microscopes still have their 100-watt tungsten-halogen lamps. That won't likely change anytime soon. However, the two that have been converted to high power, high CRI LED work just fine as well.

[PeteM]

@PeteM : Very nicely explained. Just two small comments
- UV extends up to 390nm; wavelengths above that are visible: violet (around 400nm) to blue (around 450nm).
- I once read that the ideal daylight for microscopy, that they used 100 years ago or so, was reflected sunlight (from clouds) rather than direct sun rays. I believe that such reflected sunlight is of >6000 K color temperature. Just a thought.

Good points. You're right about the cloudy sky - and even above 6000K:
.

Kelvin color temperatures.png (65.01 KiB) Viewed 1202 times

.

You're also right about the usual definition of visible light beginning around 400 nm.

My point about the LED bump around 450 nm is that while we generally define visible light as around 400-700 nm, we humans vary in what we can see. That 450nm peak (violet-blue) will be clear to most younger folks but not so clear to older folks like me and some others here. Whether that blue cast bothers someone is a matter of both taste and individual perceptual differences. It can be removed with filters and partially resolved in photos by changing a camera's white balance. Imaging tools provide further tools to match up what the camera sees to our standard expectations of subjects viewed in "daylight."

[apochronaut]

You neglected to mention loss of contrast due to scatter and lack of coherence. That's the led I'm talking about. How do you know that a particular diy led illumination system is as good, especially as a retrofit? I'm still looking for proof. Heat has been easily dealt with in good systems. I don't see that as a justification. If you are getting heat at the stage you need another microscope, not leds. Field microscopes have the best of illumination systems and free energy.
I'm totally open to improved microscope imaging but I have not the time or enthusiasm to be altering unbroken systems. The best way to improve a microscope is through correct alignment and better objectives.
As Vivaldi once said. If it ain't Baroque don't fix it.

[PeteM]

You neglected to mention loss of contrast due to scatter and lack of coherence . . . As Vivaldi once said. If it ain't Baroque don't fix it.

We've addressed the scatter thing before. It's best to make use of the microscope's existing illumination optics. Ideally, this means placing a single LED die at the same location as the original filament. The LED die should have about the same size as well. Until recently, these smaller single-die LED lamps weren't very bright. Some are now available (at a cost) that work well in terms of lumen output, CRI, and illumination angle.

An alternative, with some risk, is to use a brighter COB LED and then collimate it (using either the microscope's original sub-stage optics or something purpose-built). Retrofit companies sometimes add a lens. Microscopes with collimating optics, field diaphragms, and a proper condenser with an iris usually do fine. The light exits the field lens more or less collimated, can be shaped by the field diaphragm, and is further controlled at the condenser. It may not be exactly Kohler illumination, but as you've said yourself, critical illumination can do just fine.

I might add that some LED flashlights come with a focusable reflector that produces a near-parallel path to shine a small spot many yards away. It's rare that they have good CRI, and the dimmers, even when included, may cause flicker at high shutter speeds. But in principle and actuality (other headlights), LED light can be collimated as easily or more easily than tungsten bulbs.

Many of the older tungsten and tungsten-halogen illumination systems had their own problems. In cheaper scopes (say, an A.O. 150 or a Leica ATC 2000), the lamp is right below the field lens - with little space for collimating optics and a direct heat source under the stage. Even in more expensive scopes (20-100 watt systems from the "big four"), attempts to capture all the light emitted mean mirrors behind the bulb. Even after careful alignment of the filament up-down and front-to-back (often ignored), there's a lot of scatter. These scopes have collimating lenses and baffles to block that scatter - and those systems usually work well for whatever scatter may come from a LED as well.

We're both Vivaldi fans, it seems.

[apochronaut]

You win. I'm getting a new microscope tomorrow, well actually I am going to diy a new microscope. Perfect for an led illuminator. Won't melt it.
https://ideas.lego.com/projects/55c1d1a ... c1df0b90ae

All those graphs are so persuasive.

[PeteM]

The "winners" are guys like DTSH, who have repaired and modernized wonderful yet affordable older microscopes with LED retrofits, as well as research scientists with superior illumination systems. "Losers" might be people who throw away a perfectly good tungsten-halogen system and replace it with an inferior LED retrofit.

Getting some of the facts right is just in the hope of helping others, like "Scoper," decide when/if a retrofit makes sense and how to go about selecting and placing the right components. But sure, the older microscope illumination systems work wonderfully (if still working) for you and often for me.

[dtsh]

You neglected to mention loss of contrast due to scatter and lack of coherence. That's the led I'm talking about. How do you know that a particular diy led illumination system is as good, especially as a retrofit? I'm still looking for proof.

What would you accept as proof?
I am willing to do any side by side comparison you request that is within my capability.

For diatoms I have the following examples: Stauroneis pheonicenteron, Gyrosigma attenuatum, Gyrosigma reimeri, Navicula oblonga, and Pinnularia nobilis.

Heat has been easily dealt with in good systems. I don't see that as a justification. If you are getting heat at the stage you need another microscope, not leds.

Why not solve the source of the problem itself, the excess heat? Why switch to some other system when the one you have just needs better lighting?

Field microscopes have the best of illumination systems and free energy.

But only at certain times. If that's the best illumination, why bother with the heat and space requirements of halogen?

My main motivation is to fix the scopes I have without acquiring even more microscopes. If I can help someone else in the process, even better.
The 17w 1036 lamp is a tad anemic when one is using light hungry methods of microscopy, you've said so yourself in the past; so when the cord is ruined and the transformer who knows where, does it not make more sense to modernize it than to replace it with the same anemic lamp? That it can allow the instrument to run cooler and take up less space is a benefit to my mind and if it can perform as well or better than the original, why wouldn't I want to modernize it? Is it as good or better? It seems to be to me, but I am not expert; however, I am open to learning and exchanging some ignorance for knowledge.

This is my preferred conversion, but it requires drilling and tapping 2 holes which makes it more difficult for most people and that's leaving aside the entire aspect that it's a non-reversible change which many would not want.

ao10_rear_control.jpg (88.72 KiB) Viewed 1144 times

[PeteM]

I hadn't seen or noticed that control before, DTSH. To my eye, it's an elegant solution - especially if it allows ditching the external power supply and cables from early A.O. models.

The science fiction Heinlein once wrote up a paragraph about things a man should know how to do, including changing a diaper. IMO, these days it would be things a man or woman could learn to do - probably including drilling and tapping two holes. The harder part is the design, which you're happy to pass along.

[dtsh]

I hadn't seen or noticed that control before, DTSH. To my eye, it's an elegant solution - especially if it allows ditching the external power supply and cables from early A.O. models.

The science fiction Heinlein once wrote up a paragraph about things a man should know how to do, including changing a diaper. IMO, these days it would be things a man or woman could learn to do - probably including drilling and tapping two holes. The harder part is the design, which you're happy to pass along.

The drop-in I posted earlier, the last post on the first page, takes much of the same design but puts it in a small remote so there's no need for drilling and tapping; it does however, take up a bit more space on the bench. If I remember correctly, the dial knob is the same for both designs. I personally prefer the compactness of the rear bolt-on, but I admit the remote is nice in that it could be positioned to the left or right as one likes and doesn't require permanent modification. Power is supplied through a typical wall wart. I have one that takes 12v in, but boosting it to 34v for the LED produces more heat than taking 24v and boosting that to 34v, so when I needed more power supplies I ordered 24v versions. I'll start a thread with all the parts and details over the next couple of days and upload all the models.