American Optical Series 10 LED replacement for 1036A

Here you can discuss DIY adaptations to the microscope.
Message
Author
dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

American Optical Series 10 LED replacement for 1036A

#1 Post by dtsh » Thu May 25, 2023 11:52 pm

First off, this is a work in progress. I've tested it and it's passed everything I've thrown at it, but I know it's not perfect and I sincerely hope others will help me continue to develop it.
This started because I had a couple of AO10's with 1036A illuminators in which the wiring had failed long before they came into my possession, none of which came with transformers. The stock 1036A comes with a 17w lamp that's largely sufficient, but for light intensive methods it's not not up to the task, especially at higher magnifications. Given the shortcomings and the expense of acquiring missing parts and repairs, I decided to try replacing the lamp with an LED. I've done this a few other times with different LEDs like the 8w version I've been using for the past couple of years, but I wanted something that didn't require any permanent modification to the instrument. I'm going to start with the end result, then I'll go over all of the parts and assembly. Total cost should be somewhere under $50. I'll try to give the price I paid for components where I can, but it's important to remember that this is the kind of project where you can build one for X dollars, but due to shipping costs the price per unit drops significantly when buying multiples. As always for a DIY project, get a few more parts than you need as sometimes it's easy to accidentally damage a component.

There shouldn't be much advanced skill needed; the ability to solder a few components, use a multimeter, the ability to cut a mostly straight line with a hacksaw and finish it up with a file, the ability to drill a few holes, tap a couple of threads, and the use basic of hand tools. It will also help having the ability to not recoil in abject disgust at some of the very hacky things that will follow.

I invite criticism, but I expect it to be alongside suggested ways to make an improvement.

To save everyone the effort of reading a thread only to see the end and realize it's not what you're looking for, here are some images of the final project before you invest much time.
AO10_drop-in_full.jpg
AO10_drop-in_full.jpg (117.94 KiB) Viewed 5772 times
AO10_drop-in_full_bottom.jpg
AO10_drop-in_full_bottom.jpg (94.71 KiB) Viewed 5772 times
AO10_drop-in_remote.jpg
AO10_drop-in_remote.jpg (63.66 KiB) Viewed 5770 times
AO10_drop-in_remote_bottom.jpg
AO10_drop-in_remote_bottom.jpg (105.15 KiB) Viewed 5772 times
Last edited by dtsh on Fri May 26, 2023 12:16 am, edited 2 times in total.

apochronaut
Posts: 6272
Joined: Fri May 15, 2015 12:15 am

Re: American Optical Series 10 LED replacement for 1036A

#2 Post by apochronaut » Fri May 26, 2023 12:15 am

Just what did you thow at it?

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#3 Post by dtsh » Fri May 26, 2023 12:21 am

apochronaut wrote:
Fri May 26, 2023 12:15 am
Just what did you thow at it?
Money? Somewhere less than $50. It's hard to estimate as I had some parts already, but should be less than $50 with those added....should be. The price of the 3D printed parts should be minimal, but I have a 3D printer.
Profanity? Not much actually! I only burned myself a few times.
Wasted sleep? A few hours perhaps. Most of that in the preparation of the aluminum plate. I was using the plate layout I'd designed for my first LED (8w) and this one doesn't need such a large plate and one could make it a bit smaller. Could *probably* even 3D print it too, but I haven't tested that.
I left it on high for 12 hours and temps were still tolerable.

apochronaut
Posts: 6272
Joined: Fri May 15, 2015 12:15 am

Re: American Optical Series 10 LED replacement for 1036A

#4 Post by apochronaut » Fri May 26, 2023 1:21 am

Was actually asking for what regimen of performance tests. I thought " every thing I threw at it " meant that it had passed muster in terms of performance? So, I guess in the end you just threw money at it?.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#5 Post by dtsh » Fri May 26, 2023 2:24 am

apochronaut wrote:
Fri May 26, 2023 1:21 am
Was actually asking for what regimen of performance tests. I thought " every thing I threw at it " meant that it had passed muster in terms of performance? So, I guess in the end you just threw money at it?.
I've been asking for quite some time for specific tests and I've done what's been offered. The tests I've done have largely involved imaging and getting results at least on par with the original lighting. I've never claimed to be an expert, just a motivated person who likes to tinker and learn.

If you've got some tests for me to try, I will gladly do what I can to see where it's strengths and weakness are. I don't think it's perfect, but it is far better than the broken illuminator it started as.
I'm not an optical company and I'm not making this for profit, I'm just trying to make some nice instruments usable again. I have no doubt it has shortcomings, my optical and electronics skills are weak at best, but I'd like to know how to evaluate it in a useful manner so that it can be improved.

Phill Brown
Posts: 603
Joined: Mon May 24, 2021 1:19 pm
Location: Devon UK.

Re: American Optical Series 10 LED replacement for 1036A

#6 Post by Phill Brown » Fri May 26, 2023 11:42 am

Probably test with some diatoms.
Well done for taking on the project.
I'm not saying it's the way to go but could provide useful information to you,remove the head and place it sheet of white paper above the objective to get an idea of your projector,it will be clearly visible through the paper.
If you have colour banding around the edges see if anything works to reduce them.
Happy to have my suggestion shot down in flames as irrelevant.

apochronaut
Posts: 6272
Joined: Fri May 15, 2015 12:15 am

Re: American Optical Series 10 LED replacement for 1036A

#7 Post by apochronaut » Fri May 26, 2023 1:58 pm

All of the systems that the series 10 was designed for in all magnifications with 176,176A and 184 eyepieces. BF at 40,60,100,150,200,400,450,500,600,675,750 and 1000X ; DF from 40 to 675X, Dark and Bright phase 100 to 1000X. B-Minus at 100X. Even fluorescence.
High grade illumination systems are complex and highly engineered. Most led illuminators are no better than student illuminators of the past, just brighter and bluer.

Both you and Pete are making a point that putting hours and hours into diy'ing an led illuminator is important as a replacement for a broken or missing illuminator. That's a pretty spurious argument given that the series 10 was one of the most widely sold microscopes in history and there are dozens and dozens of orphaned illuminators floating around.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#8 Post by dtsh » Fri May 26, 2023 11:12 pm

apochronaut wrote:
Fri May 26, 2023 1:58 pm
All of the systems that the series 10 was designed for in all magnifications with 176,176A and 184 eyepieces. BF at 40,60,100,150,200,400,450,500,600,675,750 and 1000X ; DF from 40 to 675X, Dark and Bright phase 100 to 1000X. B-Minus at 100X. Even fluorescence.
High grade illumination systems are complex and highly engineered. Most led illuminators are no better than student illuminators of the past, just brighter and bluer.
I have some of that and will provide what I can after I've covered the build. Part of the reason I've been offering one is because I know I don't have all the things nor skill necessary for the various testing that an optics lab would do.....but as I understand it some of our members do, or at least have access to such.
apochronaut wrote:
Fri May 26, 2023 1:58 pm
Both you and Pete are making a point that putting hours and hours into diy'ing an led illuminator is important as a replacement for a broken or missing illuminator. That's a pretty spurious argument given that the series 10 was one of the most widely sold microscopes in history and there are dozens and dozens of orphaned illuminators floating around.
I'm not pushing this as far as it goes, but it's supposed to deliver > 1450 lumens, but I don't know how to verify if that's even close to reality.

I'm enjoying the learning and tinkering at probably equal parts as I do the microscopy; if you think it is pointless then I suggest you don't build one.
Care to steer me to a replacement that's better than the 1036A and will cost me less than $50?
Also, it would seem that your recommendation should do fluorescence as well by your own standards to be worth bothering with. I look forward to your reply, it will save me much time. Thanks.

Phill Brown wrote:
Fri May 26, 2023 11:42 am
Probably test with some diatoms.
Well done for taking on the project.
I'm not saying it's the way to go but could provide useful information to you,remove the head and place it sheet of white paper above the objective to get an idea of your projector,it will be clearly visible through the paper.
If you have colour banding around the edges see if anything works to reduce them.
Happy to have my suggestion shot down in flames as irrelevant.
Thanks for that advice, it made me consider the question of how the lightpath would look without the telan lens; something I have wondered before, but had never tested. It prompted me to do a quick test, the paper worked nicely, but the ceiling works too. This was with a 45x Cat.1238 dark phase objective (just what was there). It is interesting to see how the infinity system projects and it is always nice learning something new. I had been under the misconception that the infinity bundle would be more parallel; the projected image of the field iris is about 10"/250mm in diameter.
AO34mm infinity projected.jpg
AO34mm infinity projected.jpg (99.39 KiB) Viewed 5606 times
And with the field iris.
AO34mm infinity projected2.jpg
AO34mm infinity projected2.jpg (84.26 KiB) Viewed 5599 times
Last edited by dtsh on Sat May 27, 2023 12:37 am, edited 6 times in total.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#9 Post by dtsh » Fri May 26, 2023 11:49 pm

Use of existing parts motivated a lot of my parts choices based on what I had. Once I get to assembly, those of you with low tolerance for abuse of electronics components assembled by someone with a nearly complete lack of skill and understanding might want to look away.

To provide the right output voltage I used a XL6019 30W 5A DC boost converter; there are probably better solutions out there and I'd like to hear them especially if they don't significantly add to the budget.

The LED is a Cree Xlamp CMT1412 (CMT1412-0000-000N0U0A57G) 5700K which
https://www.mouser.com/ProductDetail/Cr ... cyCode=USD

The wall wart is a 24V power supply 3A; make sure the plug matches the jack, I used a 5.5x2.5mm jack.

Perhaps 250mm of 22 gauge 2 conductor jacketed PVC or silicone wire, something that's fairly flexible.

A 25mm x 25mm x 15mm heat sink

A tube of thermal glue

A 1K potentiometer. I have many of these little ones with an integrated on/off built into them so that's what I used. I have no idea how long they will last, but they've help up to my occasional use for about a year, YMMV.

A 10k trim pot.

AO10_drop-in_misc_bits.jpg
AO10_drop-in_misc_bits.jpg (204.34 KiB) Viewed 5594 times
Last edited by dtsh on Sat May 27, 2023 3:25 pm, edited 1 time in total.

Phill Brown
Posts: 603
Joined: Mon May 24, 2021 1:19 pm
Location: Devon UK.

Re: American Optical Series 10 LED replacement for 1036A

#10 Post by Phill Brown » Sat May 27, 2023 6:51 am

I've done the ceiling projection using a sheet of Matt photo paper,( my ceiling is 350 year old plaster lath)on a board.
From my testing a single pinpoint source is optimal as a starting point.
As it's possible to get a smaller pinpoint from LEDs up to 3w than an incandescent there's a good chance it'll work.
Higher power are COB so things get messy.
For DF it's less of an issue.
The colour banding is as expected,
Possible a well aligned pinhole aperture at the led could resolve the light source?
I would expect the curved surface of the LED being collimated to be cause of the fringing.

jfiresto
Posts: 342
Joined: Fri Feb 15, 2019 8:19 am
Location: Northern Germany

Re: American Optical Series 10 LED replacement for 1036A

#11 Post by jfiresto » Sat May 27, 2023 1:58 pm

dtsh wrote:
Fri May 26, 2023 11:49 pm
... To provide the right output voltage I used a XL6019 30W 5A DC boost converter; there are probably better solutions out there and I'd like to hear them especially if they don't significantly add to the budget....
The XL6019 is an XLSEMI boost regulator IC – the heart of the DC/DC converter module you are using?

As I mentioned in another thread, I am using a Mean Well IDLC-25 90–295VAC dimmable (6~100% on) constant current LED supply that would combine your wall wart and DC boost converter. It may be better or worse for your application depending on your requirements. For the other guy, it was apparently worse.

It was around 24 euros ($26) which I think is a pretty good deal for what by all reports will be a safe, reliable, and flickerless LED supply. It expects a standard 0–10V or 10V PWM dimming signal which I will draw power for off a small ca. 6 euro PCB mount power supply that will power the 12VDC fan cooling the LED. The IDLC-25A version would include a 12VDC@50mA supply, but I have not checked its price and availability.
-John

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#12 Post by dtsh » Sat May 27, 2023 3:18 pm

Phill Brown wrote:
Sat May 27, 2023 6:51 am
I've done the ceiling projection using a sheet of Matt photo paper,( my ceiling is 350 year old plaster lath)on a board.
From my testing a single pinpoint source is optimal as a starting point.
As it's possible to get a smaller pinpoint from LEDs up to 3w than an incandescent there's a good chance it'll work.
Higher power are COB so things get messy.
For DF it's less of an issue.
The colour banding is as expected,
Possible a well aligned pinhole aperture at the led could resolve the light source?
I would expect the curved surface of the LED being collimated to be cause of the fringing.
I think I could use a bit of aluminum foil to produce a pinhole mask and place it after the frosted lens. My schedule is rather full the next couple of days so I might not get to it right away, I'm not sure what to expect but that's an interesting test to try.

jfiresto wrote:
Sat May 27, 2023 1:58 pm
The XL6019 is an XLSEMI boost regulator IC – the heart of the DC/DC converter module you are using?
It is indeed based on the XLSEMI chip; however, the limit of the XLSEMI is around 40v so it won't boost to the level needed for the Mean Well's > 90v.
I've previously used some LED drivers (LUXDrive, CinCon) and was happy with them, but I was having trouble finding ones that worked with the LEDs I was intending to use. I found the LUXDrive BuckBlock to be great with it's 0-10v dimming circuit, very easy to work with and I discussed that in an earlier thread on converting an AO10 to LED (I believe around 8w for that one). Eventually a friend who knows more about electronics than I do suggested this route so I tried it and have thus far found it satisfactory. He speculated it might need some capacitors on the output to smooth it out, but that hasn't (thus far) been found to be needed.
jfiresto wrote:
Sat May 27, 2023 1:58 pm
It was around 24 euros ($26) which I think is a pretty good deal for what by all reports will be a safe, reliable, and flickerless LED supply. It expects a standard 0–10V or 10V PWM dimming signal which I will draw power for off a small ca. 6 euro PCB mount power supply that will power the 12VDC fan cooling the LED. The IDLC-25A version would include a 12VDC@50mA supply, but I have not checked its price and availability.
I know at some point I will likely have to accept a fan if I use anything higher wattage, but I have been trying to avoid them if I can so I'm not adding a potential source of vibration (I don't know that it matters with such a heavy scope (about 6kg/15lb), maybe I should try it someday).

jfiresto
Posts: 342
Joined: Fri Feb 15, 2019 8:19 am
Location: Northern Germany

Re: American Optical Series 10 LED replacement for 1036A

#13 Post by jfiresto » Sat May 27, 2023 3:40 pm

dtsh wrote:
Sat May 27, 2023 3:18 pm
jfiresto wrote:
Sat May 27, 2023 1:58 pm
The XL6019 is an XLSEMI boost regulator IC – the heart of the DC/DC converter module you are using?
It is indeed based on the XLSEMI chip; however, the limit of the XLSEMI is around 40v so it won't boost to the level needed for the Mean Well's > 90v....
The IDLC-25 takes 90-295VAC mains voltage and directly drives the LED: "Direct connecting to LEDs is suggested. It is not suitable to be used with additional drivers." I could not find a 350mA COB LED with a 49–70V drop which makes me think MEAN WELL's IDLC-25-350 customers generally string two or more LEDs in series. The 500mA output IDLC-25-500 I bought should just manage the single ca. 36V@500mA LED it will be directly driving.
I know at some point I will likely have to accept a fan if I use anything higher wattage, but I have been trying to avoid them if I can so I'm not adding a potential source of vibration (I don't know that it matters with such a heavy scope (about 6kg/15lb), maybe I should try it someday).
The ca. 1W DC fan I will be using produces less vibration than the old AC fan it is replaces. There are some really quiet, lower power, lower speed fans if you just need to keep hot air from building up around a heat sink.
-John

JohnSG
Posts: 12
Joined: Tue Mar 28, 2023 8:07 pm

Re: American Optical Series 10 LED replacement for 1036A

#14 Post by JohnSG » Fri Jun 02, 2023 4:28 pm

Following this with interest as I slowly get my AO 10s repaired and set up.

I have looked at LED light sources over the years for a variety of applications, so I can suggest a type that may be of interest. The following COB LED has a very high CRI of 98, and notably does not exhibit the giant spike in the violet. It is worth a look at the datasheet. While the total lumen output is less than the Cree, the emission area is smaller by a factor ~4, which may help in other ways. The series of LEDs has a full range of color temperatures, I just picked 4000K as a middling value. I have no meaningful experience in microscopy, just passing along something I have found.

https://www.digikey.com/en/products/det ... 3/12349859

Datasheet:
https://media.digikey.com/pdf/Data%20Sh ... un2020.pdf

Hope it's useful,
John

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: American Optical Series 10 LED replacement for 1036A

#15 Post by PeteM » Fri Jun 02, 2023 7:35 pm

apochronaut wrote:
Fri May 26, 2023 1:58 pm
. . . Both you and Pete are making a point that putting hours and hours into diy'ing an led illuminator is important as a replacement for a broken or missing illuminator. That's a pretty spurious argument given that the series 10 was one of the most widely sold microscopes in history and there are dozens and dozens of orphaned illuminators floating around.
This is my first look and first post in this thread. A few points:

The LED conversions I've done have been of two types. Dozens have been quick and cheap replacements for otherwise good microscopes for kids. Typically under $35 and an hour or two of work. Sometimes under $10. These are often old tungsten (not tungsten-halogen) systems, maybe 10 watts, gone bad. They restore a microscope to better-than-new illumination with little compromise in even field illumination. Most never had Kohler illumination. Perhaps twenty kids wouldn't have a microscope today, save for these replacements.

The second type has been a few replacements for high-end microscopes such as an Olympus BX-40 for which good replacement power supplies are rarely found (they tend to go bad) and are very costly. The original lamp is also too dim for proper DIC and darkfield, especially for taking movies. Commercial LED replacements (Retrodiode etc.) are far cheaper and better. Saul makes one that includes both a LED for viewing and a flash for stop-action vibration-resistant imaging. Along with a nosepiece adapter, they turn a BX-40 into the DIC-capable equivalent of a far more expensive BX-50. I've also used one on a BX-50, simply because replacing the broken power supply and missing lamp would cost hundreds more. With it's 4000K constant color temperature, it also works as evenly and nearly as brightly as the 100 watt bulb - without the shorter life, heat, and color temperature varying from something like 1K-3K from tungsten-halogen bulb from 30 years ago.

As for LED replacements on American Optical microscopes, I've only done one - an old finite-era A.O. phase scope - purchased years ago, Phil, on your recommendation. I've not had to replace an illuminator on an American Optical Series 10, but can surely see why DTSH's DIY approach might be of use to many who might happen upon this thread.

The 1036a illuminator has an older 18-watt tungsten bulb and requires a somewhat inconvenient external power supply. Search "American Optical 1036a" on eBay and just the hard-to-change bulb costs $34.11. Three might burn out in a hobbyist's lifetime of use - so there's $100. The cheapest replacement illuminators and power supplies currently showing on that same eBay search are well over $100 - typically for a whole new microscope. So, now we're up to maybe $200.

That 18-watt tungsten (not tungsten-halogen) bulb was what was available sixty years ago, but dim by modern standards despite a fairly decent condensing system. In 1960, the very first commercial digital camera wouldn't appear for another 15 years - and cost something like $3000. If someone in 2023 wants to make movies, say of protists, at high power or under darkfield - the old bulbs are too dim. Check YouTube - these videos are all the rage and good for kids and science. Ditto for darkfield and imaging at low ISO. There's also the constant fiddling with neutral density filters and white balance for imaging with the sixty-year-old bulbs.

Where we agree, Phil, is that it's (currently) silly to replace a perfectly good tungsten-halogen system with a DIY LED in most cases. The newer Series 10, with the 20-watt tungsten-halogen bulb, might be an example. I say might, because I've had two users concerned about the heat from these and also their tendency of these models to oxidize the contacts.

For $50 and some DIY fun, DTSH has a solution that should last a lifetime, ditch the cumbersome external power supply, cost less, be bright enough for most any task, and -- judging by the micrometer-slide images in the other thread -- provide results that look quite good.

I'm grateful he's willing to share it - and also that he's willing to try tests to be sure he's getting the best possible even illumination and freedom from aberrations these old (but still excellent) microscopes are capable of delivering.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#16 Post by dtsh » Fri Jun 02, 2023 11:35 pm

JohnSG wrote:
Fri Jun 02, 2023 4:28 pm
Following this with interest as I slowly get my AO 10s repaired and set up.

I have looked at LED light sources over the years for a variety of applications, so I can suggest a type that may be of interest. The following COB LED has a very high CRI of 98, and notably does not exhibit the giant spike in the violet. It is worth a look at the datasheet. While the total lumen output is less than the Cree, the emission area is smaller by a factor ~4, which may help in other ways. The series of LEDs has a full range of color temperatures, I just picked 4000K as a middling value. I have no meaningful experience in microscopy, just passing along something I have found.

https://www.digikey.com/en/products/det ... 3/12349859

Datasheet:
https://media.digikey.com/pdf/Data%20Sh ... un2020.pdf

Hope it's useful,
John
Thanks for noting that, it looks like it would be pretty easy to adapt that since the typical voltage is around the same point as the one I have been using. This has been a very busy week for me, but hopefully by monday or tuesday I should have a post on the wiring and component assembly, a slight adjustment of the trim pot is probably all it would take.

PeteM wrote:
Fri Jun 02, 2023 7:35 pm
As for LED replacements on American Optical microscopes, I've only done one - an old finite-era A.O. phase scope - purchased years ago, Phil, on your recommendation.
I'd love to know what your opinion has been of that scope since I've been tempted to pickup phase for the Series 15...if I ever find such.

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: American Optical Series 10 LED replacement for 1036A

#17 Post by PeteM » Sat Jun 03, 2023 12:43 am

dtsh wrote:
Fri Jun 02, 2023 11:35 pm
. . . I'd love to know what your opinion has been of that scope since I've been tempted to pickup phase for the Series 15...if I ever find such.
If memory serves, mine was a Series 4 with an old trinocular head - made before the switch to infinite optics. It had a dim tungsten bulb accessible from the back and a mirror below the field lens that had lost most of its silvering and had to be replaced. The plus was that several different types of phase objectives were (are) available for it - something like three levels of bright, dark, and B- phase and thus a possible three dozen or so options. That was of possible benefit to researchers and live cell pathologists those decades ago trying to get just the right contrast, but mostly a waste of time swapping objectives in and out for something like kids and pond critters.

The minus is the older or absent coatings lead to poor contrast, and the field size is smaller than I prefer. To my eye, the plan infinity phase objectives from the Series 10 era provide better images and several more recent microscopes better still.

I've also picked up a half dozen Series 50 phase microscopes for kids and pond critters along the way. These have lots of limitations - a 15-watt tiny spherical appliance bulb directly under the field lens, only three places on the turret, just a monocular head, difficulty in adjustment, and older achro optics - but they have been by far the best value in a phase scope I've seen for kids. Several parents use them with an old cell phone permanently attached and linked to a monitor.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#18 Post by dtsh » Sat Jun 03, 2023 2:55 am

PeteM wrote:
Sat Jun 03, 2023 12:43 am
dtsh wrote:
Fri Jun 02, 2023 11:35 pm
. . . I'd love to know what your opinion has been of that scope since I've been tempted to pickup phase for the Series 15...if I ever find such.
The minus is the older or absent coatings lead to poor contrast, and the field size is smaller than I prefer. To my eye, the plan infinity phase objectives from the Series 10 era provide better images and several more recent microscopes better still.
I was hoping to hear relatively similar performance to the non-plan 34mm infinity.

PeteM wrote:
Sat Jun 03, 2023 12:43 am
I've also picked up a half dozen Series 50 phase microscopes for kids and pond critters along the way. These have lots of limitations - a 15-watt tiny spherical appliance bulb directly under the field lens, only three places on the turret, just a monocular head, difficulty in adjustment, and older achro optics - but they have been by far the best value in a phase scope I've seen for kids. Several parents use them with an old cell phone permanently attached and linked to a monitor.
Does the 50 also have the darkfield patch like the 150's phase turret? It's been a nice instrument with the various modes and tends to be the one I grab if I'm traveling. I definitely agree the lamp is anemic, at least the socket for the bulb is easier to find.

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: American Optical Series 10 LED replacement for 1036A

#19 Post by PeteM » Sat Jun 03, 2023 5:42 pm

dtsh wrote:
Sat Jun 03, 2023 2:55 am
. . . Does the 50 also have the darkfield patch like the 150's phase turret? It's been a nice instrument with the various modes and tends to be the one I grab if I'm traveling. I definitely agree the lamp is anemic, at least the socket for the bulb is easier to find.
The phase condenser for the Series 50 has just four positions - one brightfield and three color-coded annuli. The largest annulus works for darkfield at 10x. The other two are for 10x and 40x phase objectives. The condenser attaches directly to the bottom of the stage and is a bit clumsy to work on. Probably fine for educational use back in the day - since students wouldn't be messing with centering knobs and the like.

I'd be inclined to stick with your 150, with four or five places for objectives, the usual phase condenser, and possibly one of your LED retrofits for a robust and battery-powerable travel scope.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#20 Post by dtsh » Sat Jun 03, 2023 8:29 pm

PeteM wrote:
Sat Jun 03, 2023 5:42 pm
The phase condenser for the Series 50 has just four positions - one brightfield and three color-coded annuli. The largest annulus works for darkfield at 10x. The other two are for 10x and 40x phase objectives. The condenser attaches directly to the bottom of the stage and is a bit clumsy to work on. Probably fine for educational use back in the day - since students wouldn't be messing with centering knobs and the like.
I don't use it often and misremembered the turret, it would seem the same design with 3 annuli and brightfield; same as the 50.

PeteM wrote:
Sat Jun 03, 2023 5:42 pm
I'd be inclined to stick with your 150, with four or five places for objectives, the usual phase condenser, and possibly one of your LED retrofits for a robust and battery-powerable travel scope.
It has been a good travel instrument so that I can take an instrument on trips and if it were destroyed or desired I'd not feel out much if it didn't return home. It came with an AO hardshell, which itself needs some repair, but it makes a convenient package with just enough room to tuck in slides and other sundries...even a box of kimwipes. I don't use it often, but I think if I ever have to work on the lighting it will get an upgrade. It certainly would be nice to install a rechargeable battery into the base, a portable phase contrast scope that doesn't require power or cables would be convenient.

Here's my 150 if anyone is interested. https://www.microbehunter.com/microscop ... 24&t=12867

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#21 Post by dtsh » Tue Jun 06, 2023 6:54 pm

Here's where we get to the gross parts where I do things in a very unprofessional manner.


I'm being extra verbose here because when I first tried this sort of thing a few years ago I would have appreciated some extra notes for someone who wasn't experienced with building circuits (I'm still not, but I can build this no matter how gross it is)
Using a bit of heatshrink for solder joints is recommended since things will be packed in tightly to reduce the chances of an accidental short. Remember to pre-tin all parts before soldering to get a good bond.

Get out a multimeter, set it for ohms (greater than 11k) and connect it to the center pin and either of the other two pins of the new trim-pot. You should be able to read the resistance between them and as you turn the pot with a small screwdriver, the resistance should change. I initially set this trim-pot resistance to about 7.6K ohms. While it isn't necessary, I trim the 3rd pin off the trim-pot so it doesn't interfere.

Now we want to solder the trim pot to the potentiometer. First we need to determine which pins on the potentiometer we are using since we will only be using 2 out of the 3. If you look at the potentiometer with the pins facing you, you will see 3 pins; the left one is #1, center is #2, and the right one #3. We will be soldering one of the trim-pot's pins to the center pin on the potentiometer. Connect your multimeter to the green potentiometer on pins #1 and #2 and turn the potentiometer through it's range of motion, when you feel/hear the click of the switch inside the potentiometer the resistance should be 0 and around 1K ohms at the other extreme; if you're getting the opposite, use the other outer pin. You can use a jumper, but I soldered the pin on the potentiometer directly to a pin on the trim-pot.
AO10_drop-in_pot_ass.jpg
AO10_drop-in_pot_ass.jpg (148.88 KiB) Viewed 4984 times
After soldering them, connect your multimeter to the remaining pin on the trim-pot and the pin on the potentiometer and test again. Rotate the shaft on the potentiometer so that when the internal switch activates the resistance should be about 7.6k and when at the other extreme is should be around 8.6K, it doesn't need to be exact but it needs to be around 1K ohm of difference.
Set the potentiometer and trim-pot aside for now.

At this point we're going to locate the pins for the blue trim pot on the XL6019 boost converter, they're easy to see on the bottom and with a bit of desoldering braid the trim pot should come off.
AO10_drop-in_boost_back.jpg
AO10_drop-in_boost_back.jpg (120.17 KiB) Viewed 4994 times
AO10_drop-in_boost.jpg
AO10_drop-in_boost.jpg (134.28 KiB) Viewed 4994 times
Once that trim pot is removed, solder two small jumper wires into the two outer holes, the middle one isn't used. Since this is a resistance circuit to control the output voltage it doesn't matter which wire connects to which pin on the potentiometer/trim-pot assembled we made, only that the resistance is right. Connect one wire to the pin on the green potentiometer and the other to the pin on the trim-pot and set the whole works aside for a minute. The green potentiometer I use has a built in switch that uses the lower two pins, we'll use those in a later step during assembly.

Now we assemble the barrel jack. We will need a lead for the + and - on the barrel jack. The jacks I have get soft and malleable when heated, so I found it handy to insert the plug to hold the pin in place. Once those are soldered in place we can test the whole assembly. Connect the + and - from the barrel jack to the associated + and - terminals on the XL6019 labelled as "IN +" and "IN -", the center pin on the barrel jack should be positive(+). You can destroy the boost converter if the leads are reversed, so make sure it's + to + and - to -.

Now switch the multimeter to voltage and set the range to something higher than 48v DC and connect the probes to the "OUT" + and - terminals. Plug the power supply's jack into the barrel jack and plug it in to the wall; you should see an output voltage on the meter. Turn the shaft of the green potentiometer clockwise and the voltage should rise; turn it all the way and note the voltage on your meter. We want to shoot for the typical voltage of the LED, in this case 34.2v and we get there by adjusting the trim-pot until it reads 34.4v or a bit under with the potentiometer turned to max. Now as you turn the potentiometer shaft counter-clockwise, the voltage should drop. If not, check things again.

The next step is to disconnect power and to disconnect the terminals of the boost converter leaving the barrel jack with it's lead wires from the XL6019 boost converter with potentiometer/trim-pot assembly.

Next up is putting it in the remote housing. It will be a tight fit so wires should be no longer than necessary once assembly begins.
Last edited by dtsh on Sat Jun 10, 2023 12:23 am, edited 1 time in total.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#22 Post by dtsh » Sat Jun 10, 2023 12:16 am

I wasn't happy with the images I took, so I ended up assembling another one.

Here is an image of the XL6019 and potemtiometer assembly soldered up, ready for testing.
AO10_drop-in_XL5019_ass.jpg
AO10_drop-in_XL5019_ass.jpg (130.31 KiB) Viewed 4870 times
Here it is temporarily assembled for testing. Note, the switch built-in into the potentiometer isn't connected, power is unswitched for this testing; watch for smoke.
The datasheet says 34.2v, so try to adjust the trimpot so it's close to that. A little turn is an understatement; this would probably be better with a resistor and a smaller 5k or 1k trim-pot, but I was trying to keep this simple.
AO10_drop-in_assembly_test.jpg
AO10_drop-in_assembly_test.jpg (120.21 KiB) Viewed 4870 times
Here it is assembled into the body, note the barrel jack needs to be inserted before soldering.
AO10_drop-in_assembly1.jpg
AO10_drop-in_assembly1.jpg (124.61 KiB) Viewed 4870 times

And how it looks once you put the pot in place.
AO10_drop-in_assembly2.jpg
AO10_drop-in_assembly2.jpg (99.64 KiB) Viewed 4870 times
All it needs now is the two leads to an LED.

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#23 Post by dtsh » Mon Jun 12, 2023 3:40 pm

Here is a link to the models for the body and dial.
https://www.thingiverse.com/thing:6074859

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#24 Post by dtsh » Thu Jun 15, 2023 7:44 pm

Silly as it may sound, the part I'm most proud of is the simple sheetmetal clip that holds the LED and heatsink. I went through a number of iterations of ways to hold the heatsink in place, but the easy ones seemed to not allow for much adjustment to get a decent alignment. After a few attempts and some time to think, I thought about something like a machinist's T-slot clamp and considered ways to put a slot in the plate for someone without access to a mill and then I came up with this with the slot cut into the clamp as it should be easier for someone with a dremel or perhaps just a drill and some files. It is simple to fabricate, nothing more complex than a drill, some small files, and a vise or other way to make the bends. It allows adjusting in X and Y and can be shimmed to gain Z if needed.
AO10_drop-in_LED_clip.jpg
AO10_drop-in_LED_clip.jpg (117.19 KiB) Viewed 4559 times
I ended up using a threaded insert instead of tapping threads into the hole as it allows a more threads and thus more travel for the M3 bolt to give clamping action.

A simple ziptie to provide some stress relief to the lead wires and it's good.
AO10_drop-in_LED_plate.jpg
AO10_drop-in_LED_plate.jpg (102.91 KiB) Viewed 4559 times

JohnSG
Posts: 12
Joined: Tue Mar 28, 2023 8:07 pm

Re: American Optical Series 10 LED replacement for 1036A

#25 Post by JohnSG » Mon Jun 19, 2023 2:30 pm

Thanks for making this public.

John

Scoper
Posts: 153
Joined: Tue Jan 24, 2023 4:06 pm

Re: American Optical Series 10 LED replacement for 1036A

#26 Post by Scoper » Mon Jun 19, 2023 4:47 pm

I agree…thanks posting this very useful info.

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: American Optical Series 10 LED replacement for 1036A

#27 Post by PeteM » Mon Jun 19, 2023 5:38 pm

Clever bracket design.

JohnSG
Posts: 12
Joined: Tue Mar 28, 2023 8:07 pm

Re: American Optical Series 10 LED replacement for 1036A

#28 Post by JohnSG » Tue Jun 20, 2023 8:13 pm

Will you be posting the dimensions of the T-shaped aluminum baseplate?

Thanks,
John

dtsh
Posts: 977
Joined: Wed May 01, 2019 6:06 pm
Location: Wisconsin

Re: American Optical Series 10 LED replacement for 1036A

#29 Post by dtsh » Tue Jun 20, 2023 8:27 pm

JohnSG wrote:
Tue Jun 20, 2023 8:13 pm
Will you be posting the dimensions of the T-shaped aluminum baseplate?

Thanks,
John
Ooooh, where's your listening device in my home?!?!? I was just getting to that.

I had a template I had been using for my own purposes, but it wasn't suitable for use by anyone else so I had to redesign it. I added a 3D modeled baseplate to the thingiverse design as a reference; the two main measurements are that the two mount holes are about 57.5mm apart and the hole for the M3 threaded insert is setback 12mm from the centerline of the two mount holes. It should be possible to 3D print the mount plate and use it as-is with a thin plate of something with a reasonably large surface area between the plate and the heatsink, but I haven't tested that to prove it will work for extended use. Speculating, a think bit of metal or possibly even cardboard stock might do the trick for anyone unable to get/make a plate. If you try it let me know your results. The biggest problem I foresee would be the material raising the LED higher than it needs to be; that could be countered by putting a recess into the plate to accommodate... My minimum test thus far is letting it run at the highest setting for 8 straight hours.

The modeled plate is 5mm thick, but the ones I've been making are 6mm so there's a little wiggle room if needed.

Do you think a marking template would be useful? It wouldn't be difficult to model one based off the rest.

JohnSG
Posts: 12
Joined: Tue Mar 28, 2023 8:07 pm

Re: American Optical Series 10 LED replacement for 1036A

#30 Post by JohnSG » Wed Jun 21, 2023 6:05 pm

dtsh wrote:
Tue Jun 20, 2023 8:27 pm
Do you think a marking template would be useful? It wouldn't be difficult to model one based off the rest.
Yes, at least for me. My guess is that most of the LED heat is transferring to the baseplate, so I think using a 3D printed baseplate will result in substantially higher LED temperature. There will be very little convection inside the base cavity, so the fins on the heat sink won't be doing a lot. The clamp arrangement doesn't help with this. But, if the heat sink has good thermal contact with the baseplate, the latter has a lot of surface area to transfer heat to the surroundings, so this is probably a better path. This is a guess, but I have a number of years of power electronics experience and a big part of that is experience with thermal management, so it is an educated guess. If I'm correct, a little thermal grease between the heat sink and the base plate will help. If you try it someday, use a very thin layer of thermal grease.

I can be pretty sure I won't be building this soon, because I can barely carve out time to get some microscopes going. I have just about gotten my bright field AO 10 into usable shape, and it's taken months. Just planning ahead.

John

Post Reply