Glare problem with 144 LED right light for microscope

[Electro Fan]

I think this is one of your best so far and after following your posts I think you are on a good path.  Plz remind us of what scope and camera configuration and settings were used in this one.  Thx

The quality of that one is totally down to lighting. I have two LED gooseneck lights shining sideways at a sharp angle from about an inch away. The sideways lighting prevents glare and the close placement provides extra light for the camera sensor.

That video is taken with the Eakins trinocular microscope (AmScope clone) and the Eakins IMX290 autofocus camera. This camera only works well on this microscope with very strong illumination (bright to the point of absurdity I would say.)

I always use the RaspberryPi HQ camera now. The sensor performs much better in normal lighting (2x2 binned mode) and it's easy to capture a good image without being obsessive about moving the lights as close as possible. I ssh into Linux on the RPi to record directly and don't need a HDMI-USB capture device either. Much cheaper and easier to get good results, usually no post processing to boost exposure needed, highly recommended for Linux aficionados.

Your results and figure it out tenacity are very cool.

I’m pretty sure that much of microscopy, like photography itself, comes down to lighting.  And I’m pretty sure that leads to the amount of light, the direction/angle from which it shines, the type / color temp of light, and the number of sources - so I’m looking to experiment with LED rings and goosenecks to see what causes what.

On the camera, as much as I like making tools to make tools so I can have a tool to use, I’m inclined to start a tad simpler.  (Maybe I’m missing something - probably a lot - but “raspberry pi camera” sounds like an extra step or three in experimenting.)

I think I’d prefer a mostly off the shelf camera that would capture and allow for a display similar to what my eyes would see when looking into the microscope.  So for starters I’d like to avoid post processing and even a PC - just see the live image on a monitor via a simulfocal camera (although a camera with a built-in card writer would be very nice.)  Any suggestions on what specs (resolution, fps, aperture, etc) and/or what hdmi? manufacturer/model might suffice?  Thx!

[lukego]

Another example of overhead ring-light vs side lighting (gooseneck leds).  Both views were instructive.

That's a really nice comparison.

The directional lights produce a much more pleasing image. I do think that if you had something wet in the frame, like flux or IPA, than it would be almost invisible under that illumination though. That's why I like to have a lot of sideways illumination for picture quality and a small/adjustable amount of top illumination for adding "just enough" glare to see texture details.

[lukego]

On the camera, as much as I like making tools to make tools so I can have a tool to use, I’m inclined to start a tad simpler.  (Maybe I’m missing something - probably a lot - but “raspberry pi camera” sounds like an extra step or three in experimenting.)

This sounds right. I'm totally sold on the RPi, but it's a complex beast to setup, and I'm leveraging experience with Linux/ssh/ffmpeg/nfs/etc that I already have.

I would suggest finding a camera that works well in low light. This will make your life easier because the camera on a trinocular port will be able to capture the same image as you are seeing through the eye pieces. RPi HQ performs well, thanks to its large sensor with decent sensitivity and ability to use the full sensor at appropriate resolutions. I'm sure other more traditional cameras perform well too but I'm not sure how to evaluate them on paper or what to recommend.

Avoid the trap of using a camera like the Eakins IMX290 with low sensitivity. Then you will end up needing way too much illumination. It will look uncomfortably bright through the eyepieces, and you will still probably need post-processing to get a decent image. If you look at Steve Gardener from SDG Electronics he is even going so far as to build a custom ring light for his Eakins IMX290 camera that will probably get a decent image at the expense of blinding him on the eyepieces... I went pretty far down that path myself spending a few hundred $ on illumination before switching cameras.

[Electro Fan]

On the camera, as much as I like making tools to make tools so I can have a tool to use, I’m inclined to start a tad simpler.  (Maybe I’m missing something - probably a lot - but “raspberry pi camera” sounds like an extra step or three in experimenting.)

This sounds right. I'm totally sold on the RPi, but it's a complex beast to setup, and I'm leveraging experience with Linux/ssh/ffmpeg/nfs/etc that I already have.

I would suggest finding a camera that works well in low light. This will make your life easier because the camera on a trinocular port will be able to capture the same image as you are seeing through the eye pieces. RPi HQ performs well, thanks to its large sensor with decent sensitivity and ability to use the full sensor at appropriate resolutions. I'm sure other more traditional cameras perform well too but I'm not sure how to evaluate them on paper or what to recommend.

Avoid the trap of using a camera like the Eakins IMX290 with low sensitivity. Then you will end up needing way too much illumination. It will look uncomfortably bright through the eyepieces, and you will still probably need post-processing to get a decent image. If you look at Steve Gardener from SDG Electronics he is even going so far as to build a custom ring light for his Eakins IMX290 camera that will probably get a decent image at the expense of blinding him on the eyepieces... I went pretty far down that path myself spending a few hundred $ on illumination before switching cameras.

Ok, I’m following your lead but why does this guy seem to be such a happy camper both in terms of the camera and his eye piece viewing?  His lighting tutorial is compelling.

https://youtu.be/5VzrGNmIApw

[jfiresto]

Ok, I’m following your lead but why does this guy seem to be such a happy camper both in terms of the camera and his eye piece viewing?  His lighting tutorial is compelling.

https://youtu.be/5VzrGNmIApw

I did not see any eyepiece viewing in his video. Instead, he appears to mount the camera on a monocular macroscope and solely view its video output on an LCD monitor. Compared to a Greenough stereo microscope, his macroscope has much simpler internal optics, has no beam splitter to halve the light, and can have and has a bigger objective that gathers much more light. He can also crank up the illumination as much as he wants to compensate for the camera's insensitivity, with no danger of blinding the operator to produce a bright enough image for his viewers. Look at how much he lights up the table for his macroscope video. [EDIT: Look at the dimmer slide. If he could, I bet he would push his goose neck spots to 11.]

[Electro Fan]

After watching the video again I think I can answer my own question above.  NorthridgeFix gets such good results with the ringlight plus gooseneck for both eye piece viewing and the camera because he has decided that ~4” working distance is enough to do his nice soldering work.  He points out that he isn’t using a Barlow lens because while that would provide more working space it would change the lighting angle from the light ring which would negate his very effective lighting.  I have been following the general consensus? path often advocated around here that a 0.5 Barlow provides an optimum soldering working distance of about 8” - which I can believe provides good soldering ergonomics.  But maybe that’s a bit of herd or early adopter reasoning?  I don’t know since I am a rookie at soldering for through hole and a complete beginner with SMD.  Maybe the answer is that we should be soldering with NorthridgeFix’s four finger rule of thumb.  I can completely buy into the concept that as the lens rises the light ring angle is just bouncing the light down and back up into the user’s/camera’s view rather than washing the light across the PCB area of interest.  I saw the discussion about a polarized ring light and that seemed to get somewhere but not quite all the way there.  I also saw that Amscope sells a ring light that has selectable segments which might help but probably still is dependent on the steepness or shallowness of the angle as dictated by the ringlight height above the PCB.

On a related note I “liked” a comment around here that said don’t worry if a tri/simul port config loses some light vs regular stereo because you can always add more light.  That got me happy about going with a 3rd port (vs 2) but now it’s clear (haha) that just turning on a ton of light is not the full equation - we need enough light from the right angle(s), and likely from 2 (or more?) directions.

Net, net:  if the two primary sources of light (beyond ambient) are ring and gooseneck lamps, then we need to think about either the height of the ring - or maybe the diameter of the ring?  It seems likely to be unwieldy but maybe a larger diameter ring light (kind of like people use for Zoom videos?) might be able to create the desired angle NorthridgeFix likes, but at the ~8” working distance made possible with a 0.5 Barlow.

Just some thoughts....

[Electro Fan]

Ok, I’m following your lead but why does this guy seem to be such a happy camper both in terms of the camera and his eye piece viewing?  His lighting tutorial is compelling.

https://youtu.be/5VzrGNmIApw

I did not see any eyepiece viewing in his video. Instead, he appears to mount the camera on a monocular macroscope and solely view its video output on an LCD monitor. Compared to a Greenough stereo microscope, his macroscope has much simpler internal optics, has no beam splitter to halve the light, and can have and has a bigger objective that gathers much more light. He can also crank up the illumination as much as he wants to compensate for the camera's insensitivity, with no danger of blinding the operator to produce a bright enough image for his viewers. Look at how much he lights up the table for his videos.

You are right - he seems to only be using the camera without any eye pieces.  That’s not the dual use case but I think his insights on working distance and ring light angle are still very relevant.

[jfiresto]

I find microscope ring lights tend to wash out an image because of reflections, even more so as you increase the working distance and angle the illumination closer to vertical. I would gladly accept the light loss of using cross polarizers to eliminate glare and restore contrast. In the meantime, more oblique lighting from multiple angles works for me.

[Electro Fan]

Kind of off topic but 2 other observations:

1) NorthridgeFix doesn’t seem to have any issues with boom jitter.  Maybe at 4” vs 8” it’s easier to let the highly flexible boom do it’s ease of positioning flex magic and not have to wish for the stability of a more rigid but less flexible boom.  Maybe he benefits from the shorter working distance and/or maybe he is less likely to bump the head since he goes without eye piece viewing.

2) I only heard of Eakins for the first time about a week ago and it seemed like folks were ready to put their products in some tier beneath Amscope but so far it looks to me like Eakin’s products can compete (and some might come from the same source).

[Electro Fan]

I find microscope ring lights tend to wash out an image because of reflections, even more so as you increase the working distance and angle the illumination closer to vertical. I would gladly accept the light loss of using cross polarizers to eliminate glare and restore contrast. In the meantime, more oblique lighting from multiple angles works for me.

Ok, so it has been written, so it shall be done. 

[lukego]

My barlow lens preference is 0.7x, then 0.5x, then 1.0x. I have all three and switch as needed. (Sure, I know the 1.0x is really just a piece of glass to protect the objective and not strictly speaking a barlow lens.)

If you are working with hot air then 1.0x can feel very cramped. If you are working over a preheater then 0.5x might be too long and stretch your neck. Similarly, at 1.0x you can only see a really small area of the board while at 0.5x it can be hard to do very fine close-up inspection. I find 0.7x a good compromise between all those factors that works 90% of the time for SMD work/rework.

I agree with all the comments above that some Youtubers get really excellent image quality using a simple ring light on a monocular microscope with a short working distance. I suspect that you'd want a second strong directional light to create a shadow that you can use to perceive depth. Maybe a good monocular setup would be competitive with the binocular/trinocular scopes overall, I'm not sure, maybe I will try to set that up using my Eakins autofocus camera and spare lenses.

[Electro Fan]

one more data point from an Amazon review

https://www.amazon.com/gp/aw/d/B00JZJO7YC/ref=psdcmw_393248011_t1_B07ZRJQ177#aw-udpv3-customer-reviews_feature_div

Very good light ring for low distance work.
I like this light ring for my amscope Sm-4b. It works very well. The brightest is well bright enough! The variable brightness knob is very useful,and helps to get the best view on your project,depending on how much light you need.

One thing people don’t really seem to mention,is that at a longer working distance(like with the 0.5x Barlow lens),the quality of optics is greatly deteriorated. The lights from the ring are angled and designed perfectly for the 4 inch working distance the Sm scopes provide. But when you go to 8 inches,the angled lights shine DIRECTLY onto what you’re looking at,from all directions,and this causes gray distorted colors. When you hold the light ring in your hand,you could see this. Have it aligned with the head of the scope,and move it closer to your object(for me it’s a circuit board I solder on) ,and you see it gets colorful and clear,since the lights shine around what you’re looking at. But when you move it farther away,the angled lights have more time and distance to cross over and shine across each other and directly onto what you’re viewing. This causes the gray and reflective distortion. A gooseneck may be best in this case,or somehow attach the light ring to “hang down” from the scope so that it is 4 inches from the object of view.

I only mention the above to help anyone that gets the ring and is confused about optics being grayed. I saw a review on here about that. They probably used a Barlow lens. Perhaps the 0.7x wouldn’t be as bad since it’s a 6.4 inch working distance.

Otherwise,this light ring is very useful. For the price,I can’t argue!

[jfiresto]

I find microscope ring lights tend to wash out an image because of reflections, even more so as you increase the working distance and angle the illumination closer to vertical. I would gladly accept the light loss of using cross polarizers to eliminate glare and restore contrast. In the meantime, more oblique lighting from multiple angles works for me.

Ok, so it has been written, so it shall be done. 

Sorry, I am probably parroting the obvious. I need to find a quieter place where I can get more sleep.

Perhaps I should stop posting for a while. I should definitely stop making significant purchases on ebay. I probably did not need the lightly used E-PL7 I bought yesterday, but it was cheap compared to the cost of a good microscope camera with a 4/3-inch sensor.

[Electro Fan]

I find microscope ring lights tend to wash out an image because of reflections, even more so as you increase the working distance and angle the illumination closer to vertical. I would gladly accept the light loss of using cross polarizers to eliminate glare and restore contrast. In the meantime, more oblique lighting from multiple angles works for me.

Ok, so it has been written, so it shall be done. 

Sorry, I am probably parroting the obvious. I need to find a quieter place where I can get more sleep.

Perhaps I should stop posting for a while. I should definitely stop making significant purchases on ebay. I probably did not need the lightly used E-PL7 I bought yesterday, but it was cheap compared to the cost of a good microscope camera with a 4/3-inch sensor.

Nah, keep posting.  Your posts are always helpful, for sure.  Read, eBay/Amazon/AliExpress, Test, Write, Sleep.  Keep up the good R&D 

[Electro Fan]

Man, I didn’t even know what a E-PL7 is.  I just looked it up, that looks very nice.  I don’t have any four thirds anything.... now you have me thinking maybe it’s a system path that should be explored.  My plan was to get a microscope with a simulfocal port and just leave off the video camera for now, just make it Phase II.  Actually, Phase III.  Phase IIA was maybe learn how to solder through hole better and Phase IIB would be SMD.  We are all showing each other a ginormous field of rabbit holes.

[Electro Fan]

Hey Rabbit Buddies - look over here.  That’s cool, I was just there.  No, look over HERE.  Yeah, check that hole out!  Yep, it goes vertical and then horizontal and then vertical again.  Hmmm, mine had a diagonal.  Maybe your rabbit hole is newer than mine.  Guys, guys - over HERE.  Check this out!  Rabbit holes, rabbit holes, Endless rabbit holes.  🐇

[lukego]

Seriously relatable with the rabbit holes!

I've pretty much sorted out my soldering lab and techniques now, including the microscopy. Took about a year. I haven't actually soldered by big 1500-pin 4x4cm BGAs yet but I'll burn that bridge when I get to it.

I'm now diving down the rabbit hole of automatic PCB routing: How do you write a program to layout a PCB including 25Gbps SERDES links with sufficient signal integrity? Maybe I'll know the answer to that in another year...

Meanwhile I also need to learn how to debug the stuff that I'm assembling! Maybe that will require a one year detour along the way...

[Electro Fan]

Seriously relatable with the rabbit holes!

I've pretty much sorted out my soldering lab and techniques now, including the microscopy. Took about a year. I haven't actually soldered by big 1500-pin 4x4cm BGAs yet but I'll burn that bridge when I get to it.

I'm now diving down the rabbit hole of automatic PCB routing: How do you write a program to layout a PCB including 25Gbps SERDES links with sufficient signal integrity? Maybe I'll know the answer to that in another year...

Meanwhile I also need to learn how to debug the stuff that I'm assembling! Maybe that will require a one year detour along the way...

I'd say if you are looking at 1500 pin BGA soldering or working on 25 Gbps PCBs that your rabbit holes might be somewhere in the Earth's crust and and headed for the mantel.  My soldering and PCB work is more like at about the depth of the cup on a putting a green.

[Electro Fan]

Ok, new/adjacent rabbit hole.

It has dawned on me that while I think I might be soldering some SMDs that there is a lot of discussion about alternative techniques.  So maybe before I settle on the appropriate microscope working distance for viewing, lighting, etc. perhaps I should better understand the likely tools - in case they are more than / different than a soldering iron.

I've searched a fair amount... lots of bits and pieces but no big/full picture yet.  Anyone have a link to a tutorial that gives a good overview of conventional soldering vs hot air, rework stations, soldering tweezers (these look useful but I haven't seen as much on these as on hot air), etc, etc?  Thx

[lukego]

I've searched a fair amount... lots of bits and pieces but no big/full picture yet.  Anyone have a link to a tutorial that gives a good overview of conventional soldering vs hot air, rework stations, soldering tweezers (these look useful but I haven't seen as much on these as on hot air), etc, etc?  Thx

Good question. I don't know of one. I've been trying to work this out first hand buy acquiring most of the interesting-looking tools and seeing how much I use each one. I've been a really big fan of hot tweezers (JBC NASE) for fine SMD work like replacing 0402 passives but I now have a smaller hot air station (Quick TR1100) that might be a much cheaper and easier solution to the same problem. I'm thinking a lot lately about the question: if I were starting over, what would I want to get and in what order? (e.g. start with a big iron and a big hot air gun, later add smaller ones as a luxury, also get a preheater if working lead-free, something like that.)

On the microscope though I really appreciate flexibility. I was really frustrated when I was using a camera that could only produce a good image in very specific circumstances e.g. at a short working distance with very strong and very close illumination. I wouldn't like to have a microscopy setup that's optimized for only one working style.

[jfiresto]

... I've searched a fair amount... lots of bits and pieces but no big/full picture yet.  Anyone have a link to a tutorial that gives a good overview of conventional soldering vs hot air, rework stations, soldering tweezers (these look useful but I haven't seen as much on these as on hot air), etc, etc?  Thx

You might also research reflow ovens and hot plates. I picked up a 2000W electric Ceran grill to explore and refine the latter, and will return to that once I finish some microscope projects.

[lukego]

You might also research reflow ovens and hot plates. I picked up a 2000W electric Ceran grill to explore and refine the latter, and will return to that once I finish some microscope projects.

That is a really important point. You have to decide what are your own goals and whether you are more interested in assembly or rework. You can get a good assembly setup much cheaper and easier: an oven, a $20 vacuum tool, and some solder paste. Rework might only need a hot air gun but it can also rapidly become expensive with the endless temptations of specialist tips, irons, hot tweezers, air tools, etc that really are a great pleasure to use.

I decided early on that I'm mostly interested in rework. I further decided that I will treat assembly as a special case of rework: a fresh new board is "really" just a board on which every component is faulty (missing) and needs to be corrected (soldered on.) So here I am now with almost ten different iron/tweezer/air soldering tools but no reflow oven and having never tried applying a stencil, just because I am not interested in trying that until after I have mastered rework.

Most people think I'm being very eccentric and suggest it would be much cheaper, faster, and more reliable to use stencil and reflow oven. I'm sure that's true when everything goes well. I'm just imagining that sooner or later I'll need to do some important rework of expensive boards and I'll be glad to have mastered the skills instead of making my all my beginner mistakes "on the job" with production boards. (I also wonder how other people developed their rework skills in the first place, and how much money they spend on buying fresh components for each prototype instead of reusing more old ones.)

Lots to think about and some very personal decisions to make. I suppose that it comes down to budget too: do you want to do things as economically as possible or is there a particular amount of money that you are trying to invest in developing your electronics capabilities.

[Electro Fan]

This might be a way to roughly halve the working (and lighting) distance from approximately 8" to 4" (plus add some stability to a PCB) in case anyone wanted to see how the 4" view and working distance compares with the 8" view and working distance.  Or it might be more cost effective to use a "1.0" vs a 0.5 Barlow. 

Accessories can be like shallow rabbit holes

[lukego]

I use Omnivise when I'm working. This raises the PCB and then you have to raise the microscope by the same distance to make it focus. That's too high for my neck with the 0.5x Barlow.

Aside: He mentions at the end that it's hard to use an Omnivise with a small board and a preheater. I have a solution to this. I put the PCB into a Stickvise and then that into an Omnivise. Photo.

[lukego]

btw: the main issue with working distance for me is not the elevation of the PCB but rather how much vertical space I have to maneuver tools e.g. hot air gun that ideally wants to be vertical to direct air downwards onto the chip.