I’ve been looking to buy a microscope that can capture images of small tool edges and features with decent resolution and quality. Here’s the big question I’m looking to have answered – are digital microscopes any good yet?
I last shopped around for a microscope in 2019, and at that time I was discouraged by the poor quality of consumer models and the astronomical prices of professional equipment.
It seemed that there was no middle ground between toy-like mini digital microscopes and professional optical microscopes with pricey cameras attached.
Advertisement
Looking around a bit recently, I found that there are brand new models, some of which look good, at least on paper.
I all but dismissed this category, based on the couple of demos I’ve seen over the years, and so I wasn’t prepared for all of the many new options.
In shopping around, I noticed a few things.
First, there seem to be very few manufacturers. AmScope, shown above, has many different models but from what I have seen only a few have unique-looking designs. To AmScope’s credit, they provide greater details and specs about their scopes than most other brands.
A lot of the other brands I’ve looked at have very appealing marketing materials, but my optimism falls apart when I realize several other imaging brands offer the same or nearly identical product for less.
I’ve also been seeing exaggerated claims.
Advertisement
Andonstar, for example, boasts that their digital microscope can deliver more than 4000X magnification.
Maybe, if you’re looking at say a 1/64 object on a 65″ TV. But what kind of quality can one expect?
There are plenty of $100 to $150 options, and then features and capabilities vary wildly between $150 and $1500 price points.
I’m looking for something with better image quality than the average webcam.
Some look to offer built-in measurement and annotation features, which I suppose can be nice since I’m loath to deal with software that might not be supported down the road.
I have been looking at the higher end of the spectrum as well.
My goal is to examine things that are incredibly difficult to photograph with a camera – sandpaper grit, knife edges, drill bit wear, screwdriver tip wear – things like that.
Being able to measure features is going to be the difference between consumer and professional models.
OC White has an interesting product line, but it doesn’t look like they’ve updated any of their models in a few years.
The Leica Emspira 3 looks fantastic, from what I’ve seen so far, but likely has an industrial-grade price tag.
I’ve requested a pricing quote.
That’s how you know something is going to be extremely expensive, when a company won’t disclose pricing until a salesperson is assigned to talk with you.
The Olympus DSX1000 is likely an order of magnitude greater than what I could reasonably hope to budget for.
Looking at the spectrum, from $100 models to “let’s talk about your needs and budget” industrial and institutional options, it seems to me that the big difference is whether we’re talking about a webcam sensor with Barlow lens or focusing tubes, or an integrated optical and digital microscope system.
The Zeiss Visioner 1 looks good too, and has special tech to expand the depth of field. It likely costs more than my car, which puts it out of the realm of imagination.
Here’s another AmScope, an HDMI model that can capture images and video to MicroSD cards. It’s under $900.
Depth of field is going to be a challenge for many microscopes, where not much will be in focus unless something is flat. This is partially why most digital microscopes show off circuit board components in their sample images, aside from this being a key target user application. Circuit boards are relatively flat with short components. With good lighting, it’s easy to make them look good.
Something like a Phillips screwdriver tip is going to be a blurry mess. With a standard digital camera, capturing closeups of items with depth often requires focus stacking, where multiple images are taken at slightly different distances and stacked together in software. It’s extremely tedious and impractical.
All of these consumer models, whether $90 or $900, seem limited to very flat objects, such as stamps, coins, and circuit boards with surface-mount components.
I want something that can look at a screwdriver or drill bit tip with full depth of field, and in seconds rather than the many minutes it takes to set up lights for a camera with macro lens. If it can do that, it can examine pliers jaws, punch points and other such things was similar ease.
The image shown here isn’t very good, and takes an inordinate amount of time to capture.
If I could take a better image and in seconds, that would be great.
Measurement features are likely baked into proprietary (read extremely expensive) software. I can do without that if it makes the difference between affordable and “yeah right” pricing.
Shown here is higher quality diamond abrasive particle that I imaged back in grad school.
Every couple of years, I look at metallurgical microscopes and check benchtop SEM pricing just in case the tech has become affordable for individuals (it hasn’t). I’ve looked at inspection scopes as well, but I haven’t had much need for one without good imaging capabilities.
One brand – Keyence – claims that their latest digital microscope is “delivering images that rival an SEM” thanks to their “optical shadow effect mode.” That might be true, but I’m sure their prices are well out of reach.
There doesn’t seem to be much between consumer models and true digital microscopes at “I’d better skip lunch for 10 years” prices.
Even at the $1000 and up price points, everything I’m seeing short of institutional research and industrial microscopes – but not inspection scopes – are designed around small and flat objects.
That brings me back to the big question. Are digital microscopes good yet?
Or rather, are the affordable ones not too bad?
I feel as though I might be missing something.
And yes, I know this is not a typical tool, but I’m hoping there are enough ToolGuyd readers in industrial or high precision fields who can point me in the right direction.
Plus, in looking into digital microscopes, I’ve seen a number of popular tool-related applications, such as inspecting wood and metal cutters for damage. It might make sense for me to start off with one while I decide whether or not to shelve the idea of a “real” microscope again.
I am eager for any pointers regarding digital microscope brands or tech!
At the higher end of the spectrum, I can think of dozens of ways to use a professional scope for ToolGuyd purposes, but that could just be wishful thinking. The scientist in me likes to be let out.
Before anyone says “just try extension tubes/macro lens/so-and-so phone attachments,” I have tried different techniques over the years and looked into others. There’s a broad spectrum of other things I could try, most of which are expensive, time-consuming, or both.
I’m looking for something I can just turn on and use.
Are digital microscopes good? Which ones? I’ve got ToolGuyd’s credit card at the ready.
Is there reader interest in extreme closeups of tools such as drill bits, screwdriver bits, pliers jaws, tool wear, abrasives, saw teeth, and so forth, on an ongoing basis?
Adam
The author of my favorite sharpening resource (which is now defunct, sadly) used an Intel QX3 Computer Microscope to inspect blades. Please have a look at the archived page here for detailed info on his usage and results:
https://web.archive.org/web/20190131124842/http://www3.telus.net/BrentBeach/Sharpen/qx3.html
Also, I highly recommend poking around the rest of the archived site, as it is truly the most straight forward and detailed info I’ve ever seen compiled on the subject of sharpening. I’m so sad the the original site is no longer active.
Blocky
That’s a great read. Thanks for sharing.
FEM
Ha ha. I have one of those QX3 scopes! We bought one for work over 23 years ago, when digital microscopes were way too expensive. The image sensor was 0.3MP. For us the results were mostly unusable.
ToolGuyDan
I’ve had good luck with a MOYSUWE MDM10 (https://www.amazon.com/dp/B0BV6H185S). It claims a laughable 1300X magnification, which requires accepting video output as a form of magnification, but otherwise I’ve been quite impressed, especially given the price-point. It has many compromises, to be sure, but none of them are deal-killers. The only thing I truly miss on it is additional raw magnification; one can easily see printing details or closely examine materials of any kind, but something like identifying bacteria—trivial with a $500 desktop compound ‘scope—is simply beyond its level of capability.
Still, pound-for-pound, I haven’t been disappointed in the least, and in fact carry mine with me (it conveniently packs a rechargeable battery, and the stand is removable) for authenticating certain rare objects “in the field”. I’ve probably “sold” three or four of them now to people in my industry who see me using mine and immediately ask where I got it.
Oh, and the software is almost decent, in the sense that it gets the hell out of your way. Images are saved to a removable card in a standard way, or the device can present itself as a normal camera to the host OS when it’s plugged in.
Tony
For my own needs I’ve found the best compromise to be a sub-$100 USB microscope and a large computer monitor (which I already have). I settled on this solution because I found the higher priced digital microscopes to be quite off-putting (both in terms of price and their overstated capabilities). I remember buying one of the several hundred dollar models and then immediately returning it after it didn’t provide the features claimed.
I wanted a digital microscope that works with the computer’s native Camera App (no additional software required), provides clear images, and I wanted something that can easily be removed from its stand to inspect non-flat objects. I ended up going with the Tomlov DM101 (which doesn’t seem to be available on Amazon currently). I paid $56 for it when it went on sale at Amazon a year ago.
This model is great for extreme close-ups of tools. For example, with the naked eye it appears that the lasertip etchings on my new Wera 162i PH1 Phillips screwdriver are very refined. However, when I put it under my digital microscope, the lasertip etchings are actually somewhat crude. They don’t look anything at all like the polished pictures you see on the Wera website. And yes, this is on an unused (brand new) tool. I’m not dinging Wera at all. They make great tools and the lasertip feature works great. It’s just that the lasertip etchings are not nearly as refined as the company would have you believe from their marketing materials.
Michael F
This is something I’m also interested in. I try to keep 40+ year old computers and CRT monitors running (like my Commodore 64) and frequently would like a digital microscope for inspection. I haven’t bought one for the same reasons you outlined in this post. One suggestion is that sometimes you can find used industrial grade equipment for a good deal on eBay. Since these expensive devices are usually high quality and used in a very controlled environment you can generally expect a long lifespan.
Stuart
The issue with used industrial equipment is that there are various levels of wear and tear, and the imaging tech is likely outdated. In academia, especially, money isn’t spent unless it’s absolutely necessary or when new grant money comes in.
Bart Youngblood
I guess it depends on what your expectations are.
I’ve been looking at one for my electronics work, particularly when I’m working on PCBs and have a hard time making out tiny markings on components, or trying to track down the solder joint gone intermittent from 40-50 years of heat cycling. Most of the ones in the $150 range seem to be decent enough for doing that kind of work.
Only reason why I haven’t pulled the trigger on one is I’m often working on (heavy) stereo amps or radios that I’m often having to reposition on their top/bottom/side. I can’t seem to find a good quality one that doesn’t have the display permanently mounted on it or can otherwise be mounted on a boom to be quickly repositioned.
MM
I don’t have any current knowledge on these, but when I was last doing this sort of work my go-tos were either:
a) used high-end brand stereomicroscopes aka “dissecting microscopes”, using a 3rd party adapter to mount a modern DSLR camera. At the time I was using Canon EOS 5D MkIIs, then getting video into a PC via the USB cable & the software which came with the camera.
or
b) a variety of video options from Edmund Optics.
Something else to consider is simply a good macro lens for a digital camera. I suspect you probably already have one or more good digital cameras, and it’s amazing what one can do with a macro lens.
My experience has been that the name-brand solutions are incredibly expensive. When I had my university job we had an Olympus BX60 microscope which was fitted with a film camera, no doubt state of the art when the microscope was first purchased. I was tasked with upgrading it to digital. Our Olympus dealer came back with a $17,000 quote for their digital camera and adapter, and despite the sky-high price the resolution of the digital camera was already well out of date and it couldn’t record video. I ended up purchasing the aforementioned EOS 5D MkII which was state-of-the-art at the time, plus a mounting adapter from Leeds Micro, for under $4k for the whole package.
Stuart
Back in 2019, Olympus quoted me $45-55K for an inversion stage metallurgical microscope, and then $10K (~$4K + $6K accessories and imaging) for a student-grade scope.
Metallurgical prep (mounting, polishing) has hefty fees, and I determined the costs weren’t worth the limited content or exploration potential.
I had been looking for an inspection scope at the time too, but nothing looked suitable.
There’s a lot more I can do with a scope that has greater working distance and depth of field.
What do bi-metal vs. carbide OMT blades look like after cutting steel screws? Drill bits after 10, 50, 100 operations? Screwdriver tips when turned to a certain torque? Different grades of router bit cutting edges? Sandpaper? How clean are the edges of holes drilled by Forstner vs. spade drill bits?
Some of the industry-grade microscopes look to provide mixed capabilities depending on how they’re equipped, and might be able to do this far quicker and easier than otherwise.
The difficulty with a photography macro lens is that the depth of field is very shallow. Add in ways to increase magnification, such as with a bellows and microscope-style objective lenses with a stack of compatibility adapters, and things get finnicky with large investments in equipment and time.
I’ve looked at this company’s motorized macro rail before – https://cognisys-inc.com/stackshot-macro-rail-package.html – and it’s relatively affordable (with respect to business-expense solutions). But, there’s no real-time capture as with a microscope, and a lot of post-processing is needed.
There are others, but they’re far pricier.
I’m thinking a consumer model – if decent enough – can get me up to speed and determine the boundaries of what I can expect from any system.
There are a lot of different components:
Optics
Camera/Imaging
Controls
Everything else
Optics are expensive no matter what. The same is true about camera and imaging equipment. Control packages are pricey – some brands look to have standalone or PC-connectable equipment, while others have custom 4K displays with the user interface hardware built in.
Quick results can mean all the difference.
The “everything else” could be an interesting project. Perhaps a DIY XY stage and motorized Z axis could perhaps leverage everything I’ve been learning about CNC motions and controls. Lighting is a challenge no matter what, but there are many modern advancements. If I need something customized, I can create a custom lighting bracket to hold a high CRI LED strip.
Focus stacking, however, would be a pain no matter what.
I’ll play around while I research options.
It also depends on what level of magnification could be useful, something I can play around with using digital cropping.
That’s where I think consumer vs industrial solutions split. Consumer products promise say 500X magnification because you can look at a 1/8″ part/feature on a 65″ TV via HDMI. Industrial solutions seem to go by the feature size on the sensor, although some make it clear that video magnification is used.
Wayne R.
The Olympus OM-1 system does in-camera focus stacking. With a good Laowa macro lens, a good flash (that’ll let you catch a fast series of focus stacking shots), and maybe not even a mount/tripod/stand, you might find it to be what you need, plus it’s a regular camera and does excellent video too.
https://www.youtube.com/watch?v=pc7y28vkb3Y
Stuart
I donated the last of my m4/3 equipment last year, and am not eager to buy any more camera bodies. But, I’ll explore all options before spending a penny.
MM
I am by no means an expert on digital cameras, but I think that there are many models which have automatic focus stacking these days so it’s certainly an option to consider.
Also, while it is true that macro lenses tend to have a shallow depth of field I feel that they are plenty good enough for the applications that Stuart mentioned. When I had my machining business I had an EF-S Canon DLSR with the 60mm macro lens & a ring flash–not exactly a high end rig–for closeup product photos, and to document wear on tooling. For jobs like showing the surface finish of a drilled hole, looking at the teeth of a saw blade or the point of a drill bit? Plenty good enough for that without focus stacking. Heck, even a basic DSLR with a half-decent lens and Macro Mode is good enough for that task. You don’t need metrology grade video microscopes to take closeup photos for a website.
Stuart
Need? No. “Need?” That’s different.
If I won the lottery, I’d set up a test lab with an electron microscope that I’d look at everything with. In a heartbeat, without hesitation.
That hasn’t happened yet. If I can scratch that itch a little bit in a different way, and there’s interest in what could come from it, I’m at least going to explore the option.
Images are easy. Good images are harder. Measurements are the hardest part. How much wear to a saw blade tooth, Phillips bit tip, etc?
I’ll be looking into this in the background. In the meantime, I also need an inexpensive magnifier or semi-decent affordable digital microscope for looking at tooling, chisel edges during sharpening, and similar.
David Z
What is unique about the OM-1 is that it isn’t just taking the pictures. The camera also processes them to give you the final image.
Bob
I have strong advice on digital microscopes: Get one! Don’t deprive yourself of one because you can’t find that perfect one. I can afford any one I want, but I can only justify a cheaper one. So I grabbed a cheapie and I’m glad I did.
The very cheap, lower quality ones (mine cost $30) are VERY useful. Mine has gotten me out of binds, solved mysteries, and satisfied my curiosity many times. I just plug it into my USB port and it uses my Windows webcam software automatically. 9 times out of 10 it does what I need. The sharpness, versatility, and ease of use pleases me every time I use it. So until you find that perfect model, get yourself a $20-$30 one. My cheapie made me lose interest in a higher quality one.
Ben
We use Dinolite at work. They have been great for the past 5-6 years. I’m sure the same models are ~$100 these days.
TonyT
My comments:
1) Keyence has been over to demo their digital microscopes, and they are VERY impressive. Based on the demo, I’d say they put a lot of time into the user interface and usability. At least some models support EDOF. Price is >$50K, which is too much for our occasional needs.
2) We’ve had a Dino-lite AM7915MZTL digital scope for >5 years which we got specifically because it has EDOF (based on software stacking with the camera varying focus slightly) support. Cost was ~$1200. It has worked pretty, including the EDOF feature. https://www.dinolite.us/products/usb-microscopes/usb-edge/am7915mztl/
3) For standard microscopes, don’t forget Meiji Techno. We bought a Meiji steroscopic microscope with camera attachment ~15 years ago; IIRC, price was ~$3K. https://meijitechno.com/
Stuart
Keyence kept coming up in Google sponsored search ads, and their tech does look impressive. They were the first to call in response to my price list/quote request. They gave me a rough ballpark of ~$25K as the starting point.
For some reason I thought Dino-Lite was focused on consumer products; I’ll take a look at them too. At the surface, it seems decent. Some of their models have measurement capabilities, although it requires a calibration scale.
Their depth of field examples don’t look too impressive, and requires Windows, but it can be hard to tell just from online images. Maybe that’s the difference between realistic expectations and marketing hype. Their lighting and support accessories look reasonably priced as well.
Thanks!
I’m hesitant about a stereoscopic microscope or similar, as at least some of the cost would go to features (such as dual eyepieces) I don’t plan to use extensively.
I’m not familiar with Meiji Techno, but their equipment does look to be reachable.
I have a lot more research to do.
TonyT
As far as Keyence pricing goes, we were probably looking at a higher end microscope – had X, Y motion and such. Keyence is an interesting company; they’re basically a sensors company, but sell a wide range from $100 fiber optics to $100,000 3D measurement systems. They sell direct, have a lot of sales staff turn over, hide the interesting info behind a login (and call you every time you download or look at anything), and do have some really neat products.
I’ve always considered Dinolite more industry oriented than consumer. They used to do a lot of tradeshows; not sure about post-COVID. They’re definitely not the cheapest compared to a random Amazon USB microscope, but the extra cost could very well be worth it. The Dinolite EDOF worked well enough for my needs at the time (looking at HDD heads), but might not be a good fit for you.
I’d recommend contacting Dinolite with your questions; also you might see if they have anyone local (we bought from a local company). The Dinolite.us website does offer 30 day return or exchange.
Jared
I don’t have any input on your purchase, just here to say as a reader: yes, I want to see those things.
E.g.:
What do bi-metal vs. carbide OMT blades look like after cutting steel screws? Drill bits after 10, 50, 100 operations? Screwdriver tips when turned to a certain torque? Different grades of router bit cutting edges? Sandpaper? How clean are the edges of holes drilled by Forstner vs. spade drill bits?
Stuart
The challenge is in time and quality.
I can do a lot of that now, it’s just a matter of how long it would take to produce appealing and useful images.
Steve
I can’t help in selecting one, but as a reader I like the direction you want to go with this and I think it will be a great value add to your articles once you can use one to look for wear.
Andrew
I’ve used Meiji, Olympus, Leica, Nikon, Keyence, etc with budgets up to >$100k, including automated optical inspection tools.
I’ve used e.g. dinolite and other low budget versions at the other end.
I’ve used both categories for development, and for 24/7 production.
For your kind of use, I expect you’ll get far more bang for your buck at the lower end. A high magnification objective can cost in the $10k+ range for e.g. a Nikon compound microscope.
There’s two main architectures with two different costs:
1. Cheap moulded lenses plus low cost CMOS sensors
2. Custom multi element optics, mechanics, automation and software.
There’s just not much in the middle because there’s no real market there.
MandM
Maybe look at astronomy cameras? I’d guess that’s a much bigger market, and the sensors have advanced dramatically over the years.
I’ve got no idea on the implementation, and I’ve never used one, but I know that folks can get some incredible astronomy images with cameras that cost around $500.
https://astronomy-imaging-camera.com/
Nathan
See if some of them have a trial period. Might be what you need. Which one has his quality camera sensor but no frills. Manual lens selection. Manual focus or auto focus with manual adjust. Etc.
I assume here it would run through a computer and capture setup like Photoshop. Would you want video?
Stuart
That’s the difficult part – the “frills” are what make the difference.
A motorized z axis allows for autofocusing, which I can take or leave, but also automatic focus stacking. Software packages then create a composite image from a stack of photos.
Frills can make the difference between clicking a few buttons and getting a usable image or handling a camera for two hours.
For the depth I need, we’re talking about minutes per each composite image. There’s one scope that can create images in real-time using cutting edge tech, but… it starts at $40K.
There are a couple of more affordable options, and they’re all vastly different from each other.
I am also looking at ways to leverage my existing camera with greater time efficiency.
TonyT
Maybe you should actually lay out your non-negotiable requirements and negotiable wishes. How much depth of field do you need at what feature size/magnification? and so on.
To give you an idea of what you can do without focus stacking, another approach is to use a lens with a very high f/number, and absolutely blast the light. We used that approach on the HDD head project; we used the Dinolite for initial research, but the production machine used a VST VS-TC8-110 lens with 8x magnification (meaning a 1mm circle will be a 8mm circle on the sensor), f/89 aperture, and 100 micron depth of field, which is very good for 8x (and we needed all of it, and more)
Stuart
I started working backwards, determining what capabilities I can afford, and then working from there.
Feature size would be 1/8″ to 1/2″ on average. This can be done with a wide range of equipment, with the limitation being speed and ease.
This is the dream – https://www.zeiss.com/metrology/products/systems/industrial-microscopy/visioner1.html . It can deliver high depth of field in real-time, but at huge costs.
Less expensive systems look to be able to cover the same range in minutes. With consumer camera equipment, large image stacks can take hours for setup and processing, which I’d like to do but can’t.
I’ve talked briefly with Keyence, Vision Engineering, and Zeiss. I’m waiting to hear from Olympus and Leica. Olympus emailed me at least, apparently Leica is too good to provide quotes to small businesses. Theirs (Emspira 3) looked a bit simplistic, which suggested to me affordable.
I’m currently leaning towards going handheld (thank you for the Dino-Lite recommendation!) and with a camera-based macro setup that can buy me time until I can determine exactly what I need and can afford.
One benefit of this process is seeing how each brand’s options are unique.
Some have tilting camera heads, not to different from how miter saw bevel adjustments work, which might work with a Dino-Lite or similar. I just have to figure out how to design something like that, or rather a good brake to ensure it doesn’t loosen up and flop over.
TonyT
More feedback:
1) Check if your current digital camera can do focus stacking. I didn’t realize it until this discussion, but my camera (Panasonic GH5) can do semi-automated focus stacking (it takes a burst of pictures at slightly different focus points, then you select which pictures or a range of pictures to merge, and then it merges them on camera).
2) Definitely look into the Emspira – maybe try to find a dealer if Leica isn’t responding. I did a quick search, and it looks like the prices starts ~$10K.
3) The Zeiss Visioneer is truly cool, way more real time than the Keyence demo I saw. Note, however, that the resolution decreases as the DoF increases, e.g. at 2.8×2.1mm FoV (field of view), resolution is 128 Lp/mm and DoF is 1.8mm, at 5.4×4.0mm FoV it’s 64Lp/mm and 6.4mm DoF, and at 20.1×15.1mm FoV, it’s only 16Lp/mm but a whopping 69mm DoF.
4) It’s possible that a small aperture lens system could give the DoF you need, but the problem is that I suspect you’d have to build your own system, which isn’t worth the time. To give you an idea, the Moritex Schott ML-Z0315D zoom lens has a DoF of 2.4mm at 0.6x, and 1.5mm at 0.8x, and 0.8mm at 1.2x. (FoV will depend on the sensor size, for a 1/2″ sensor should be 8.0×10.7mm at 0.6x, 6.0×8.0mm at 0.8x, and 4.0×5.3mm at 1.2x).
Stuart
Thank you, I appreciate it!!
I still haven’t heard a peep from Leica, which is a big turn-off. If a company or department is this unresponsive to pre-sale requests and inquiries, how might they treat me in case of a post-sales issue or problem?
My current camera can stitch but not stack images. Additionally, I would usually use manual and adapted lenses.
An automated rail might do the trick, such as Stackshot or Wemacro. Zerene Stacker (or Helicon Focus) could help with the image stacking workflow.
I wasn’t aware about the Visioner’s loss of resolution at higher magnification, which makes me feel better about it being so far outside my price range.
I’ve been looking at pseudo DIY systems which at the least could help me decide what might work should I budget for a higher-end scope for analysis purposes.
The world of machine vision lenses is completely foreign to me, but I’ve started to pick up the basics.
It might also make sense for me to use traditional macro lens and potentially (hopefully) speedy automated image stacking flow, and then adapt the same to a setup with 5X microscope objective lined set up with a tube lens and adapters.
Either optical assembly could potentially allow me to move to a vertical focus rail setup with tilting control. I’m accustomed to aiming a camera and then positioning the work (such as a screwdriver tip). Positioning the work and pivoting the camera could help simplify things and allow for repeatability. It’s something I hadn’t considered before, but is apparently common in higher end digital microscopes to compensate for having a flatter image compared to stereo inspection scopes. If it’s done with the rotational axis across the specimen plate, that can allow for much easier and repeatable examination without refocusing.
The question there, if I go the DIY route for now, is how to built a tilting head (similar to miter saw bevel adjustment) that’s rigid, repeatable, and easy to adjust.
Is there such a think where a rotational pivot arm can be counterweighted and dampened, similar to how videography fluid heads work?
A geared mechanism can bear a lot of weight, but wouldn’t allow for free and smooth adjustment, such as to quickly adjust from 0° to 30-45° by hand.
The undoubtedly pricey Olympus has a tilting head. https://www.olympus-ims.com/en/microscope/dsx1000/tilt-model/ How might one emulate that functionality with off-the-shelf parts?
The Olympus looks to have an pivot lock on the side, and a thick shaft or bearing for the head to pivot around.
A Keyence can do something similar – https://youtu.be/C4IBBj20OpE?t=34 .
A premium Leica has a spring-release lever – https://youtu.be/Ir9qOpEXlTk?t=14 .
A different Keyence has a cranking lever – https://youtu.be/pX9f57yHyXo?t=9 .
Or, the Taig micro mill has a rotating z-axis shaft. But, there’s little friction between its metal contact surfaces unless the locking nut is fully tightened. There could be a way to adapt that design but with added friction to help prevent a fully loaded camera-equipped z-axis from crashing down.
Sorry, I’m thinking aloud in case you or anyone else has ideas on how this can be done.
I have seen one example of a tilting stage, but that doesn’t seem as advantageous except for focus rail rigidity.
TonyT
It’s good to think out loud – otherwise no one else can provide any feedback. I do think you eventually decide if you want to make this a project, or spend more to get something that “just works”.
I think using low cost approaches first to get a better idea of what works is good. Also, you might consider using several different tools instead of having one do-it-all tool.
To be accurate, the Visioneer’s resolution decreases of the FoV increases, which makes sense (much bigger area for the same optics), but my feeling was that something has to give for the extra DoF, and that appears to be resolution. The ML-Z0315D lens I references is an old lens that’s nothing special, but has almost twice the resolution as the Zeiss at roughly similar settings (assuming I did my math right): 122 lp/mm vs 64 lp/mm for the Zeiss at a 4mm x ~5.3mm FoV.
BTW, if you’re interesting in measuring, you should investigate telecentric lenses – they are very commonly used for measuring, and this page from Olympus explains why: https://www.olympus-ims.com/en/insight/telecentric-optics-101-all-the-basics-you-need-to-know/
I’m not a mechanical designer, but here are a few places that might give you some inspiration:
Affordable Misumi manual rotary stage (we’ve used it, not meant for quick adjustment, but might give inspiration): https://us.misumi-ec.com/vona2/detail/110300198470/
igus slewing ring systems https://www.igus.com/info/slewing-ring-overview
McMaster rotary tables
https://www.mcmaster.com/products/rotary-tables/
Stuart
@TonyT
Thanks, much appreciated! I figure that if I’ll post about this separately if I continue to run into the same walls.
I suppose there’s always a tradeoff, even with new tech.
Right now I’m looking at ways of using infinity corrected objectives with a standard digital camera. Well, that’s phase 2. Phase 1 is seeing how well a motorized platform works for focus stacking.
Phase 3 might involve building an optics assembly with telecentric lense and c-mount camera. Or maybe a microscope with full annotation capabilities. At that point I might have my needs strictly defined and expectations fully matched with realistic capabilities.
A slewing ring might be close to what I’m looking for, if it can handle loads. But, I’m not seeing applications close to what I’d potentially want.
Rotary stages and tables are really designed for positioning. They can handle weight, but I haven’t seen any designed for large off-axis loads.
Face-mounted crossed roller bearing kind of look like they could work for this. But all of this is also secondary. I can rework a z-axis from fixed to rotational/tiltable.
If the optics path is consistent, I can always use a calibration scale for manual bitmap measurements, regardless of lens.
TonyT
I do think having a motorized system makes it much more useful, the more motorized axes the better 🙂 One thing expensive microscopes give you is nice motorized stages, up to piezo stages with linear encoders.
I wonder if you can combine some learning between this and the CNC. Yes, you’ll want to use different components, but learning about motion control for one application should be helpful in the other.
Some rotary stages are better than others for off axis loads. We’ve made a few systems with significant off axis loads (for example >5 lbs ~6″ in front of the vertically mounted stage). Bell Everman’s ServoBelt rotary stages have pretty high load limits. https://www.bell-everman.com/products/rotary-positioning/servobelt-rotary-stage
We also did an experimental system basically using a Harmonic Drive gear as the stage (with basically a 12″ wide 8 lb dumbell-shaped load mounted parallel to the face of the vertically mounted HD gear).
I did a couple quick searches, and you can get used Servo Belt stages with Teknic motors for <$300, and HD gears also for <$300. I would not recommend buying a used worm gear stage, because worm gears definitely do wear out.
Final note: Prior Scientific https://www.prior.com/ is famous for microscope automation, and you might get some good ideas from browsing their web site (and seeing what's available used). (Their approach doesn't fit mine, but then again, I'm in a different market).
Stuart
Thanks, I’ll look into those as well!!
Another Keyence sales rep called me today – “I see you’re looking at XYZ on our website.”
I had looked around at the machine vision options for ideas, and apparently that was enough to trigger another call by a second salesperson.
You said:
You weren’t kidding!!
TonyT
Yeah, Keyence sales. My boss loves Keyence and doesn’t mind being called by them, but I don’t like bothered, so last time we had a scheduled Keyence demo, I downloaded a whole bunch of interesting stuff (on machine vision, 3D sensors, and such) the day before 🙂
Stuart
It reminds of a time when I requested a bunch of Corning samples – centrifuge tubes, culture flasks – plastic sample holders kinds of things.
The local rep practically chased me down as I left to go from my office to a lab room.
I really like it when a sales rep is available to talk or field questions. But there’s a limit to that.
However, I found both calls from Keyence – the first before 9am, and today’s after I simply revisited their website, to unbalance me a bit.
ToolGuyDan
Have you reached out to any local universities? Land grant schools, in particular, often have an entire staff dedicated to agricultural outreach; if you can spin a project as being ag-related, or click with some other institutional priority (e.g. if you can pay a modest fee for ‘scope time on their SEM, and that fee is reduced because the operators are trainees, or kids from an underserved population, etc.), you may find great success. Personally, I’d track down a materials scientist who’s also a tool nerd, and get them to write a grant proposal to buy a fancy ‘scope for evaluating wear modalities in real-world usage with an eye towards aerospace applications—a screwdriver in space needs to be tough, yes, but it also needs not to shed any debris as it wears, and that requirement gets more intense the longer it’s in space (think of something like a Mars mission, where it’s much harder to predict how often the tool will be needed, and nearly impossible to do a resupply of fresh tips).
KMR
A bit about ToolGuyd:
https://toolguyd.com/so-i-finally-earned-my-phd/
ToolGuyDan
I took a look, and sure enough—there was Fred in the comments. I think he might be immortal. 🤣
He asks a good question, though: when do we get to see your thesis (and/or the patents that arose therefrom)? Also, you mentioned in the post that you might write a short book; did you?
Stuart
I don’t remember? I was doing a bunch of freelance stuff at the time. Most were published, some fell through.
I can send you a link or copy via email if you’re interested, but it’s not at all tool-related.
“As far as long-term plans go, meaning a few years from now, I intend to build and acquire testing equipment. Wouldn’t it be neat to compare the crystalline structure of cheap vs. expensive drill bits? Broken wrenches? Sand paper? Cordless drill motor windings? Circular saw carbide teeth?”
I guess we’re now at the “few years from now” point.
Stuart
It’s been nearly 11 years now, and I have yet to pay an hourly rate to sit behind an SEM. It seems time to figure out what I need or can access without telling myself I’ll find a way to get scope time “later.” An SEM is also destructive, requiring sample piece removal and prep.
mark999
If you’re willing to 3d print, source components, and assemble, check out the PUMA microscope. Designs are on GitHub, which also has links to in depth YouTube videos and to the author’s site. The latter has preassembled scopes and components.
I’ve yet to assemble one but it sounds quite capable; IIRC the author claims it exceeds any commercial scope available in some ways.
https://github.com/TadPath/PUMA
FA Guy
Scope
I run failure analysis labs. Optical, SEM, TEM, and FIB are all daily workhorses in my career.
Keyence are really nice scopes. Have had a couple of $100k automatic everything scopes and they’re amazing. Their salespeople are aggressive to try to get you to allow them to do a demo for you. Believe part of their compensation is based on how many demos they give so even if you say you’re not buying anything they’ll still come and be nice. Their demos are amazing.
What you’re trying to do isn’t a demanding application. You can get a great setup without having to spend a lot. Get a used trinocular stereo zoom microscope. These things don’t break. At most they’ll need a cleaning which is easy. Do not buy new…it’s never worth it for this type of scope. Olympus, Nikon, Leica, Zeiss are all good brands. Lots of reputable used lab equipment businesses deal with this. Get one on a solid base and not on one of those fancy articulating arms that you can move it in any direction as those aren’t stable and image will come out blurry. Get a boom type arm if you need that extra depth. Trinoculars have a c-mount which allow you to place a digital camera on while still being able to use both eyepieces. Get one which goes to ~40x with an adjustable knob, not fixed lenses.
To get that large depth of field you need an adjustable aperture. Low, and even pricier, digital scopes do not do this. It’s the lens system that matters. Closing the aperture is what gives you that large depth of field. Less light will come in though. It’s the same principle as using an SLR camera.
This eBay listing is perfect. Just need to add camera and light.
https://www.ebay.com/itm/204037506893
Amscope makes great cameras that are priced reasonably. 9 megapixel is plenty. MU900 is a perfect camera. Use these in the lab and they work great. You don’t need any fancier ones or low light ones. Software that comes with MU900 will take measurements but it gets a little tricky with stereo zoom scopes with the adjustable knob for magnification as you can’t calibrate the software to it. You’d need fixed lenses if you want to make measurements super easily. One trick is to take a picture of your sample, then keeping the magnification the same take a pic of calipers opened to a known dimension and then use an image program to count how many pixels is the width of the caliper opening and now you have a conversion for pixels/mm for your sample. Paint.net is a free program which can do this. It may be possible to do a calibration each time with MU900 software but not sure.
For light, get an led ring light. Ideally one with multiple zones that you can turn on/off as it can help with reducing glare. Amscope LED-144A is what we use and it’s been great. Fiber optic ring lights are expensive and really aren’t any better. I’d stay away from fluorescent ring lights as well.