About Me

My Photo

I previously worked on Virtual Reality and other hardware at Valve.  I currently work at Google[x].

Prior to starting at Valve, I built computer peripherals such as keyboards, mice, and joysticks that were designed to be used inside MRI machines.  My company, Mag Design and Engineering, sold these devices directly to researchers at academic institutions who used them to publish scientific papers in peer-reviewed journals.

After work, I spend time on many different types of projects that usually involve circuit design, machining, material selection, and general fabrication/hacking.  My favorite place to be is my home workshop.

ben dot krasnow at gmail

http://www.youtube.com/user/bkraz333

http://www.linkedin.com/pub/ben-krasnow/4/6a9/679

http://www.twitter.com/BenKrasnow

Monday, March 21, 2011

DIY scanning electron microscope - Overview video


Today, I finally produced an image with my DIY scanning electron microscope. I've spent the last few months working on this project, and am encouraged by today's success. There is still a lot of work left to do in making the image higher resolution, and eliminating sources of noise, however this image proves that all parts of the microscope are operating as designed.




147 comments:

  1. Do you have any measure on the x-ray output of this thing? You might want to put some lead around the bottom of the bell jar to help keep you safe.

    Awesome project. I didn't think anything could top your LN2 setup, but this does.

    ReplyDelete
  2. AdShea, you are right, decelerating electrons will cause X-rays to be generated. However, my accelerating voltage is 6KV, so the resulting X-ray wavelength is very long and will not penetrate matter very far (like a couple micrometers). Even the aluminum and glass of the chamber will contain all traces of the X-rays. This is why older black-and-white TV sets (and oscilloscopes) did not need thick leaded glass -- their acceleration voltage was under 10KV, and only produced soft X-rays. Color TVs needed higher acceleration voltages 25-35KV, and would produce penetrating X-rays, thus necessitating the thick leaded glass shield.

    ReplyDelete
  3. Wow! That is one impressive project! What do you recon is your current resolution, and how far do you think you might improve on that?

    Hat off to you sir!

    ReplyDelete
  4. Fascinating project, and something I've wanted to build for years. Can't wait for more details - especially of the electron gun (is that copper tube at earth? Looks like it).

    One question I'd love to know the answer to first: Why do you need to convert the secondary emissions into light? Can't you detect them directly by, say, a plate connected to a high impedance amplifier?

    ReplyDelete
  5. Really incredible... great work.

    ReplyDelete
  6. Love the project. Any more information about how it works and how it was build would be great.

    ReplyDelete
  7. Hey Ben, you just inspired a young materials scientist whose main thesis is in characterization (currently preparing TEM samples). Please build a HR-TEM next!

    ReplyDelete
  8. Very interesting DIY and an exceptional blog overall!
    Keep up the awesome work Ben
    (I linked this project to my blog - http://sciartmag.blogspot.com)

    ReplyDelete
  9. Just found the site off of Metafilter. You have definitely got a new follower in me. Cudos on an amazing project. I can't wait to see more.

    ReplyDelete
  10. > Why do you need to convert the secondary
    > emissions into light?

    Rupert, this question bugged me from the first day that I started to plan the SEM project. I searched through many descriptions, patent filings, and academic journals, and never found a comprehensive discussion. I believe the short answer is that the Everhart-Thornley detector has a dynamic range many orders of magnitude larger than a high-impedance electronic amplifier. In addition, the ET detector has a much lower noise floor than any comparable electronic amplifier. The nature of the highly-biased phosphor screen allows even single electrons to create photons, and those photons can be counted by the photomultiplier tube. Building an electronic amplifier that can sense single (or a few dozen) electrons would be very difficult indeed. It does seem like a very Rube Goldberg solution, but all modern SEMs use ET detectors, so they are still considered state of the art.

    Yes, all of the copper pipe and support structure is at 0V -- earth ground. The electron gun is biased at -5000V.

    ReplyDelete
  11. Zigurana, the lock washer is only being magnified 10 or 20x. In this test video, I wasn't too concerned with focus, so the resolution is probably 50um or more. Eventually, I hope to achieve resolution of about 1um. I will probably experiment with different objective lens placement, final aperture size (currently 100um), deflection plate placement, etc. I was so excited to see an actual image, I had to make a video. I'll spend the next month or two trying to get actual decent images out of the microscope.

    ReplyDelete
  12. Ben, you are extremely gifted at relating technical concepts in relatively simple terms. Congrats on your success!

    ReplyDelete
  13. Your project is fucking awesome! Good luck with it!

    ReplyDelete
  14. http://www.reddit.com/r/science/comments/g8igo/real_garage_science_man_builds_scanning_electron/

    ReplyDelete
  15. Ben -

    Truly excellent project. Your video explaining the physics of the SEM is at least as good as or better than most of the SEM instructional material that's out there and certainly more accessible. Have you considered adding an EDS detector?

    ReplyDelete
  16. I enjoyed your video and look forward to future posts.

    ReplyDelete
  17. Yossarian, thanks! The EDS detector would be difficult to home-brew. I've seen EDS parts on eBay, but using one is not nearly as fun as creating cool SEM images with a secondary detector.

    ReplyDelete
  18. Hi Ben,

    This is amazing, what do you that you can do this at home?!

    I'd be interested in doing a USB board and software so you can get the image into a PC, rather than as a Oscilloscope image, if this would be of interest to you?

    Thanks

    Steven

    ReplyDelete
  19. Keep these videos coming! This is fascinating stuff.

    ReplyDelete
  20. Ben, Now you've got me wanting to build one too... When you going to build a mass spectrometer? ;-)

    ReplyDelete
  21. That's really neat, kinda like two of the most inherently cool elements in the whole universe, radium and mercury. And just like those two elements, without due caution, your machine will silently fuck you and anyone else around it totally up. You really need to start this whole conversation with "don't try this at home unless you're familiar with the dangers of working with x-ray producing equipment."

    You might want to look into regulations governing the operation of high energy devices capable of producing ionizing radiation. I'm sure there's someone in your state government who is more than willing to help you out...

    ReplyDelete
  22. As others have said, really, really cool!

    Just wondering, what are the spark plugs for?

    ReplyDelete
  23. Foxs, this was my first plan. I originally wanted to use a microprocessor to generate the raster scan pattern and capture the data for storage/display on an LCD. I may still do this, but there are a few problems: finding a microprocessor with enough RAM to store the image matrix, using a DAC to generate the raster scan pattern at live video rates, and interfacing with a nice LCD (ie generate SVGA video). The Parallax Propeller seems like it might work, but needs a lot of development. Have you handled live video with a microprocessor or USB device?

    ReplyDelete
  24. Alan, the spark plugs are very inexpensive and effective high-voltage feed-throughs or pass-throughs. They insulate all of the high voltage lines running into the vacuum chamber, and prevent air from leaking into the chamber and spoiling the vacuum.

    ReplyDelete
  25. Anonymous, if you knew a little more about physics, you would know that electrons with energies below 10KeV produce "soft" X-rays when decelerated, which will not penetrate glass or aluminum that is more than a few micrometers thick. It is because of this that performing EDS with electron microscopes requires special detectors in the chamber, often cooled with liquid nitrogen.

    If you were in the habit of reading comments, you would see that I have already discussed X-ray generation at the very top of the page in response to someone with a similar "concern."

    ReplyDelete
  26. Great work, no doubt.
    I have read some of the remarks especially regarding the concerns related to radiation. I think you have already studied this carefully.
    Take care!

    ReplyDelete
  27. Oh, PLEASE be explicit and detailed. I was considering dropping $25K to buy a used SEM "just for fun." This would be WAY more cool.

    ReplyDelete
  28. Wow! Amazing project! Good luck!

    ReplyDelete
  29. Absolutely amazing... great work... I've always wondered how difficult this would be. I'll be following with great interest! Cheers

    ReplyDelete
  30. Excellent work!
    I second the request for detailed explanations (as time permits). Better still if you eventually release open-source hardware documents so others can repeat, and possibly improve, what you've achieved!

    I have three CRT monitors I've been hanging onto in the hopes of doing some form of microscopy.

    Here are two Scanning Tunneling Microscopy projects (no affiliation):
    www.geocities.com/spm_stm/Project.html
    sxm4.uni-muenster.de/stm-en/

    And the Gnome X Scanning Microscopy suite ("The complete open source SPM control soltuion"):
    gxsm.sourceforge.net/index.php

    ReplyDelete
  31. This is incredible! Looking forward to more videos. Also, thanks for the detailed explanation on how SEM imaging works, never quite got it before.

    ReplyDelete
  32. You really need to get a shatter shield on the bell jar. As the friend who gave me mine told me, "Plosions are bad, dosn't matter if they are in or ex, plosions are bad." A professor at Portland State told me once he saw glass put through a car door from a ploding bell jar.

    ReplyDelete
  33. Ben,

    Great stuff! I have a few things I'd like to look at under a microscope of this magnitude. What was the total cost of the build?

    ReplyDelete
  34. Fantastic stuff!

    Regarding digital control: I think an inexpensive (say $100-150) FPGA development board would serve you well. You can easily find them with VGA and/or DVI out along with enough memory to buffer an image, plus USB for transferring stills. An audio DAC might be sufficient to drive a deflection amplifier. If not you can easily drive a separate DAC.

    Email me if you'd like recommendations for specific boards.

    ReplyDelete
  35. Great, inspiring project, Congratulations and many thanks for sharing. I'm looking forward to any news.

    ReplyDelete
  36. Ben,

    Excellent project! Regarding digitization, have you considered using just an MCU to digitize a frame, rather than provide a live feed? You could sweep the beam at any speed and then upload the information and reassemble the image in the computer. This would allow you to get a very detailed image (with some >16bit DACs to control your X/Y). You could use your scope for centering and then connect the MCU to do your digitizing.

    ReplyDelete
  37. Most excellent. As a young physics student I love seeing physics used and explained by someone who obviously has a love for learning and discovery. Thanks for the inspiration. I'm awaiting future videos.

    ReplyDelete
  38. Very interesting build! How many of the parts are off the shelf and how many are homebrew?

    ReplyDelete
  39. Wow, awesome, awesome project. Please keep the videos coming! I'd love to learn more about it.

    Are you going to add stigmation correction to it?

    Also, just to make sure I understand your explanation of the spark plugs -- they act as insulators until the voltage across them is sufficient to ionize air in the spark gaps, at which point they conduct? So, they basically prevent any voltage below a certain threshold from reaching your electron emitter and focusing hardware?

    ReplyDelete
  40. Outstanding job Ben -- and no magnets in sight! All E-field, utterly cool!

    One of my forum members linked here, and it's way worth seeing, thanks.

    You'd be welcome over there -- I'm fairly transparently trolling for talent -- people who really do things rather than just talk endlessly. There aren't enough of us!



    Doug Coulter

    ReplyDelete
  41. Awesome project, can't wait for the other videos :)

    ReplyDelete
  42. Your video is crystal clear, what kind of video camera do you use? Very very cool project, look forward to following. Thank you for posting...

    ReplyDelete
  43. Rando, my camera is a Lumix GH1 with the stock 14-140 lens. It's capable of truly great-looking video, but I've always been in too much of a hurry to setup proper lighting, make a neutral background, frame the shot, etc. I try to remember to turn off autofocus, set the white balance manually, set a manual exposure, etc. For tripod shots with little or no panning, it's a lot easier. Thanks

    ReplyDelete
  44. Doug, I ended up using magnets, but they were permanent magnets! Initially, I had a lot of problems aligning the beam with the column. Since my column does not have adjustment screws to precisely move the electron gun, I had no easy way to direct the beam. One day, I had a random thought to bring a small permanent magnet near the bell jar while the beam was on, and found it very easy to move the magnet around the outside of the jar, then tape it into place when I had aligned the beam. I marked the bell jar location on the base plate, so that the magnets always return the same spots after I reassemble the chamber.

    ReplyDelete
  45. Ben,

    Absolutely spectacular, really...

    ReplyDelete
  46. Richard, my ultimate goal was only to achieve low magnification for specimens like insects, pollen, small man-made mechanical items, etc. Astigmatism may not be much of a problem at those levels of magnification, but it could, in which case I would try adding an electrode between the X and Y deflection plates. Alternatively, I might be able to correct beam astigmatism by adjusting the common mode voltage on either the X or Y plates.

    The spark plugs are just simple feed-throughs. I cut away the sparkplug "ground" electrode and just soldered a wire onto the high voltage electrode. It's just a zero-ohm entrance to the vacuum chamber

    ReplyDelete
  47. Holy crap. I did not think this kind of thing could possibly be DIY.

    ReplyDelete
  48. Ben,
    I'm really interested in trying this myself as my senior project for my undergraduate BS degree. I wanted to build an SEM, but it seemed very difficult, however your project creates great hope in me that it can be done.
    Do you think it would be possible to build a similar device in about a year? I have some more technical questions too, as you mentioned the possibility of some sort of documentation. Perhaps I can get in touch with you?

    Thanks for the post, it's really outstanding.

    -Ethan

    ReplyDelete
  49. Amazing work! Really an inspiration.

    ReplyDelete
  50. HI Ben,
    Fascinating stuff! I like your narration in the video too - very easy to understand.
    Hey I was wondering whether you'd like to join us on the next episode of the Boys of Tech podcast as our guest? I haven't been able to find an email address for you, so I'd appreciate it if you could contact me using our Web contact form here: http://boysoftech.com/contact
    I look forward to hearing from you.
    Regards,
    Edwin.

    ReplyDelete
  51. Ethan, it certainly is possible to build a SEM by yourself in a moderately equipped workshop (as I have shown.) Start by assembling your vacuum chamber with mechanical pump and high-vacuum pump (diffusion, turbomolecular, ion, etc). Get a high-vacuum gauge so that you can be sure your pump is working, and test your chamber for leaks. Even though I used a glass bell jar, I think in retrospect, it would be much easier to use a metal chamber. Feel free to ask me questions: ben at magconcept period com.

    ReplyDelete
  52. macona, I welded up a metal cage for the bell jar, but I don't use it very often. I think the risk of the jar spontaneously imploding is pretty low. If something fell on it, that would obviously be a problem.

    Mad, I bought all of the high voltage supplies. Building them would be time consuming, and would take away from the time I could spend on the electron column, which is really the interesting part of this project. The column itself is completely machined from raw materials, except for the pre-formed tungsten filament, which I bought on eBay.

    ReplyDelete
  53. plexxer, yeah that sounds like a good plan. I don't have much experience with high performance microcontrollers, though. I've used Arduinos, Arduino Megas, Parallax propeller, Atmel AVR chips, etc. None of these can handle a 512x512 8bpp image, as far as I know. Do you have any experience or schematics for image handling with a microcontroller?

    ReplyDelete
  54. Hi if you want a simple way to display your image on a PC you could generate a composite TV signal by setting your horizontal scan signal to 15khz and vertical scan to 60hz adding them together with an adder op amp circuit and feed that into a cheap composite USB TV adaptor. Once the PC has the live video stream you could use whatever tools you want to manipulate the image. Do a search on composite video generation.

    ReplyDelete
  55. Bravo, Ben!

    I am myself trying to make a SEM. I almost completed the Vacuum system (vane pump, oil diffusion pump, jauges and valves), and I built a 10 kV HV power supply.

    I also collected various components (photomultiplier and scintillator to make an E.T. detector, electronics, bellows, etc...). I learned to pierce thin gold leaves by electro-erosion, in order to make cleans diaphragms of any size.

    The mechanical architecture will use threaded rods like yours, but outside vacuum. The column will consist of stainless steel bellows segments, normally used in vacuum valves, separated by Teflon plates with Viton joints. I hope to escape the very tedious obligation to have to make a full return to atmospheric pressure for any setting...

    I plan to use a LabJack USB instrumentation interface (http://labjack.com/u3) for the scan ramps generation, an to retrieve the useful signal in numerical form. The image frames will be stored in the control computer, but not at video rates!

    I will use a permanent magnet focusing, according to a patent used in the LVEM5 SEM (http://www.cordouan-tech.com/). I could send you the original scientific publication with plans for the required magnetic assembly (you will need a lathe)

    So well done, you encourage me to continue my project, which dragged for over two years!

    ReplyDelete
  56. Ben,

    By not doing it in realtime and offloading the image generation to the computer, you can use a simple MCU to control the beam, digitize the output and send the data via USB to the computer. You could probably get away with using a Teensy (http://www.pjrc.com/teensy/) or an mbed (http://mbed.org). Either one has a USB interface and enough sample code to get you going. You could probably get by with using the PWM output (coupled with an RC filter) at first to generate your X/Y voltages, but I would probably move towards using a high resolution DAC afterwards.

    I've never done a project like that, but I don't think it would be too difficult to do. If your interested in more details, send me an email - plexxer (at) gmail.com.

    Joel

    ReplyDelete
  57. I have been slowly working towards doing something similar. Have the diffusion pump.
    One online book I have found useful for understanding charged particle optics and acceleration is the online book
    Principles of Charged Particle Acceleration at
    http://www.fieldp.com/cpa.html

    Not light reading, but I have found it a good source on the theory.

    ReplyDelete
  58. Love this!!! Now I want to build one.

    ReplyDelete
  59. BB... Brilliant Ben. Use your gift of knowledge for good.

    ReplyDelete
  60. Awesome! I'm speechless. I use SEM's all the time but until you've seen something built I don't think you can really understand it. More details please.

    ReplyDelete
  61. I wouldn't even bother with microcontrollers. FPGAs are the way to go if you want to do high performance video with tight timings. This is a nice mixed-signal FPGA for only $99. All you'd need to add would be an external memory for the framebuffer.

    ReplyDelete
  62. that is awesome.. ur an inspiration!

    ReplyDelete
  63. How long did this take you to build? Also, do you think there are logical break points such that it could be broken down into a series of smaller projects?

    My supervisor and I have been considering proposing building one of these as a upper year undergraduate engineering project(or a series of such projects).

    ReplyDelete
  64. Just saw your video on wimp.com, I was highly impressed. You said "optically it works out better if the electron gun is farther away" in reference to your lensing. I am not sure this is a technically correct statement, in that this is not an optic lens. But the fact that I commented on it makes me a nitpicking asshole, so I am sorry.

    ReplyDelete
  65. I hope you know that this is on the EM/OLM list server!

    Great video man! Anything is possible in this world!

    ReplyDelete
  66. Saw this on HackaDay.com. And well, I hope to be considered a "quite large" response. I'd LOVE to see a write-up and specs of this build. As a high school biology teacher (and by necessity, a part-time MacGyver) anything that helps my students --and me-- better understand the complexities and some not-so-obvious simplicities of scientific devices is an amazing help in bringing interest to topics.

    I'm impressed you managed to put this together with a bunch of old parts. I'm wondering what made you think of building it to begin with... i.e., you had one part and thought, "Hmmm... I could use this as part of a DIY electron microscope." What was that one part and why did you think of this?

    And please: Details, details, details! While I may never have everything I need to do it myself (including the "time factor"), just knowing how and being able to explain part by part would be amazing!

    Thanks and Good Luck in refining to a higher magnification/resolution.

    ReplyDelete
  67. Very interesting project.

    Please do continue your work as it is very informative.

    John Bozzola
    Southern Illinois University
    Carbondale, IL

    ReplyDelete
  68. That is really great. Please post more details on the components.

    Jafar

    ReplyDelete
  69. Is that a Hitachi V-212 scope I see? For whatever reason I keep crossing paths with these scopes. I have owned 3 of them at various times in my life.

    ReplyDelete
  70. This shit is excellent, sir.

    ReplyDelete
  71. Really amazing stuff!
    I wanted to encourage more detailed follow ups. I am interested in applying the benefit of your experiences to DIY electron beam melting fabrication, if I get a chance. (Although that's going to be incredibly complicated!)
    See http://reprap.org/wiki/MetalicaRap for the most promising project I'm aware of. At this stage I'm still just researching the feasibility and mechanics, and have not involved myself with the group yet.

    ReplyDelete
  72. Greetings,

    Excellent work! Very interesting how you resolved various issues.

    James Jefferson
    Hitachi High technologies America
    Electron Microscope Division

    ReplyDelete
  73. Very impressive, Applied science at it's finest

    ReplyDelete
  74. A wow project, please keep posting videos.

    ReplyDelete
  75. Absolutely amazing! I wish I had 1/100th your obvious skills of ingenuity! Seeing people 'simply' decide to go ahead and tackle something like this is really inspiring. There is hope for the human race yet.
    I look forward to subsequent video reports!
    D. from Canada

    ReplyDelete
  76. Santee, I've always wanted to build a project that involved particle physics. It just seems cool to manipulate matter at such a fundamental level. My first plan was to build a cyclotron, but the device will emit lots of hard X-rays, and the product may emit gamma rays. In addition, the most interesting thing a cyclotron can do is transmute one element into another at small scale, which is really neat, but I don't have a use for radioactive isotopes. I thought a primitive, but working SEM would at least allow me to take some cool pictures of insects, etc. I already had the vacuum chamber, which I originally used for metal vapor deposition, and this saved a lot of time in the SEM build.

    ReplyDelete
  77. Inspirational! You have officially become one of my heroes.

    ReplyDelete
  78. Ben

    This is just amazing. For video output, have you considered using a modern GPU? I know FPGAs were mentioned, but the massively parallel nature of some modern ATI/nVidia cards might make both computations and video I/O manageable on one device. I'm not taking about the Tesla/Fermi architecture as much as I am perhaps a used/refurbished OpenCL/CUDA card for computation and presentation. Might be worth considering, and OpenCL CUDA is usually a lot more fluid than VHDL/Verilog.

    ReplyDelete
  79. My university has an electron microscope that is going to be tossed out. Are there any parts you are interested in? Like, there is a backscatter detector chilling on top of the whole thing.

    ReplyDelete
  80. I saw this on Wimp.com - you are awesome!! A great instruction that even novices can understand. Thanks so much, and congratulations! I'm not sure if you are a teacher, but you certainly would be a great one if you ever want to enter that field. Kudos!

    ReplyDelete
  81. Your work is very nice, first image must have been very exciting! Have you thought of using quadrapole magnets for tighter focus? My father designed ion sources and did beam development for cyclotrons for many years if you have any vexing problems, I can ask him. He never worked on SEMs but loves a good physics problem.

    ReplyDelete
  82. Andrew, sure you can break this down into smaller projects. If you are buying surplus parts, the first big project is just to construct the vacuum chamber itself and attach a high-vacuum gauge to be sure all of your pumps are working. Depending on your budget and luck, this alone may take a few months.

    ReplyDelete
  83. Amazing project. mind to visit simf1.blogspot.com I build F1 simulator at hom

    ReplyDelete
  84. Hello Ben,
    beautiful work! This is super exciting. As a part of my graduate work I used SEM to pattern hydrogels, tiny patterns:
    http://tinyurl.com/4d3cwyr
    Today, I work on various creative and artistic projects in DC (www.albuscav.us/krsko).
    I'd love to discuss a project that could utilize your wonderful SEM not only for imaging objects but also for fabricating them. Would you be interested, Ben? Thank you for the video that explained the SEM in very nice way.

    ReplyDelete
  85. WOW!! very cool indeed... you had me at hello... BUT please continue with more video...

    Cheers,
    Dan

    ReplyDelete
  86. very interesting video!
    I would love to see more videos of the microscope.

    greetings from Sweden
    Christian

    ReplyDelete
  87. You're the coolest guy I know of on this puny planet.

    ReplyDelete
  88. this has certainly be the most interesting diy project i've ever seen.. you must have a very big brain to even comprehend such a thing. i'm amazed!

    ReplyDelete
  89. I'm a grad student in materials science and I use a 5 kV SEM practically every day. Really loved your video and learned some things that I probably should've already known (had never looked into the ET detector before). I'm looking forward to seeing you at Maker Faire this year. Great work!

    ReplyDelete
  90. Dude your voice is so soothing. You should go into voicing sciencey things all the time

    ReplyDelete
  91. Awesome project! As was the nitrogen generator. Can't wait to see what is next. These are the kinds of DIY projects that inspire people in a really meaningful way. Keep up the excellent work!

    ReplyDelete
  92. Peter K, thanks for the note. You have a nice website -- very cool art projects. You can email me at ben at magconcept period com. My SEM is quite crude, so I am not sure how well it would be able to handle electron beam lithography. It currently has no way to modulate the beam, even though that feature could be added fairly easily. Talk to you later

    ReplyDelete
  93. Inspiring - thanks for putting so much effort into sharing your progress

    ReplyDelete
  94. You should team up with chemhacker he brought his scanning tunneling microscope to the New York Maker Faire. http://www.chemhacker.com/topics/stm/

    ReplyDelete
  95. Hi Ben,
    Absolutely superb project. I use a variety of electron microscopes in my work, and I find it absolutely fascinating that you were able to build a functioning SEM when my lab pays sometimes millions for them. Well done, and good job on the explanation of the system as well.

    If you ever do decide to add in extra plates to cope with stigmation, I recommend you use an octupole or at least hexapole geometry. Those plates will also give you some extra freedom for shifting the image or if you want to do e-beam patterning. As for your comments on the ET detector above, you are correct. The phosphor/photomultiplier set-up is used to get more useable signal. With a good setup every incident secondary electron will great multiple phonons, which can then create multiple electron/hole pairs in the photodetector, providing gain and a much higher detection sensitivity. Some SEMs and charged particle optics systems use a detector called a CDEM, but I don't know that you could make one on your own.

    Anyway, again, great job. I look forward to your other videos on this build.

    ReplyDelete
  96. Incredible! My dad helped Finders Uni in South Australia get its first Electron Microscope in the early 1960's. I was awestruck when I first saw. Huge, grey, taking a whole room. And now its DIY on the kitchen table. Astounding work. Well done.

    ReplyDelete
  97. Amazing work!

    I second the idea to, at very least, release your project as an open-source if you elect not to write detailed how-to.

    Thank you for inspiration.

    ReplyDelete
  98. If you have time, I would love to see more about this project. So interesting!

    ReplyDelete
  99. Hello Ben, I am in my first year of engineering. I was wondering if you are self-taught or whether you have formal training of some sort? I find that I learn concepts much more thoroughly when I work on this alone without the pressure of multiple assignment deadlines!

    Chris

    ReplyDelete
  100. This is really an amazing project. Please keep sharing more info about it.

    ReplyDelete
  101. ctiebs, I have a BS in mechanical engineering, so I took some lower-division physics courses, but most of my research for this project was done on the internet. In general, I learned many of my day-to-day skills from my father, reading articles on the internet and books, and just experimenting in the shop.

    Pressure from multiple deadlines is something that most people are faced with -- whether in work or school -- so I'm afraid it doesn't change a whole lot after college. It's true there aren't really any exams or tests in the real world, but there are similar stressful situations like preparing for a presentation, conference call, client meeting, etc. The main problem that I had with college curricula is that the knowledge being tested on exams is so completely arbitrary, and generally useless after graduating college.

    I had one professor tell me that college was a mental exercise, and that the content of the curricula was not as important as the act of learning it. I completely disagree with this assertion, but unfortunately, many employers feel as my prof did, and insist that their applicants have a degree. Apparently having a degree is simply proof that one can follow directions, memorize certain things, and show up on time for about 4 years.

    So, basically I feel that undergrad degrees are simply overpriced "quality control" certifications that employers and universities have developed for potential workers. Unfortunately, a degree is necessary to be seriously considered by most employers.

    Having said all of this, I really liked UCSB, and have great memories from my time there. Learning useful things in class would have made it all that much better.

    ReplyDelete
  102. Edmonton, in what condition is the microscope? I may be interested in those parts, please email me at ben at magconcept period com.

    ReplyDelete
  103. That is amazing! I hope you don't mind that I blogged your video here:

    http://snuze.blogspot.com/2011/03/kickin-diy-ass.html

    *hats off*

    ReplyDelete
  104. Great job. I am as impressed with the pro job on the 19inch cabinet as I am on the gun, etc.

    ReplyDelete
  105. Anon, I didn't build the cabinet myself. When I said "built from scratch" in the video, I was referring mostly to the electron column. For projects like these that will require lots of "tweaking time", I think it makes sense to save time on easy-to-buy items and I already had the cabinet in my shop.

    ReplyDelete
  106. Just beautiful. Looking forward to more videos.

    ReplyDelete
  107. Hello again, Ben... I just now found my "round tuit" and read these comments after posting elsewhere on your blog. I agree with your perspective on a college education. It took me ten years of part-time attendance (but full-time study!) to obtain a BEE degree while getting paid full time as an electronics technician and working mostly as a non-degreed electrical engineer for those technician wages. On the up side, my employer paid all my tuition and the classes were just down the hall (usually) from the lab I worked in. UC Santa Barbara is a nice place to spend four years; you probably enjoyed it a lot. Plus, it looks like you took some time to dabble in a little physics, some electronics, maybe a bit of chemistry... who says the Renaissance Man is dead? I am probably getting a little old for this sort of thing, but your blog has given me inspiration! I have tons of “stuff” acquired over the last fifty years or so that I could probably cobble together into something the current generation of budding mad scientists/geniuses would like to pick up and run with… maybe and online blog is the place to do that.

    As others here have said, I really like your style. You should maybe be working with ILM instead of fMRI people.

    Hop

    ReplyDelete
  108. Awesome, really. I'd love to see details of the electron optics, and also the electron gun structure. Can't wait for more video!

    ReplyDelete
  109. Very very very cool. Of homebrew contraptions I've seen this top the list.

    ReplyDelete
  110. Anon, my first plan was to use potentiometers for the focus and maybe Faraday cage voltage. The problem is finding high-value, high-voltage potentiometers. With a 10KV supply, a 10Mohm pot would draw 1mA, and dissipate 10 watts. This would not work, so I really need a 100Mohm pot that can handle 10KV -- very difficult to find. I bought a flyback transformer for a color TV with a focus control, and it indeed contains a high-value pot, but it is purpose-built, and very difficult to remove since it is potted in epoxy and contains other fixed-value integrated resistors. I found it easier and cheaper to just buy the supplies for each voltage that I needed.

    ReplyDelete
  111. Hey Ben,
    I had no clue what your vid was about, just started watching it and not much clue about physics or scanning microscopes either but I was really really impressed.

    Great job, I bet it's very rewarding to see that image after a few months of hard work and best of luck with your project. It seems to me it already attracted quite a lot of interest!

    Best regards,
    Serban

    ReplyDelete
  112. WHY IS THERE A APPLE LOGO ON THE FIREEXTINGUISHER? :O pruduct placement?

    ReplyDelete
  113. Awesome! Here's one vote that you should post the how-to step-by-step videos! You should win some kind of prize for this work, no doubt.

    ReplyDelete
  114. Anon, I actually don't own any Apple products except for an original iPod nano, which was a gift. I attached the included stickers to a fire extinguisher and shop vacuum. I thought it was funny -- iExtinguish and iSuck?

    ReplyDelete
  115. Congratulations Ben! Great Project!

    ReplyDelete
  116. Hi Ben,

    i am really impressed. i am expecting further
    detail information to be publicized soon..

    thanks and congratulation for your job.

    ReplyDelete
  117. hey Ben i was very excited when i saw your video. i have been thinking about building my own sem for several years. I read you have expierience with the propeller chip. If you generated the various voltages needed for scanning and focussing the beam in the propeller, you could do the image acquisition via a A->D converter stage + USB2.0 FIFO chip. I am myself waiting for this 15$ thing to hit the market(announced for may) http://www.ftdichip.com/Products/Modules/DevelopmentModules.htm#UM232H . you can directly connect this to an 8 bit ADC of your choice, all you have to do is let the propeller generate the clock signal for the ADC/FIFO combination. programming the windowsXP/linux for the module interface is easy

    would like to see some digital clean shots.
    i notice some ripple in the picture. is this the influence of the power supply?

    best regards

    ReplyDelete
  118. Simon, take a look at my most recent post. The waviness in the image was caused by stray oscillating magnetic fields. There are other sources of noise that I am still trying to reduce.

    Thanks for the link to the new FTDI module. It sounds like the most difficult part of using it would be writing a custom interface for the high bit-rate drivers.

    I am leaning toward FPGA for production of the scan signals. The Propeller is a very fast microcontroller, but the "dot clock" for 500x500 at 30fps is 7.5MHz, I am not sure the Propeller would even have enough speed to update the scan D/A and simultaneously control an A/D. It certainly could not also control a video output device.

    The analog display system has lots of benefits (like being very simple, very adjustable, already built, etc). There is still a lot of work left to do on the electron column itself, so getting high quality digital images would probably not be too helpful until the column can outperform my analog display device.

    ReplyDelete
  119. Hey Ben,
    you should take a look at this board. http://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=139&No=502
    24 bit video DAC is on board. Can be connected with http://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=73&No=360
    2xDA, 2xAD highspeed module
    Of course, money is the issue here, but note that you can get a discount if you work at university.

    I think each of the propeller heads can work at 80Mhz but im not entirely sure.
    7.5 MByte/s would not be a problem for the FTDI
    I have done some programming. FTDI is famous for having a very nice set of drivers for both platforms(win/linux) and there are demo programs in C++ and other languages for download.
    Simon

    ReplyDelete
  120. I just wanted you to know that you're video has given me a massive nergasm... bad. You sir are a god among men!

    ReplyDelete
  121. Hi

    If you get sick of all that oil in your oil diffusion pump you could head our direction in our open hardware EBM 3D metal and solar cell printer

    [[High vacuum Ion pump or turbo pump ]] ( Use electron guns to pump chamber down 50 50L/s max 1.5x10^-11Torr)

    Breakthrough ! This turbo pump cost has been a block to costs coming down, But after redesign putting 300 x 4mm holes in a stainless plate with titanium plates on either side covering the holes,thereby the titanium plates acts as cathodes and perforated stainless acts as anodes situated around the outside of the lens and deflection coils providing magnetic fields and a 7KV supply hooked up , we have create the first electron gun and ion pump combination.

    Turbo pump is best option but expensive ( Turbo pump e.g. turbovac 50 50L/s 1.5x10^-9Torr) .

    Also if you feel like going 3D imaging have a look at this , using 6 PIN diodes as pick ups.http://www.ecmjournal.org/journal/smi/pdf/smi98-17.pdf

    If you are interested in this we would be greatly interested as we intend to add it to our Home 3D metal and Solar cell printer http://reprap.org/wiki/MetalicaRap
    open hardware.

    ReplyDelete
  122. Hey, Ben...I am working at the Kagawa National College of Technology, and one of the departments is really focusing on SEM...one researcher is focusing on coating the filament with Yttrium Oxide...

    ReplyDelete
  123. Men, its a very complex project! Congratulations!

    ReplyDelete
  124. As a chemist who barely passed the electricity and magnetism portions of physics back in undergrad days (I'm also a product of the UC system--UC Riverside) I understood some of what you were saying! When I saw the title of your video I questioned whether it was some kind of hoax, having only ever seen the usual hulking commercial SEMs before. I add my accolades to those already posted here--very nice work.

    ReplyDelete
  125. Do you still think you are gonna post the detailed videos? I'd be really interested.
    Is the structure of the sample just determined by the fact that contours on the sample which are farther away from the sensor will appear less intense?
    Thanks a lot for the videos

    ReplyDelete
  126. Azra, check out my blog and youtube channel for more info. I am also writing a book about the SEM and other projects with the No Starch Press, which will be available next year.

    Yes, the high intensity parts of the image indicate that lots of electrons made their way from the sample to the detector from that region. Low intensity areas either emitted fewer electrons, or the electrons were blocked by the topology. There are a few other smaller order effects, but that is the main idea.

    ReplyDelete
  127. hello,

    Im a french particle physics enthousiast, I was on the way to build an electrostatic accelerator to learn the basics and i discover your DIY SEM. I ve built a complete remoted vaccuum system with alcatel primary pump and turbomlecular secondary pump. I ve decided to change my goal and follow your way. I already have a lot of equipment to build the same system, but i need th schematics of your raster scan generator. Is there a way to do that ?
    My website : www.accelerateurs.fr
    My adress nicolasleboucher@orange.fr

    sory, first version in french, but i will certainly translate it in english.
    Congratulations for your project and hope to read you again !!!!

    ReplyDelete
  128. Anonymous, the raster scan generator is composed of two 555 ramp/sawtooth generator circuits like this: http://freecircuitdiagram.com/2009/06/06/sawtooth-wave-oscillator-using-555-ic/

    The horizontal and vertical generators are not synchronized, which doesn't matter since the generator drives both the display device and the SEM. Good luck!

    ReplyDelete
  129. hello,

    thanks for your reply about the raster scan generator. I thought the circuit was more complicated !!
    one question: i ve 2 oscilloscope tubes that can be easily open. Why don t you use the electron gun of such a device instead of building it from scratch ?
    Have you study this way ?
    I m on the way to translate my website, and i will soon put an article about your project.

    Thanks again

    ReplyDelete
  130. The electron gun inside most CRTs uses a metal-oxide coated cathode that cannot be exposed to atmospheric oxygen. They are assembled, sealed inside the evacuated tube, then energized to activate the metal oxide coated. If the activated coating is exposed to air, it will be oxygen-poisoned and not emit as many electrons. Another problem is that the large cathode is designed for high emission, not a small spot size. The electron gun in a SEM uses a tiny electron source (originally a bare tungsten wire) to produce a very small source.

    ReplyDelete
  131. one more question about circuits: can you give more details about the amplifier for the photomultiplier ? thanks

    ReplyDelete
  132. nicolas, the anode of the PMT is capacitively coupled to a single op-amp. The op-amp gain is about 100, and the offset is controlled by a potentiometer voltage divider. The output of the op-amp is fed into the blanking control of my oscilloscope.

    ReplyDelete
  133. This is so COOL! I just ran across your blog and I had to comment. You were linked to by a colleague (he saw you in the Maker mag)on the Microscopy Society of America list server so you may be getting some traffic. I've been running commercial SEMs for 30+ years (also training other folks to use them) and I think this is the first home-built I've ever seen. The tube power supplies and all the knobs brings back memories.

    ReplyDelete
  134. Hi there Ben, after seeing your videos, you have definitly got me inspired and willing to produce some "science reckon". Ok, my interest are mainly aerogels and nanotubes, both used together could be usefull for firefighting ex.:(beter suits with high temperature resistence), and other applications maybe (subject to testing).

    Anyway, this is only one of the main purposes. Yet like someone said, "if you can't measure it, you can't improve it." (someone that is known but I can't recall his name).

    I've got a few questions to you about the SEM you have made.

    Right now I'm designing 3 Vaccum chambers, one to grow nanotubes, the second to make metal catalyst nanoparticles and the third for super critical drying (using your CO2 method).

    The thing is, before I get into nanotubes/nanotechnology (wich are the most interesting fibers around the world/and the most futurish problem solving technology) I need to have a not so costly way of watching their behaviour from each process method, I need an electron microscope.

    There are some parts that I can understand, but because igonrance is allways present, I'm having difficulties understanding the lens system.

    Does they have to be coiled? Or Could I just use a few Nd magnets (disc magnets) with a hole(North) in the middle /ousitde(South) to surround the copper beam tube to do the focusing and apperture jobs? (induced current in the south pole will influence the north pole?)

    I might be having some trouble understanding... but, nobody knows everthing!! :) Thanks in advance! Oh sorry for my English there..! :)

    ReplyDelete
  135. dz.dice, the electron beam lenses can be electrostatic (as in my design) or electromagnetic, or potentially permanent magnet. I'd guess the permanent magnets would have to be very large and unwieldy. You'd also need shimming coils to adjust the field strength, so I don't know of any SEM designs that make use of any permanent magnets for shaping the electron beam. Most (if not all) modern designs use solenoid coils with shaped pole pieces to concentrate the field into a small area. Machining the pole pieces and winding a coil is a lot more difficult that just using copper electrodes for electrostatic beam control. Good luck!

    ReplyDelete
  136. Do you have an EDX module on it? It should be a nice SEM like this ;-)

    ReplyDelete
  137. i actually worked on a raster scan controller based on arduino hardware during the past months :). It was originally intended for a simple STM.

    All the measurement, serial transmission and display stuff is done, just the x/y sawtooth output has to be completed (based on PWM and RC filter). Resolution would be 10bit at about 10-20Hz line frequency for a 512^2 image. Voltage range for input and output would be 0-5V.

    It'd be cool to see it actually doing useful things :) If you're interested i'd be glad to finish it soon.

    ReplyDelete
  138. BEN awesome,
    Thankyou for being the first.
    i am reading a book on holograms, you think haha
    what does holography have to do with an electron microscope but read holography H. Arthur Klein.
    I was thinking nm res too voltage focusing aberrations.
    But it is the more efficient and best way to provide.rhe focus. The book is real good they had the same prob you do I want to do EBL so your groundwork is awesome.
    WILL

    ReplyDelete
  139. Interesting project. A few comments.

    I've done considerable work with high vacuums (primarily for optical coatings). The use of aluminum in high vacuums is less than ideal. I am guessing that your vacuum is probably limited to about 10E-4 mmHg. Aluminum is porous and will generally outgass contaminants indefinitely. You may wish to add a nickel plate to it so that you can seal it. The mean free path for your electrons is going to be very short which means that you are going to get a distorted image.

    A questions: would it be easier to move the state (say using a piezo stage instead of deflecting the beam?) What are the upsides and downsides of doing that instead.

    Very nice project. Thank you for doing it.

    ReplyDelete
  140. needs more turboencabulator...

    ReplyDelete
  141. Sam, thanks. I'd say that moving the stage via piezo might be feasible for tiny displacements (high magnification), but not very useful for low magnification. The electron beam can sweep a 1 cm^3 area 60+ times per second, which would be physically very difficult for any mechanical system. Tunneling electron microscopes use a piezo-actuated probe tip.

    ReplyDelete
  142. Great project .I am interested for sure in making something similar as i want to make a DIY FIB .So to couple my unit to something like you have done will be a big success .

    Please post as much info as you can .As i am drooling(hommer's ahhh) enough!

    ReplyDelete
  143. Great project .I am interested for sure in making something similar as i want to make a DIY FIB .So to couple my unit to something like you have done will be a big success .
    Question : Couldn't you have used a CRT tube assembly and the whole TV electronics to make the same?

    Please post as much info as you can .As i am drooling(hommer's ahhh) enough!

    ReplyDelete
  144. HI BEN
    I am abhishek from India

    saw your homemade SEM project it was just AMAZING couldnt believe u did it alone.
    we are making an SEM as our collage project...
    its at its early stage and we are procuring the tools
    may be u can give me some ideas about it....

    ReplyDelete