I was the rare individual who was intrigued by junk lying around for collection, outside my university's science departments, a proclivity not shared by my friend. She thought it a bit weird. Fact is, I shoulda been going between classes in a pickup.
Like I said kids, stay in school, don't do drugs, don't get girls knocked up and you will be able to get a job that will pay you enough to collect scanning electron microscopes. I'm a childfree supremacist myself...
Great video Ben, I am glad that we managed to get it shipped from Sweden to the USA. I hope it comes to good use so that all your viewers can appreciate the vintage technology that comes with it! :) /Richard Andersson
Absolutely awesome Ben. Can never get enough of SEM photos, simply fascinating - and I love how the MDO is being used as the data acquisition engine. The HiRes sampling mode, with the amount of boxcar averaging you're getting, is giving you many more bits of vertical (video) resolution. Simply great!
Alan, thanks so much! I wondered if the Hi-res mode allows the scope to output more than 8-bits of vertical information. It looks like it does from the smallest deltas in the CSV file. Very handy feature!
Applied Science Yes, you'll get an extra 0.5 bits each time the # of bits averaged doubles. Thus, If Hi-Res averages 2 bits, you get 8.5 bit resolution. If it averages 4 bits, you get 9 bit resolution, and so on. The MDO3k samples at 5GS/s on a single channel. When you set the record length and horizontal scale, you effectively set the waveform sample rate (reported on the bottom right of the display). The number of samples averaged in hi-res mode is then 5GS/s divided by the indicated waveform sample rate. Thus, if the waveform sample rate says 500MS/s, then 10 samples are averaged to create each waveform sample point, giving you just over 9.5 bits. Fixed point math hardware limits the progression at some point...
Perhaps you could take scan requests from us laymen and let us glimpse into the microscopic world all around us. Not everyone in the world has the opportunity to use an electron microscope so! Maybe a monthly episode dedicated to showing some of the stuff people requested to be seen? I'd love to see some pollen/plant scans :)
Very nice. Since a SEM can seemingly be a rather simple device with low speed low cost image acquisition electronics and USB interface, I wonder if you might consider seeking venture cap funding and do a kind of makerbot SEM for the people. You could just start it up and pick some people to implement your directives so you don't actually have to do anything.
Applied Science What is the resolution on the microscope you were shipped? I am in desperate need of a SEM. It would be very cool if you work on Dan Frederíksen's idea. Me and countless other citizen scientists would be all over that.
How about a timelapse of something getting degraded by the electron beam. BTW how long does it take to pump down? Does it pump the whole chamber or is there an airlock for the sample chamber?
You may not need a timelapse to see damage caused by the beam. I was looking at strands of hair, and at 25KV with a large spot size, the beam will start to "bubble" the hair. It creates a fairly large blister in 5-10 seconds. Yikes. This SEM pumps the whole column down as one unit, and doesn't have an airlock or nitrogen purge. It only takes about 10 minutes to change samples, though. It also has two separate stages for small and large samples, which can be manipulated separately, and a few unused blank plates (I'll add electrical feedthroughs).
Applied Science would love to see some video of this and whatever else fancies you. Electron microscopes are just fascinating. I could look at anything under one.
Applied Science Hello Ben, are you afraind of X-ray radiation when the electrons are de-accelerated? I mean, the glass bell do not protect you very well.
Alexander A. The DIY SEM used acceleration voltages of under 10KV, and X-rays generated by these relatively low speed electrons will be stopped by a very thin layer of glass, so the bell jar is more than sufficient. Black and White televisions did not have lead shielding for this reason. The JEOL uses acceleration voltages up to 25KV, but the chamber is relatively thick stainless steel, and this sufficient to stop the X-rays generated by the beam, which is already extremely small (50uA emission current).
Ultra long exposures, that's an awesome solution! Could you use the same trick as astro-photographers by stacking many frames to get even finer detail?
That's an interesting idea. I think it would probably work really well. I'm imagining an image that progressively improves in realtime as more frames are acquired. Then, it resets when focus or pan is changed on the SEM. Cool!
Applied Science I'd love it if you'd do a piece on "Stack and Stitch" photography. Using a regular telescope you could get some fantastic pictures of the moon with super high resolution!
Also I'm very surprised at the images you managed to capture without having to plate the specimens in gold like most SEM operators have to do. The fly image especially is just amazing for what it is. I wonder what difference gold plating would do if you were to rescan the fly now that we know it works well enough without it? Would you get any better resolution/detail at all do you think?
Coating with metal will definitely help -- especially for higher magnifications, which require smaller spot sizes, which really need as much signal as possible. Most labs that have a SEM also have a dedicated sputter coater, so it's easy for them to always coat specimens, and write it into their protocol. My sputter coater is working, but was recently moved, and hasn't been setup. I tried scanning an uncoated insect and was surprised how well it worked. You'll never see a proper SEM lab do this since they already have a full complement of tools. Sometimes it's good to just try something and see what you get! In general though, coating with metal will help the image quality in every way.
I am amazed by your level of information and understanding. I'm almost lost for words after this video and my only explanation is that you must be Einstens second coming! I have no idea what so ever what you did or how you did it. Basically all I understood was the word "shipping" XD That being said it was still very enjoyable to watch, just as all you other videos! It's 04:28 here in Norway at the moment because I've just been up watching video after video from you!
With only 5V peek to peek, a relatively slow sample rate, and the low resolution ADC of the scope; would a 24bit/96KHz sound card not be better suited as a capture device?
If there is significant high frequency noise in the signal (and there is), then although the scope is only 8-bit, the effective bits after digital low pass filtering is much higher (as a rough rule, add a bit for each halving of the bandwidth). The best screens (DICOM monochrome LCDs, $$$) are only 12 bit output, and most screens are only 8 bit (though you can get cool temporal dithering software to extend this), so anything over 12 or 14 bits is overkill. Still it might be the cheapest option to use a 24/96 USB soundcard if you need a 48kHz pass band. Anyway, the proof is in the output; the SEM of that MEMS device is ultra-sharp and the SEM of the fly is well defined and amazing beautiful micrography. I would use something else, not for the quality, but because I believe every tool in my lab should be accessible and ready to use, and I couldn't afford to have an expensive scope setup like that permanently.
Ben do you do these experiments with others or are you doing this incredible work as a product of your own mind? I am in awe my friend..... my research is material as well but the level you go is without equal. Mudfossil University on youtube sends deep respect.
I am very hard to impress, but this is first time i see something original and useful with MDO3104. Still, i would simply take normal arm cortex m4, FT2232D chip, and create labview app that can take images at 16b adc, and that could yield more than 20-50Mpixels with more dynamic range Only tricky part will be how to figure out hsync triggering on mcu If you need, i can help you with interface with labview and mcu program to collect data and send to pc in semi real time
Cheers to Mr. Anderson for seeing the opportunity and to you for taking advantage of it. The scope gets a new lease on life in education! Looks like a high quality piece of equipment (especially after what you said regarding the shipping damage) Do you think you could use an audio card as a poor man's scope (with some conditioning of the input of course) to record these images? The MDO is lovely at it, of course, but it's not exactly a low cost item, lol. I believe the Wolfson card for the Raspberry Pi is ~$35 and the ADC seems to be able to record at 192ksps.
Yes, when I started making this video, I intended to use a USB audio adapter as the capture device. Mine is a cheapy model, and only does 44KHz, so it would work for the SEM's 60-second scan, but would not have enough temporal resolution for the 10-second scan. The audio device is 16-bit as opposed to the Tek's 8-bit, so there would be an increase in dynamic range. Also, all audio adapters will need to have their input coupling capacitor bypassed. I did this on mine, and it worked well. However, some models include digital high-pass filters implemented in the ADC itself, and these are essentially impossible to bypass (like the ones in the Handy H4n). The DC input range on my adapter was something like 2-3 volts, so the signal from the SEM would need to scaled and offset -- not too difficult, though. Other alternatives are the LabJack, Measurement Computing, and NI USB-600x models.
Very cool! And thank you for the short list of alternatives at the end, couple of things in there look like they will be useful! Also, ordered shirt. Paypal via Chrome was a bit wonky in case you hear anything. IE was fine (go figure, lol)
jcims You can use a USB audio card, but you need to remove any DC decoupling caps from the interface, and you don't get very good control on the actual sampling bandwidth. If you want to eliminate (or enhance) super-nyquist aliases, you should use a sample and hold ADC. You can get plenty of cheap uCs or FPGA boards with 250ksps ~ 2MSPS sample and hold ADCs, and it's kind of trivial to build an opamp circuit to bias, filter and amplify the input signal. If I had an SEM to play with, I would attach it to my Terasic C5G board with HDMI, so the picture shows up on the screen as it captures. It has a 500ksps 12-bit ADC which is surely fast enough for the task at hand. Unfortunately while FPGA boards are cheap and widely available, SEMs are expensive and rarely sold used in my neck of the woods. If I was to build an instrument it would be a scanning confocal microscope, rather than an SEM, but that's my preference, not a recommendation for others.
If you want to send a very slowly changing signal near DC frequencies, you can set a 555 timer as a square wave oscillator at a frequency that the sound card can hear then hook the signal to a pin on the chip which will modulate the frequency, then you can use an interpolated zero-crossing demodulator in a C program to recover the voltage trend.
Hi Ben, Another cool trick you can try with the EM is imaging the internal structure of chamber, like having a mirror inside the SEM. Mount a coin shaped disc with smooth side under the beam. Apply high kV (30kv) on the disc, let it charge up for about 5 mins. Then switch to low kV ( 3kv, or 4). You will see image of chamber, with a sort of fish eye lens effect.
Applied Science This.... is freaking amazing. For the novice tinkerer that does not have access to an SEM, this is a good tutorial on how to perform capture on a NTSC video signal. Very nicely done! I look forward to seeing how you incorporate this with other projects... you alluded to metal sputtering (which I know you've already covered in previous videos) so maybe we'll get some of of these in the future!
You might have the best job in the world. This was fun and extremely informative. Love the way you married old tech with a modern o'scope. I've been using scopes since the 80s but I've never heard of a mixed domain type, but it makes sense. I'll be looking it up. Thanks for posting. Subscribing.
It's a vertical column, as is conventional. It's true that it is shorter than many SEMs. I think the JSM-T200 was a no-frills sort of machine, and so the column can be simple and compact.
Enjoyed your vids for a while now, glad I can support and buy a shirt. I love how you teach something cool in an understandable way while also doing some pretty awesome things! It is inspiring as a maker to learn things that seem like they should be best left to the pros.
Thanks! I really appreciate it. It's often a good idea to learn from people that you suspect know what they are doing, however I'd be careful about assuming there is anything "best left to the pros." No one is good at doing anything when they are born -- it simply takes dedication and interest to develop skills as well as the courage to try something and fail a bunch of times.
Applied Science Thanks! Also, if you don't mind me asking, how difficult is it to get a job at a place like Google[x]? I am pursuing my PhD at Georgia Tech and want to have a similar job where you are given projects/problems to solve but am worried I should have gone into industry and gotten more hands-on experience as a research engineer or something more applicable. On the other hand, does a doctorate degree help when applying to such a job? Thanks again for your response.
skaterzero807 As you might imagine, [x] is pretty selective. However, our publicly announced projects show that there is a need for specialists. I'm not directly involved with recruiting, however in general, I'd say that candidates who have a good track record of independent contribution and strong interest in their field are most successful.
Ben, I saw your SEM at the Maker-Faire and was really impressed. It is an inspiration of "why not do it yourself" instead of the brain numbing consuming mentality. Hope to see you at the 2015 Maker Faire. For your shot noise problem . . . to sidestep the "fundamental problem", can't you just keep the voltage low to reduce image bloom, take a large number of slow image captures, then do a median pixel average to increase the signal-to-noise? When I was a teenager, I built an oscilloscope, and I remember a short persistent phosphor CRT tubes are available that reduced the bloom. www.bunkerofdoom.com/tubes/crt/crt_phosphor_research.pdf Can you use DeepSkyStacker or other astrophotography software, or even Median Filter Photoshop? You do address averaging to reduce sampling noise, so maybe you have it covered in a way that wasn't elaborated upon. Hopefully this makes sense. Thanks for inspiration
+Jim In Ashland This problem has nothing to do with CRT tubes themselves - watch it again. In fact what he just did is based on the same concept you described :)
I captured an image with a 10-second exposure of my camera aimed at the CRT: plus.google.com/115054970849159689228/posts/L2oN8GKSpfJ It works, but not as well as digital capture, of course.
awesome, i think maybe a electronics project to do the digital capture for you at least just for the raw data. Would also be interesting to see if you could capture in more detail at the 60s scan rate in 12 or 14 bits.
With the SEM you built, it's kind of a setup process to get a vacuum going, so I was thinking. Maybe build a round stepper motor controlled plate you could put multiple samples on, then just prepare the vacuum once and you can observe multiple samples. Just a thought I had.
Very interesting, and I laughed at the end! See: just getting really beautiful images, non-coated fly to blood cells and then what? Brief "okay see you next time, bye" :D Definitely want to see more. I didn't know it's not necessary to add coating... Cool!
Use patreon.com many RU-vid channels use this for crowd funding to improve their channel. Your videos have real world practically used science in them. Very rare on RU-vid.
Man, what I wouldn't give to sit around and play with a real SEM. Like most people, I've had various microscopes all my life from cheap-o Tascos with the ubiquitous prepared slides all the way to my current one thats actually has nice german made lenses, but I've always wondered how much detail I'm missing from being able to use a real scanning electron microscope. Thanks for the videos Mr. Krasnow, I love them all.
So using the camera to collect the final image would be like using a camera with a slow shutter speed for capturing star trails or other low light phenomena. As for the beam focusing problem, I now know exactly why I was having trouble with beam focusing when I was experimenting with CRTs many years ago. Had no idea it was because the electrons were repelling each other! I came for the science, but stayed for the knowledge. Ben, a thousand hats off to you :D
A cool thing you can do is look at a live electronic chip surface. The charge of the surface makes the EM signal brighter or darker, so you can directly see the circuit running!
farvision semiconductors are very small, those traces are incredibly small, but the main reason is because there is a plastic protection film over the die. So you have to remove that clear plastic film in order to see them light up or get darker with the SEM. Chances are soooooo little that the IC will still work. But its certainly possible without breaking the IC. The electron beam interferes with the protection film. The protection film is applied to protect the die from damage when it gets capped by the mfr. But main reasons: traces are to thin, and the protection layer which interferes with the SEM. But i'm sure you can see the traces light up or get darker when those things are done. But its very hard en almost invisible. I would love to see a live computer CPU under the SEM, that would be awesome especially if you can see what its doing.
Very well done. I'm glad that the carrier did not end up destroying it. Actually, itd be kinda funny... You know some folks get new computers when they get malware or a virus - perhaps they were throwing out the scope because the monitor was unplugged? :D I enjoyed the video, thanks for sharing - thumbs up.
Hi, I hope you read this, I'm trying to do the same thing but I can't "translate" the CSV file to an image, I do not have experience with Octave but I wish you could share your Script. Thanks
cool,..hmm could you do something like you did with the x-ray chicken and produce a 3d model of something with this,...i know it would take a long time but it would be pretty cool,.. maybe you could make it produce a file for 3d printers so you could print larger versions of whatever you have in the scope
I really thought you'd have to sputter coat the fly. I thought that was just an absolute must. Every one of your videos has taught me something new. Most of them, including this one, teach me several things. Do you have a donation button on your site somewhere? I don't really need the shirt even if it is really cool looking.
Great video. My thoughts changed from 'Wow! What an amazing microscope!' - then to - 'Wow! What an amazing oscilloscope!' - finally to - 'Wow! What amazing pictures!'... Wow!
Hi Ben: A couple thoughts: 1) At some point the long exposure of the CRT will have diminishing returns. As the exposure increases so will the accumulated noise on the sensor. However I think a bigger area of image degradation will come from scattered light, although the scan beam is a “spot” the phosphor is going to throw photons around causing some “blooming” of the spot and might become a problem with slower scans and longer exposures. Someone suggested averaging multiple exposure to improve S/N; I would think you could do similar and possibly better by averaging multiple analog captures. Although, what you have shown just “out-of-the-box” look great. 2) You mentioned the edge distortion seen in your un-cropped analog captures. You might be able to correct this as they do for glass lenses. Manufactures of photographic glass provide a distortion profile for their lenses (e.g. barrel/pincushion, angle of view, vignetting, etc.). These lens specific profiles are provided within Adobe Photoshop and Lightroom, and others. Adobe has outlined a procedure and software to generate custom profiles (www.adobe.com/support/downloads/detail.jsp?ftpID=5492). Basically, you image a rectilinear grid and the software calculates an orthorectified correction matrix, Anyway that's my 1.99999 . . . cents. Cheers, Mark
hmm. How is it that the oscillocope shows a live display of the actual sample? Whenever I've seen an oscilloscope used, it's a sin wave. How do you convert the sinwave unto an image? I see in the video Mr. Krasnow explains, but I'm having trouble bridging the ideas
Look up the video on how analog TV works, instead of drawing a wave shape which would be a graph/chart of the speaker cone going in and out over time, you draw a series of lines starting at the upper left corner and going from left to right and top to bottom, then you use the Z axis to cause the brightness of the dot, if the image is synchronized and you do this fast enough from a video signal, you will get a picture.
Ben, I am amazed at how much you gained from trying what, by design and original patent, was meant to be inaccessible. You have made innovation a possible step next to online mentoring. How I wish the transparency of the developed world would be complimented by synergy from your proteges in the developing world without schisms. Do you envisage Maker faires ever possible in Africa?
I could watch this all day. Could this be done with a plebian scope like a ds1054z? Edit: Yes it can. The buffer length is large enough when hacked to store up to 12 million samples, and the timebase and gain can be similarly set. I used this for slow scan TV from an RTL SDR but same principal.
Please more SEM stuff !! Are there design improvements to be back ported to your DIY SEM from the professional one ? What is the resolution difference between your DIY SEM and the professional one ? Is the DIY SEM able to get better pictures than a good microscope ?
Hi Ben, I'm Jake, a Physics major at Boston University. I just found out about you, and I want to make my own E microscope. I am watching all the videos/info you have released about it. I am wondering if you could help me out by posting some more detailed instructions on how to make my own electron microscope. I know you are very busy so I don't expect you to post such a detailed guide when you surely have exciting specimens to analyze using your working microscope. However, I was wondering if you could direct me to some links / books you found helpful while building it and maybe send me a rough outline of how I should go about building this (ex. First read this, then buy this, then do that) . Citizen Science is by definition open and free to all - but it can only work when knowledge is shared. Ben, I think you are literally one of the only people in the world to have done this - and thus one of the only people in the world who can make it easy for everyone else out there to also do it. I want to live in a world where science funding and direction is not determined by the agenda of those who have the most money and power (ex. DARPA, the Pentagon, Lockheed Martin etc......) jakeyu@bu.edu
Funny, first time ever I see a machine which is using the Polariod cameras I got for really cheap (~20 USD). They make really amazing sharp (nearly grain free) images but only with narrow focus length.
Wow, that great: I was thinking how many can say (outside industry or the University) they have an electron microscope, but no, a "collection." Fantastic - Cheers, Mark
You inspired me! I have JSM-T20 with set of spare parts which I recuperated from the university. I never tried to make it running. Any idea what is the difference between them? I have 2 CRT monitors and a few more knobs but otherwise they look almost identical.
Thanks for the video ! ....with your amazing knowledge can you give me one example of objective reality when it comes to the "belief" people have that they live on a ball ? .....fluid mechanics shows us water is level unless it's flowing or otherwise.... It's also visible, measurable recordable and repeatable .....we know there is zero curve or proof of motion spinning faster that sound racing around and chasing the sun at 65,000 mph and 550,000 respectfully. Simple logic and understanding of the natural world around us proves without a doubt we do not live on a spinning ball.... Unfortunately the chances are that I will have receive a barrage of logical fallacies and sophistry from the "high minded" individuals that rubbish in the claim that we all can experience and lean ....pseudoscience is nothing more than a religion or a belief system for mass control. For anyone who mentions the sky, understand that it does not in any way tell us the shape of the earth we all stand upon. One of endless ways we can understand our world is to think of water flowing down the gradient of the mountain and then onto a level ocean where it levels out ....we cannot have two "downhills" in one reality where one (off the mountain) water flows downhill and the other (curve of the imaginary ball) the water "sticks" and does not flow off ! think of how mad the whole idea is ....this pseudoscience nonsense was pushed on to us from an early age ...We should be screaming it from the roof tops ! Want your children to be lied to about our very foundation ?? Or will the zombies defend what's is obviously on true .... Catch a bloody grip men and women ! We do not live on the nature define potato rock racing true what they called space..... Repetition repetition repetition ....ball earths in your face from young age use your higher minds abe wake up for the sake of the next generation.... They play with words ....Horizon equals horizontal government means to control the mind mind control "govern" "meant" .....how "intelligent people" can't see this is beyond me .....absolutely breathtaking ! ....for those who to try and defend the impossible ball all you have is a formal science (a written science) no natural science attached ....it's nothing more than a language which can be manipulated to fit their model.... Do not ask me for a "model" it's an on going investigation for which we need ....explorations... which translates as science !! Real science like the gentleman running this channel does ! the guy is amazing ! Extremely articulate a good teacher ! Who ever reads this will know all they have is rhetoric and fake images from NASA and the likes ! For those who have open minds on this objective everyday reality that we all experience ...they will come to realise the likes of Brian Cox are nothing more child abusers lying through their stinking teeth to Young innocent impressionable minds !!! That's what really pisses me off !!!! And bloody well every else who cares about the very foundation of humanity ! Good luck trying to find empirical proof you live in a ball ....you will never find it !! only in cognitive dissonants ! Pure fantasy ....who's going to be honest with not only themselves but their children ???? ....what I find amazing is instead of scientific empirical proof people will throw insults ....indeed so, not very scientific... Thank you for the video ! ....this is in no way directed to you sir .... The earth has always been a static plain ....it's what you find on the "other side" that really opens the mind up ....majority will never find this unless they find common sense ....that's not common anymore.
Hello Ben, How about building a 3D electron microscope? Whould be awsome :-) Mount the object on a stepper motor setup, rotate, interpolate and build a 3D image on your compute by calculating the raw data.
shirt logo mildly looks like steam logo. Half life 3 Confirmed. I kid. Love the videos man, this is great. Definitely grabbing a shirt. Unrelated, but any news regarding that controller you'd posted a video about using the wheatstone bridge a while back? Was a clever way to utilize the setup, was curious to see if you'd done anything with that.
Awesome, wish I had a desk top version :/ Any chance you could take an image of a drop of water ? Dip a plant leaf in a cup then drop water from that cup on a slide and see if differs from regular tap water ? Cool stuff so much to get evolved in... Hope you keep the images coming, eat more healthy fats ( plant based un processed ) and give the red blood cell another go. Eat raw onions the long green kind and raw garlic, small peace of a garlic clove and one onion chopped and place on top of your already cooked omelet will do the trick. next day see how many white blood cells pop up, you`ll be surprised I was... Back in WWI they used raw onions and garlic instead of antibiotics...
Love your channel because of the wide field of science you present that is very technical, yet 98% understandable to us that do not have a science background, thank you very much.........One more curious note: the ads RU-vid puts on your videos, like the one here about using an oscilloscope for microscopy, the ads are for a manufacturer of microscopy instruments, very interesting note..!
Brilliant use of the oscilloscope with the old SEM! Could this be done with a less expensive scope that has a capture feature? Just bought my T shirt. By the way Ben, I would support your uniquely awesome channel even if the T shirt was not part of the deal. Your great vids are more than compensation. I believe that Afrotechmods is using Patreon to allow viewers to help support his efforts. PayPal may have some sort of tip jar feature as well.
Pardon me, but could you please use one of these to analyze the cut down tu104 in the 2060ko that gamers nexus discovered and figure out why it games like a 150 but computes like a 104?
I picked up the Teespring Domination video course (google it),I pretty much turn a profit on most campaigns. Its still the best Teespring course online IMO, and I have tried about all of them.
Very nice. Now if you could just automate all the clicking and twisting knobs and copy and pasting, into a compiled script that can be included in anyone's electron scanner. Other people - who only want to look at and use the images and data - could have access to your innovation. I did not spend half an hour to review the literature on the state of coating requirements for electron microscope targets. One reason I never bothered with electron microscopes is because (well of course they are god awful expensive) it takes so much fiddling to prepare things to look at. I watched how many things you had to manipulate, and you already knew what you were going to do. Do you think anyone without substantial oscilloscope training could do that? I bet every single oscilloscope on the planet has unique and proprietary software and their own rigid settings and specifications. Not sure if you brought it to the point where everyone can have a pocket electron microscope. But this definitely helps. I think you are ahead of everyone. And it will come to nothing. You keep showing people things but your tools are so expensive, and require a highly practiced person in the loop to even replicate. I keep thinking that there must be a better way. If you make your own electron microscope one day with decent amplifiers and ADCs you can go way beyond this. At any given damage level for electron beam intensity, you could run this for days or weeks at a time. Why does everyone always assume it has to run in a few seconds? If I had a low cost one, I would let it look at the 3d electric and magnetic field variations from individual electrons as a function of time of contact. I would collect data on some targets for years or decades, just to see how far data could be pushed. Why do people only pick up data from single points? Optical sensors and amplified arrays with ADCs can pick up tiny voltage and current variations. And you use a monster oscilloscope to read one or two signals? Oh well. As much as I might dream of having a few petabytes of data focussed on a single object at high resolution, I will never even get to see the data from one electron microscope. When I was a kid, all the tvs and oscilloscopes were electron beam devices. It doesn't matter if they are sloppy, if you measure carefully and record EVERYTHING. Like a shaky camera with high resolution inertial data streams, jinking the object gives you superresolution dimensions. You want the tiny movements to be random, so the statistical analysis gets easier. No wonder kids are so depressed today. All the wonderful tools are locked away. You cannot even see the data. Or only without enough reference and environmental control datastreams to be useful. Like those NASA images in lossy formats and no audit trails in human readable or script readable form. I am so torn. Very nice work. You always do a good job with much better than average insight and creativity. But you always leave everyone with "I wish I had those tools", not "I will try that myself." Or maybe that is just me. Thanks, but something is missing. I guess I will end my life soon and never been able to play with electron microscope data. You talk about pixel resolution, but it is not the instantaneous images, rather the whole of a very long experiment and data gathering experience that determines the outcome. You did not even do your own stacking algorithm. Doesn't the beam give off photons and heat? What kind of detectors are you using? Is it just one, or a large array?
Meh, fudge it.. I'm reserving it now! If 18 more people would please do the same so the campaign can be relaunched! Ben, if you happen to read this - what size were showing in the vid?
I've heard that the Covid - 19 virus is 0.04 uM in size. If true, I wonder how relative in comparison that would be to the black dots on the fly's eye in the photo at the 11:40 mark?
Applied Science, Great job of the video and the technique to acquire a better S/N and more resolution. Since the scan rate is so slow you might consider using the sound card input for the DAQ. As you the sound card has at least 16 bits and some have 24. So with an appropriate scan rate and that resolution your pics should look even better. Thx again for the video and the insight.
