10 µm Scanning Electron Microscope From Vidicon Tube

[Segelfam] built his own scanning electron microscope. He based the machine around an old Vidicon tube, a video recording technology that was used in NASA’s unmanned space probes prior the Galileo probe in the late 1970’s. We struggle a bit with the machine translation of [Segelfam’s] original build log, but it seems that he filled the tube with helium in order to convert it for use as a microscope. But don’t worry, if you’re interested in this hack the information is all there – between the forum thread and build log – it’s just a matter of putting it all together to fill in the details.

In case you were wondering, the image to the upper right has been colored using Photoshop; the rest are straight from the SEM.

[Thanks Jerry]

13 thoughts on “10 µm Scanning Electron Microscope From Vidicon Tube

  1. helium is needed when working on the tube (cutting the frontplate off, changing subjects) to protect the cathode. when work is done, it is assembled with the metal t-adapter to form a vacuum chamber.

  2. neat, wonder if this would work with:-
    a) a dismantled under helium surplus crt from a 1980’s vintage camcorder,
    b) a surplus used VFD with hole carefully ground in the glass under helium
    c) other small crt such as one from a door camera.

    a) has the advantage of being relatively high resolution methinks, the picture on these is something like 1cm square and equivalent to VGA at least.

    I was just looking into getting some argon for making some chemicals that are unstable in air (magnetic particles for DIY magnetic solder)
    Not that expensive I might add, works out at £30 from Bougourd & Harry for an argon cylinder + mini regulator.

  3. acually, helium is only needed during the assembly and when changing probes, as he writes the kathode would be damaged being exposed to air for too long.
    anyway, nice find and beautiful hack.

  4. The site actually references the resolution as 20um, which is just 5x better than the human eye (with a loupe you can better this specific incarnation of an electron microscope…)
    The whole setup is aimed to be a demonstrator for schools (the company specializes in such). Still, very cool recycling use. The HAD blurb lets the Vidicon seem more fabulous than it is: It’s not just a space-borne sensor, it was at the heart of most television-videocameras back in the day.

  5. If anyone needs vidicon tube I found a stash at a local surplus of new-in-box vidicons for cheap.

    Also you could use argon as a cover gas as well. Probably even pure dry nitrogen. Its the oxygen that causes the problem.

  6. anyone need a old vidicon tube from a rca camcorder?
    message me on the forums if you’re interested
    it has the tube and the focusing coils(it was too cool to just throw out when i scrapped the rest of the camcorder)

  7. @loonquawl, I agree, the resolution is surprisingly low. One might do better by extracting the deflection stage from the vidicon and putting it after one or more home-brew beam condenser stages.

  8. Might be able to improve on this with better vacuum, also the problem might be that the vidicon was designed to scan AND return a signal from the sensitive screen, so the electron optics were optimised for a relatively defocussed return beam.

    I reckon that a ‘scope or camcorder CRT would work well for this, due to the small spot size.

  9. Here’s my take at a translation, hopefully it is a bit clearer than what google and babelfish put out:

    Before the advent of CCD sensors, most TV cameras were fitted with Vidicon or Plumbicon tubes. Both types of electron tubes differ only in the composition of their respective light sensitive layers. In those tubes, similar to CRT display tubes, a fine electron beam is generated with which the photo electric charges on the light sensitive layer are scanned and converted into an electric signal. In opposition to display tubes, the scanned surface is very small, hence the electron beam must be very fine (well focused) to achive the required resolution of several hundred lines.

    The fine electron beam inspired Rapp-Instruments to misuse a Vidicon as a scanning electron microscope (SEM). A fine electron beam is used in a SEM to scan the sample, whereby secondary electrons are generated which are used to create the picture. To make it happen, the Vidikon needs to be opened and fitted into a vaccum chamber. This works needs to be done quickly as longer term exposure to air will damage the delicate cathode in the tube. The best thing to do is to fill the tube with baloon gas (helium) immediately after cutting, and to then hold it with the opening facing downwards only. When changing samples later, the changer is never flooded with air but always with Helium. This way, the cathode will survive many cycles of probe changing. For the vacuum chamber, an off-the-shelf T-piece is used to which the opened tube is glued in. For operation, a pressure of less than 10^-4 bar is required, which is reached by using a diffusion pump in this setup.

    For the following measurements, a 2/3″ Vidicon type XQ 1272 was used. At 6V heater voltage, the emission regulator is used to set a current of a few µA. The secondary electron current generated at the sample will then be in the range of several nA.

    The raster currents for X and Y deflection are generated via D/A converters by a PC. The image signal is read by the PC through an A/D converter.

    Normally, secondary electron currents generated at the sample would be accellerated towards a scintillator, whose light would then be measured wiht a photo multiplier. This way, high amplification with and low noise can be achived. For this simple setup, another way has been chosen. The secondary electrons are sucked off by an electrode on high potential. This electron current can be measured at the sample holder and it is used to generate the raster image. This way, the current can be measured at ground potential which simplifies the setup significantly.

    These images of fine wireframe at different magnification levels (20-, 40-, and 80 times) demonstrate the usability of this experiment as a proof of concept. The resolution is approx. 20µm, a value that is determined by the electron optics of the Vidicon and is withing the range of what could be expecte.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.