Re: Tek 7000 Series deflection plate precision?


iglesia_cristiana_arpas_eternas
 

Hi Damian,Here my two cent about:
You need to put an extra objective lens "magnetical prefereble" in order to obtain a very small dot at least half what minimum size yo like to see.and ..or.. increase the tube length, for the same reason.

Most people use a Cristal plate with phosphor where the second emission electron emitted from the target impact and release a photon ,which in turn will be detected by the photomultiplier itself, I believe that such optical conversion is not necessary , just remove the cap on photomultiplier and allow the electron hit the dinodes directly ,If this approach work, surely will be give you a enormous gain and increase in S/N ratio, which in turn allow to use less beam intensity with all benefits.obviously all in high vacuum, inside the chamber.
End of my two cent.
Gabriel.

--- In TekScopes@yahoogroups.com, "cheater00 ." <cheater00@...> wrote:

Hi guys,
could someone chime in on how precise the deflection in a Tek 7000 can be?

I'm very new to vacuum tubes and the idea of electrostatic deflection,
so please bear with any stupid questions...

My main consideration is: with modifications, would it be suitable to
jury rig an SEM (Scanning Electron Microscope)?

I am trying to do something inspired by Ben Krasnow's SEM design. If
you don't know it, he has some info here on how he made an SEM
himself:

http://benkrasnow.blogspot.de/2011/03/diy-scanning-electron-microscope.html
http://benkrasnow.blogspot.de/2011/03/diy-scanning-electron-microscope_26.html

He made the raster generator and deflection assembly and amplifiers
himself. It was my idea to use a broken 7000 series tube and hook it
up to a working 7000 scope configured with two time bases in a raster
generator setup. Of course I'd need a Tungsten filament, the one in
the 7000 CRT would be broken by having been exposed to air.

What happens in an SEM is that the electron beam *scans* a rectangular
area on the thing you're zooming in on. It moves in a raster image,
just like on a TV, or an oscilloscope in raster config. This rectangle
has to be very small, let's say 2mm across. You'd start out with a
fairly big rectangle, say 2x2 cm, to see the overall shape of the
object, then you zoom in a bit so that the rectangle is 1x1cm and pan
a little, then zoom and pan again, and so on. Think CSI: Miami
("enhance"). You might end up with a square of 2mm or even smaller.


Here's my question: the deflection needs to be precise, so e.g. if
you're horizontally 5% of the way, it still needs to be this far
across, and not erratically in some other place.

It would be bad if it e.g. jumped from 0% to 10% to 20% and so on to
100% while skipping the intermediates.

It would also be bad if it e.g. centered around that 5%, but noise and
interference meant the beam would randomly err between 1% and 9% with
5% at its center.

The rectangle as a whole needs to be moved at least in this area of
1cm or 2cm as well. It would be fairly bad if the rectangle as a whole
drifted or if the position were shaky.

How would one modify the deflection circuitry and assembly in a Tek
7000 to achieve such a small scan area? Is the circuitry going to be
low-noise and/or linear enough so that deflection precision is
retained?

How *wide* across can the electron beam be?

Thanks,
Damian

Join TekScopes@groups.io to automatically receive all group messages.