Re: Tek 7000 Series deflection plate precision?


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.

--- In, "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

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

How *wide* across can the electron beam be?


Join to automatically receive all group messages.