Posted by
James Pawley on
URL: http://confocal-microscopy-list.275.s1.nabble.com/exorcising-spirits-from-Fluoview-tp3781225p3781823.html
>I apologize if this is a second message to the list, I think the
>first one didn't go through
>
>We are having a bizzare scanning problem. Straight vertical lines in
>an object become slightly zigzagged with a period of up to 6-7 scan
>lines, and there also may be some oscillation in the intensity. The
>period and the magnitude of this periodic noise depends on the scan
>speed and the scan size. So far we (with the help of an Olympus
>engineer) have established that:
>
>1. It doesn't seem to be the scanner controller or the galvo mechanism
>2. It does not seem to be the electric power in the building
>3. It is not a mechanical vibration
>4. It is not a computer
>5. It is not the cables
>6. It is not a 60 Hz noise
>
>Sometimes connecting the scanner controller to the outlet through a
>long extension cord seemed to help which may suggest a problem with
>grounding, but as soon as we concluded that, the trick stopped
>working. The trouble could not be reproduced at the Olympus testing
>lab.
>
>Has anyone experienced anything similar and successfully resolved the problem?
>
>Many thanks in advance!
>
>Mike Model
Hi Michael,
You say "6-7 scan lines" but do not mention the horizontal scan rate.
Assuming it is about 2 ms, then the problem would seem to have a
period of 12-14 ms (or about 83 to 71 hz), which is close the line
mains frequency. Your experience with the longer extension cord also
suggests ground loops.
Ground loops are present whenever any part of the electrical system
contains conductors that permit current to travel between any two
points in space by more than one path. Ideally, any sensitive wiring
is connected in the form of the branches of a tree, especially any
wiring such as that related to sensing small feedback (such as that
between a mirror position sensor in the scan unit to the scan
amplifier in the electronics?).
In modern microscopes, these problems are sometimes avoided by
digitizing the analog sense signal before it is returned, and also by
keeping wiring for the "earth" return of power circuits separate from
that of signal (sensing) circuits.
The efficacy of this separation can be defeated by any conductive
path that connects the two systems (such as an added accessory that
is connected to grounded plug in the wall or some concatenation of
metal items connecting the two systems - maybe a pair of scissors
resting between the laser benchtop and some unpainted part of the
microscope.)
They can also be connected inductively. A significant mains-frequency
magnetic field (more than a few mGauss) will be produced by any high
current flowing "in a loop". For instance, if some high current
device nearby, or in an adjacent room, is wired incorrectly so that,
instead of the hot and neutral currents passing through wires that
are in the same power cord (i.e., very small loop with dimensions of
mm, and therefore making a dipole field that decays with the third
power of a distance in mm) the return current may flow through some
metal parts of the device and back through some safety ground strap
(making a loop with dimensions of meters that may not be very large
but that decays a thousand times more slowly.
If such a stray magnetic field is present in the area of your
instrument, there are a number of ways that it can induce an
erroneous alternating mains-frequency currents in sensing circuits
and this will cause the amplifiers that drive the scan mirrors to try
to compensate. As they are compensating for a position error that is
not actually present, the result is to create a position error.
So I think that the most important thing is to try to find out the
frequency a little more accurately (measure the horizontal scan
period and find out the number of lines more accurately by counting
the lines in say 20 "wiggles" and diving the result by 20). If it
does seem to be mains frequency where you are (50 or 60 hz), then get
a stray field sensor (a coil of wire and a sensitive oscilloscope?)
and see what kinds of fields are present in the area. You do this by
waving the coil around and twisting it every-which-way as you do so
to get the maximum reading. Remember magnetic fields have a direction
as well as a magnitude and you will get the maximum reading only if
the plane of the coil is oriented perpendicular to the direction of
the magnetic field lines (or parallel to the loop making the field.).
See whether this reading changes when nearby equipment is turned on
or off and try to isolate the source and remove it by proper
grounding.
Alternatively, you can do the reverse and MAKE a large field by
passing mains-frequency current through a coil (a coil used for
degaussing the CRTs in old TV sets can work well, but if you make
your own, don't electrocute yourself!). If the field from this coil
creates an artifact similar to that you originally saw, then you know
that you are getting close. By varying the coil's position and
orientation you may even be able to cancel the effect out! But this
is not recommended as a permanent fix.
If the problem is new, then something has changed and It is probably
either a change in the ambient stray field or alternatively that some
connector is not making good contact (wiggle the connectors?) or that
a ground loop has been created by the addition of an accessory or the
happenstance position of some metal parts touching things they should
not touch.
It is also just possibly a mechanical effect: Mains frequency
vibration from some transformer etc. being coupled into the scan unit
and making the mirrors wiggle.
Hope that this helps. I have a couple of PDFs of articles about
hunting down stray fields and currents in SEMs and can send them if
you contact me directly.
Cheers,
Jim P.
--
**********************************************
Prof. James B. Pawley, Ph. 608-263-3147
Room 223, Zoology Research Building,
FAX 608-265-5315
1117 Johnson Ave., Madison, WI, 53706
[hidden email]
3D Microscopy of Living Cells Course, June 14-28, 2010, UBC, Vancouver Canada
Info:
http://www.3dcourse.ubc.ca/ Applications due by March 15, 2010
"If it ain't diffraction, it must be statistics." Anon.