Antoine Ghanem
e-mail address : ghanem-solvay@e-mail.com
Hello,
I realized that my first message about maxima and minima in SEM magnification
might have been misleading, so I would like to provide some details. What we
plot is the ratio of measured and nominal magnifications as a function of
nominal magnification. The measured values come from black and white
videoprints (Sony UP-850). The ratio ranges betwwen 0.75 and 0.94, depending on
the microscope setting. The average and standard deviation of 73 ratios are
0.83 and 0.05, respectively. The curves exhibit the following behaviour :
increase to a maximum about 200x, followed by a decrease to a mimium about
1000x, then an increase to about 3000x followed by a decrease down to 10000x.
Please note that the resolution of the curve is rather low (8 data points in
total), and it's not always easy to measure the magnification below 200x with a
10 micron spacing micrometer (which means the maximum at 200x is only observed
when there are data points below 200x).
I hope this rings a bell to somebody...
Antoine Ghanem
ghanem-solvay@e-mail.com
Antoine,
If you haven't received many replies to your magnification problem, it may
be because there are many sources of possible error in SEM magnification.
What the mag circuit does is define the current through the scan coils and
these establish a maximum scan ANGLE that raster will have as it leaves the
final lens. How big an AREA this raster will cover on the specimen depends
on how far the specimen is below the lens (the working distance, WD) and
the square-root of the energy of the electrons in the beam.
Old SEMs where designed to only work at one working distance to eliminate the WD variable. Later it was found that one could get a measure of the WD (at a given beam energy and assuming no hysteresis in the magnetic circuit of the objective lens) by measuring the current in the final lens and later instruments used this information internally to compensate the scan current. Likewise the scan-currents were normalized to provide the same scan angles first at a few, and then later at a large variety of beam voltages.
To make all of this compensation work, there are a number of potentiometers in the scanning amplifier to allow the magnification to be adjusted in x and y under a number of standard conditions (certain Mags, WDs and kVs). It would seem at first glance that these pots may not be prperly adjusted on your instrument.
Normally SEMs are adjusted so that the magnification refers to the ratio of the length of a feature on a photograph (usually a Polaroid) produced by the microscopy divided by the length of the same feature in real life on the specimen. You have added the additional complication of a video printer. These can be useful but often only operate at video-scan-rate where SEM image distortion is high and where the printer makes specific assumptions aboutthe dimensions of the TV raster (number of lines, ration of H to V) that the SEM Manufacturer does not fulfill (and this is particularly true if you use a mixture of PAL (625 line) and NTSC (525 lines) equipment). You may also be using a computer-based image capture system that has a video-rate display which you print with your printer. On the assumption that this image-capture system is active (generates the scanning signals) rather than passive (listens to the scan signals generated by the microscope electronics) it seems likely that the misadjustments are in the scan-generation software of your image capture system rather than in the potentiometers noted above.
A lot of this information can be found a good basic SEM text.
Good LUck,
Jim Pawley
JBPAWLEY@FACSTAFF.WISC.EDU