Raster Scan (Electron Microscope) Rotation Algorithm?

[Jeff Miller]

Hi,

I haven't made a whole lot of progress on my electron microscope
digitizing project this month, but I hope to finish in the next 6-8
weeks. Perhaps sooner. The last decision is whether to use an MCU or DSP
in the project, or to stick with discrete logic as I'd originally planned.

To recap and give some background, the heart (arguably) of a scanning
electron microscope consists of a pair of X and Y scanning coils which
are driven in a raster-scan pattern not too unlike a television
deflection yolk. In fact NTSC scan rates, interlace and all, are
sometimes used so that the output can be displayed on a TV monitor.

As it turns out, scan/image rotation is a fairly common feature of high
end scopes. I'm wondering what it would take for me to incorporate such
a feature into my project. I'm not sure whether or not this feature is
limited to digitally controlled scopes: I'd very much like to know.

I'll be driving my coils with 16 bit D/A converters: I want to
experiment with ultra-high resolution images blown up to poster size.

Of course rotating the scan in multiples of 90 degress is trivial, but
ultimately I'd like arbitrary rotation or at least in 45 degree
increments. I could probably get an idea of the algorithm I'm looking
for by drawing some superimposed squares, mapping point to point and
finding the pattern. For example: if the baseline (non-rotated) image is
generated using 0,0 as the upper-left origin and proceeds right and down
as it does with traditional TV, then for a 45 degree counter-clockwise
rotation I'd load 0,32K in for the origin.

I see a shift problem looming. I think I see the solution, though.
Although I'm using 16 bit D/A, software limitations will limit my image
to 32K by 32K pixels. In the example above, then, for the baseline image
I'd load 16K,16K in as the upper-left origin and 48K,48K as the
lower-right end of frame. For 45 degree counter-clockwise I'd load
approximately 11K,32K in for the origin: that's 16K divided by 1.414 for
X and center of range for Y.

But perhaps you're familiar with this algorithm? Are sin, cos, or tan a
necessary part of the algorithm for arbitrary rotation? Do I really need
more that 16 bits of D/A resolution to do this right? If the rotation
angle is very small, for example such that the origin of the first line
is at ~16K,16001K and the end of the first scan line is at ~48K,16K it
seems like I might need as much as 30 or 32 bits of resolution to slowly
jack up the Y value by one pixel over the course of the scan line.

I should probably be using signed notation in these examples to make
them more clear.

Is there a way to do this in the analog domain? In actuall fact I'll
probably have 4 D/A converters total, two MDACS for the scanning and two
more driving the reference of the MDACS to set gain. Normally the gain
dacs would be loaded with the same values to give equal magnification in
x and y: in fact a single DAC could be used to set both gains but I want
to be able calibrate them. At some level it seems like I'm just changing
the "slope" of the scan lines and that perhaps only a pair of analog
multiplications is needed: perhaps the output of the X DAC is simply
multiplied by a term that varies with the degree of rotation and then
multiplied by the Y term. And vice versa?

I can't quite get my head around it, though I think I'm almost there.

Anyone know how this type of thing is done in practice, or have any
comments?

-Jeff

 

[Luhan Monat]

Hey, I would do the rotation mechanically (rotate the 2 coils). Saves a
shitload of work. This is how your TV set gets its picture straight.

--
Luhan Monat (luhanis 'at' yahoo 'dot' com)
"The future is not what it used to be..."
http://members.cox.net/berniekm

 

[Rich Grise]

On Mon, 04 Apr 2005 22:22:12 +0000, Jeff Miller wrote:

I haven't made a whole lot of progress on my electron microscope
digitizing project this month, but I hope to finish in the next 6-8
weeks. Perhaps sooner. The last decision is whether to use an MCU or DSP
in the project, or to stick with discrete logic as I'd originally planned.

To recap and give some background, the heart (arguably) of a scanning
electron microscope consists of a pair of X and Y scanning coils which
are driven in a raster-scan pattern not too unlike a television
deflection yolk. In fact NTSC scan rates, interlace and all, are
sometimes used so that the output can be displayed on a TV monitor.
....
Anyone know how this type of thing is done in practice, or have any
comments?

If you really mean "yoke", ( ;-) ), then just mount it on a bearing
so it can be twirled around the gun. Use a worm and spur gear for fine
adjust.

Or, scan orthogonally, grab the picture, and rotate it in software, like
with a paint program.

Good Luck!
Rich

 

[Dan Mills]

Jeff Miller wrote:

Hi,

I haven't made a whole lot of progress on my electron microscope
digitizing project this month, but I hope to finish in the next 6-8
weeks. Perhaps sooner. The last decision is whether to use an MCU or DSP
in the project, or to stick with discrete logic as I'd originally planned.

To recap and give some background, the heart (arguably) of a scanning
electron microscope consists of a pair of X and Y scanning coils which
are driven in a raster-scan pattern not too unlike a television
deflection yolk. In fact NTSC scan rates, interlace and all, are
sometimes used so that the output can be displayed on a TV monitor.

As it turns out, scan/image rotation is a fairly common feature of high
end scopes. I'm wondering what it would take for me to incorporate such
a feature into my project. I'm not sure whether or not this feature is
limited to digitally controlled scopes: I'd very much like to know.

I'll be driving my coils with 16 bit D/A converters: I want to
experiment with ultra-high resolution images blown up to poster size.

Of course rotating the scan in multiples of 90 degress is trivial, but
ultimately I'd like arbitrary rotation or at least in 45 degree
increments.

Get a copy of the 'Graphics gems' series of books (Andrew Glassner et al),
they discuss drawing lines at arbitary angles in discreet systems in some
depth.
IIRC "The Bresenham Line Drawing Algorithm" should turn up some
possibilities.

One possible method goes like this, Vy = Line + KVx
where Vy is the value of the Y deflection signal, Line is the scan line
number, Vx is the X deflection value and K is a constant that sets the
slope of the line (school geometry). Obviously (0,0) is the point around
which this rotation operates.
Compute K to give your rotation within the quadrant and play with the signs
and deflection drive signals to select the quadrant.
I would consider using those two extra DACs to provide a controlled bleed
from Vx->Vy and vice versa, if you have no processing power driving the
DACs, otherwise just do the math in thee processor to generate the
rotation.

There are easy matrix methods for computing the values to feed the dacs to
give rotation, scaling & skew around arbitary points, I know that they
appear in the referenced series of books but cannot be bothered to look
them up now. Your uni library will be easily able to find them.

An easy alternative of course is just to post process the video in software,
which offers some useful tricks anyway....


Regards, Dan.

 

[Rene Tschaggelar]

Jeff Miller wrote:

I'll be driving my coils with 16 bit D/A converters: I want to
experiment with ultra-high resolution images blown up to poster size.

Of course rotating the scan in multiples of 90 degress is trivial, but
ultimately I'd like arbitrary rotation or at least in 45 degree
increments. I could probably get an idea of the algorithm I'm looking
for by drawing some superimposed squares, mapping point to point and
finding the pattern. For example: if the baseline (non-rotated) image is
generated using 0,0 as the upper-left origin and proceeds right and down
as it does with traditional TV, then for a 45 degree counter-clockwise
rotation I'd load 0,32K in for the origin.

I see a shift problem looming. I think I see the solution, though.
Although I'm using 16 bit D/A, software limitations will limit my image
to 32K by 32K pixels. In the example above, then, for the baseline image
I'd load 16K,16K in as the upper-left origin and 48K,48K as the
lower-right end of frame. For 45 degree counter-clockwise I'd load
approximately 11K,32K in for the origin: that's 16K divided by 1.414 for
X and center of range for Y.

But perhaps you're familiar with this algorithm? Are sin, cos, or tan a
necessary part of the algorithm for arbitrary rotation? Do I really need
more that 16 bits of D/A resolution to do this right? If the rotation
angle is very small, for example such that the origin of the first line
is at ~16K,16001K and the end of the first scan line is at ~48K,16K it
seems like I might need as much as 30 or 32 bits of resolution to slowly
jack up the Y value by one pixel over the course of the scan line.


Anyone know how this type of thing is done in practice, or have any
comments?

To start with, you'll never get 16 bits from 16bit DACs.
When you rotate the image digitally with the rotation matrices,
you shouldn't consider more than 4k pixels per side.
This due to the limited resolution of the DAC.

Rene
--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net

 

[Jeff Miller]

Luhan Monat wrote:

Hey, I would do the rotation mechanically (rotate the 2 coils). Saves a
shitload of work. This is how your TV set gets its picture straight.

Heh heh. Well the scan coils are embedded deep in the column.

-Jeff

 

[Jeff Miller]

Rich Grise wrote:

On Mon, 04 Apr 2005 22:22:12 +0000, Jeff Miller wrote:


I haven't made a whole lot of progress on my electron microscope
digitizing project this month, but I hope to finish in the next 6-8
weeks. Perhaps sooner. The last decision is whether to use an MCU or DSP
in the project, or to stick with discrete logic as I'd originally planned.

To recap and give some background, the heart (arguably) of a scanning
electron microscope consists of a pair of X and Y scanning coils which
are driven in a raster-scan pattern not too unlike a television
deflection yolk. In fact NTSC scan rates, interlace and all, are
sometimes used so that the output can be displayed on a TV monitor.

...

Anyone know how this type of thing is done in practice, or have any
comments?


If you really mean "yoke", ( ;-) ), then just mount it on a bearing
so it can be twirled around the gun. Use a worm and spur gear for fine
adjust.

Or, scan orthogonally, grab the picture, and rotate it in software, like
with a paint program.

Good Luck!
Rich

The coils are deep inside the column and can't be moved.

Rotation after the fact is a plenty fine idea. Just about the only
reason to do it in the scope, so far as I can imagine, is to avoid
artifacts if regularities in the sample beat with the scan lines, a rare
occurance I would imagine. And maybe 'cause it looks cool in real time.
I'll be carting this scope (a Mini-SEM) around to primary schools.

-Jeff

-Jeff

 

[Jeff Miller]

This message decimated with embedded responses old-school style:

John Larkin wrote:

Jeff Miller wrote:
<snip>

Is there a way to do this in the analog domain? In actuall fact I'll
probably have 4 D/A converters total, two MDACS for the scanning and two
more driving the reference of the MDACS to set gain. Normally the gain
dacs would be loaded with the same values to give equal magnification in
x and y: in fact a single DAC could be used to set both gains but I want
to be able calibrate them. At some level it seems like I'm just changing
the "slope" of the scan lines and that perhaps only a pair of analog
multiplications is needed: perhaps the output of the X DAC is simply
multiplied by a term that varies with the degree of rotation and then
multiplied by the Y term. And vice versa?

I can't quite get my head around it, though I think I'm almost there.

Anyone know how this type of thing is done in practice, or have any
comments?

-Jeff



If the deflection coils are symmetric and can be driven with arbitrary
waveforms, and you have a pair of scan sawtooth waveforms or whatever,
you can do arbitrary rotation with four multiplying DACs and two
summing amps.

D1 = k1 * saw1 + k2 * saw2

D2 = k3 * saw1 + k4 * saw2

where D1 and D2 are the analog deflection amp inputs

Saw1 and saw2 are the old analog x, y sweep sawtooths (sawteeth?)

Mmmm hmmmm. That's about where I was headed.

K1..K4 are digital vals you load from your PC into the multiplying
DACs, with the usual sin/cos triggish stuff.

Screeeeech.... what usual sin/cos stuff? Dynamically? Or statically, ie:
knowing desired angle of rotation you calculate a pair of sin/cos values
and load them into the appropriate (static) K's and then "go".

This is cool because K1..K4 do rotation and zooming, too. You can even
sort of un-bend trapezoidal distortion, like if you're not hitting a
surface straight-on. A little dynamic focussing helps here.

Mmmm hmmm, dynamic focus is next on the agenda: I might try dynamic
astigmatism correction as well. I'm even considering "zoned" or
pixel-by-pixel (ie: digital dynamic) focus: auto-tuned by FFT'ing the
samples and optimizing for the higher frequencies, of course

There are more digital ways to do this, too, but quantization issues
can get nasty.

Linear Technology has some nice fast 4Q 16-bit mdacs.

John

I think I've heard about those linear parts.... was it you mentioned
them before?

-Jeff

 

[Daniel Haude]

["Followup-To:" header set to sci.electronics.design.]
On Tue, 05 Apr 2005 10:18:40 GMT,
Jeff Miller <cornheadorama@hotmail.com> wrote
in Msg. <4Ct4e.9428$FN4.8074@newssvr21.news.prodigy.com>

If you really mean "yoke", ( ;-) ), then just mount it on a bearing
so it can be twirled around the gun. Use a worm and spur gear for fine
adjust.

The coils are deep inside the column and can't be moved.

Rotate the sample?

But really, I'd use multiplying DACs.

--D.

 

[Jeff Miller]

This message decimated with embedded responses.

bill.sloman@ieee.org wrote:

There is scertainly a way to do it in the analog domain - the older
Cambrdige Instruments elelctron microsccopes just had four sin and cos
tapered pots on one shaft.

Hmm, I think I have one of these (two actually) in the basement, I may
take a peek. 200's or 250's.

Two of the pots were fed with the display's horizontal scan sawtooth
(X-waveform) and two with the display's vertical scan sawtooth
(Y-waveform) and the microscopes scan coils were fed with a weighed sum
of the tese waveforms, weighed according to the postion of the shaft.

A point of detail - the two ends of each of the fours pots were
respectively fed with display scan waveform and its inverse.

That's a valuable point. I was thinking I might still have to
incorporate reversal of the sense of the scanning coils: otherwise I'd
be dealing with some pretty steep slopes (read: undefined). But in fact
I don't think I'll have to electrically swap poles, just... logically
swap them, if you catch my meaning. It seems to me this inverse-drive
arrangement you describe is designed to solve the same problem: keeps
all the "slopes" resonable values between -1 and 1.

Later machines used multiplying DACs for the job, though they were a
bit dodgy for TV speed horizontal scans.

Once we went over to driving the user display from a frame store, we
gratefully stopped driving the scan coils in the microscope at TV
rates.

It took rather longer for us to go over to using DACs to drive the scan
- we definitely needed good 16-bit DACs to get a acceptable blank white
picture from the 4000 line photomonitor - even a 1/4 LSB error in the
output from a 12-bit DAC gave a very obvious line structure on a blank
white screen that customers found totally objectionable.

Hmmmm. I'm not sure I see _exactly_ what the problem might be,
especially on a blank white screen. But don't feel compelled to elaborate.

The old machines all used constant current sources driving integrators
to generate the scanning ramps, with different FET-input op amps in the
horizonal and vertical scan ramp generators. You needed a fast op amp
in the horizontal scan generator and they ran quite warm, while you
needed a low-input leakage current amplifier in the vertical scan
generator, so you wanted a low-powered amplifer that ran cool.

I think I see that. You certainly wouldn't want the input of your
vertical op-amp to suck back out a significant fraction of the constant
current you were using to fill the integrating caps. Worse still would
be an amp whose input current changed with applied voltage as well as
being a significant fraction: read vertical nonlinearity.

-----------
Bill Sloman, Nijmegen

You wouldn't happen to have schematics or at least wiring/connection
diagrams for the Stereoscans, would you? I'm starting my project with a
Mini-SEM, with an eye toward taking it around to primary schools: and I
got schematics with that scope. Secondarily I want to be able to toss
the consoles on my Stereoscans but I didn't get any docs with them.

-Jeff

 

[John Larkin]

On Tue, 05 Apr 2005 10:35:56 GMT, Jeff Miller
<cornheadorama@hotmail.com> wrote:

This message decimated with embedded responses old-school style:

John Larkin wrote:

Jeff Miller wrote:
snip
Is there a way to do this in the analog domain? In actuall fact I'll
probably have 4 D/A converters total, two MDACS for the scanning and two
more driving the reference of the MDACS to set gain. Normally the gain
dacs would be loaded with the same values to give equal magnification in
x and y: in fact a single DAC could be used to set both gains but I want
to be able calibrate them. At some level it seems like I'm just changing
the "slope" of the scan lines and that perhaps only a pair of analog
multiplications is needed: perhaps the output of the X DAC is simply
multiplied by a term that varies with the degree of rotation and then
multiplied by the Y term. And vice versa?

I can't quite get my head around it, though I think I'm almost there.

Anyone know how this type of thing is done in practice, or have any
comments?

-Jeff



If the deflection coils are symmetric and can be driven with arbitrary
waveforms, and you have a pair of scan sawtooth waveforms or whatever,
you can do arbitrary rotation with four multiplying DACs and two
summing amps.

D1 = k1 * saw1 + k2 * saw2

D2 = k3 * saw1 + k4 * saw2

where D1 and D2 are the analog deflection amp inputs

Saw1 and saw2 are the old analog x, y sweep sawtooths (sawteeth?)

Mmmm hmmmm. That's about where I was headed.

K1..K4 are digital vals you load from your PC into the multiplying
DACs, with the usual sin/cos triggish stuff.

Screeeeech.... what usual sin/cos stuff? Dynamically? Or statically, ie:
knowing desired angle of rotation you calculate a pair of sin/cos values
and load them into the appropriate (static) K's and then "go".

Screech? Staticly of course, except if you want the image to spin.

This is cool because K1..K4 do rotation and zooming, too. You can even
sort of un-bend trapezoidal distortion, like if you're not hitting a
surface straight-on. A little dynamic focussing helps here.

Mmmm hmmm, dynamic focus is next on the agenda: I might try dynamic
astigmatism correction as well. I'm even considering "zoned" or
pixel-by-pixel (ie: digital dynamic) focus: auto-tuned by FFT'ing the
samples and optimizing for the higher frequencies, of course

I was thinking once of making the saws with two more dacs, so then you
could know the addresses at all times, so then you could have a focus
lookup table. Not per pixel, but in a reasonably-sized checkerboard.

But my career in sem electronics ended at that point, about when the
lawsuits began.

John

 

[Joel Kolstad]

Check out this link: http://jmargolin.com/xy/xymon.htm

The old vector based video games (and terminals) were pretty cool --
nostalga for them has led me to California Extreme once to play some of
those old classics.

 

[Terry Given]

Joel Kolstad wrote:

Check out this link: http://jmargolin.com/xy/xymon.htm

The old vector based video games (and terminals) were pretty cool --
nostalga for them has led me to California Extreme once to play some of
those old classics.

Ah, the joy of having an Asteroids machine in the garage.....

Cheers
Terry

 

[Jeff Miller]

Gary G wrote:

On Tue, 05 Apr 2005 11:05:13 GMT, Jeff Miller
cornheadorama@hotmail.com> wrote:


What SEM are you using?

If you only have slow scan display at present, you will need TV for
fast viewing and then go to slow scan for capture. Unless you have
some compelling and exotic reason to chuck the console, I'd retain it.
It will have power supplies and balanced scan and stig coil drivers.
Also, it may have ion pump and other vacuum system controls and
circuits. If the design is fairly recent, the scan coils driver board
may already have external input bypass. The way older SEMs did scan
generation (ramps) was analog, while later and today, this is done
with DACs via MPU. Either way, the scans are synched to line
frequency.

Older SEMs are digitized today using either passive capture or active
drive, depending on the design of the SEM. Soft-Imaging makes ADDA
which does passive and active. Quartz PCI does passive, and others do
this too. These are typically 12- 14-bit capture and either SEM scan
rates or at direct digital drive via DACs. In passive capture, the
pixel dwell time is set by the SEM's scan rate. In active, it is set
by the GUI.

The main capture variables are pixels X & Y and capture depth; 8- 10-
12- 14-bits.

Gary Gaugler, Ph.D.
Microtechnics, Inc.
Granite Bay, CA 95746
916.791.8191
gary@microtechnics dot com

Hi,

I want to chuck the console(s) for three reasons. WRT my Mini-SEM, I
want to be able to bring it around to primary schools. The console is
the only thing too heavy for one person to move around. Luckily the HV
power supply is in a seperate enclosure. I don't suppose I'll actually
trash the console, just put it in long-term storage.

Secondarily, the consoles for my Stereoscans take up too much real
estate in my home lab. I only have about 300' sq. ft, and have a wide
range of interests.

Finally, I hope to find a FESEM column on Ebay that's been orphaned from
its console fairly cheap. There was one kicking around for about 6
months but I don't see it listed anymore, perhaps it was sold or the
seller gave up and scrapped it.

The specs of my design are very similar to those of the units you
mention. Mine will use 14 bit A/D converters, I think. Analog Devices
has some nice 16 bit parts just around the corner but I don't think
I'll be able to wait. My design will have up to 8192*8192 imaging native
as opposed to 4K*4K in those desings you mention. Up to 32K * 32K can be
acheived by tiling as 32K * 2K blocks.

My initial solution will cost <$1500. About half that represents a
Nationl Instruments IMAQ-1424 32-bit parallel image acquisition card
with 144 MB of on-board RAM I bought on E-bay. The AD converter
evaluation module is about $100, LVDS interface converter $50, mil-spec
circular connectors about $100, etc. Initially I'll use only 16 (14
really) bits to digitize, but there's room for adding a second
simultaneous 16 bit channel for BSE etc. as the 32 bit 1424 interface
can be configured as two 16 bit channels or "taps", they call 'em.

As I get more familiar with NI and Labview I hope to be able to
transition to a 32 bit high-speed I/O card which can be had on e-bay for
$50-$150. That, or I'll get an NI AD/DA card. At that point I'll sell
the 1424 on ebay. There's a good chance I'll be using both the 1424 and
the 32 bit I/O card together initially, the 32 bit I/O card may drive
the X and Y.

My solution will have a "universal" interface box about 12"*8"*4" with a
few BNC's on it. These will feed into a "personality module" of about
the same size for each scope I have. The personality module will have
the low voltage power supplies and probably the scan coil amps, focus
controls, etc.

About how much are those commercial solutions you mention, do you know?
Of course the webs sites don't have prices.

-Jeff

 

[Jeff Miller]

Daniel Haude wrote:

On Tue, 05 Apr 2005 10:18:40 GMT,
Jeff Miller <cornheadorama@hotmail.com> wrote
in Msg. <4Ct4e.9428$FN4.8074@newssvr21.news.prodigy.com

If you really mean "yoke", ( ;-) ), then just mount it on a bearing
so it can be twirled around the gun. Use a worm and spur gear for fine
adjust.

The coils are deep inside the column and can't be moved.


Rotate the sample?

But really, I'd use multiplying DACs.

--D.

My Mini-SEM deosn't have stage rotation.

The number of MDACS I'll be using seems to be climbing. Oh well. Looks
like I'll have to design a PC board for them as opposed to using
evaluation modules.

-Jeff

 

[John Larkin]

On Wed, 06 Apr 2005 08:11:08 GMT, Jeff Miller
<cornheadorama@hotmail.com> wrote:

Gary G wrote:
On Tue, 05 Apr 2005 11:05:13 GMT, Jeff Miller
cornheadorama@hotmail.com> wrote:


What SEM are you using?

If you only have slow scan display at present, you will need TV for
fast viewing and then go to slow scan for capture. Unless you have
some compelling and exotic reason to chuck the console, I'd retain it.
It will have power supplies and balanced scan and stig coil drivers.
Also, it may have ion pump and other vacuum system controls and
circuits. If the design is fairly recent, the scan coils driver board
may already have external input bypass. The way older SEMs did scan
generation (ramps) was analog, while later and today, this is done
with DACs via MPU. Either way, the scans are synched to line
frequency.

Older SEMs are digitized today using either passive capture or active
drive, depending on the design of the SEM. Soft-Imaging makes ADDA
which does passive and active. Quartz PCI does passive, and others do
this too. These are typically 12- 14-bit capture and either SEM scan
rates or at direct digital drive via DACs. In passive capture, the
pixel dwell time is set by the SEM's scan rate. In active, it is set
by the GUI.

The main capture variables are pixels X & Y and capture depth; 8- 10-
12- 14-bits.

Gary Gaugler, Ph.D.
Microtechnics, Inc.
Granite Bay, CA 95746
916.791.8191
gary@microtechnics dot com

Hi,

I want to chuck the console(s) for three reasons. WRT my Mini-SEM, I
want to be able to bring it around to primary schools. The console is
the only thing too heavy for one person to move around. Luckily the HV
power supply is in a seperate enclosure. I don't suppose I'll actually
trash the console, just put it in long-term storage.

Secondarily, the consoles for my Stereoscans take up too much real
estate in my home lab. I only have about 300' sq. ft, and have a wide
range of interests.

Finally, I hope to find a FESEM column on Ebay that's been orphaned from
its console fairly cheap. There was one kicking around for about 6
months but I don't see it listed anymore, perhaps it was sold or the
seller gave up and scrapped it.

The specs of my design are very similar to those of the units you
mention. Mine will use 14 bit A/D converters, I think. Analog Devices
has some nice 16 bit parts just around the corner but I don't think
I'll be able to wait. My design will have up to 8192*8192 imaging native
as opposed to 4K*4K in those desings you mention. Up to 32K * 32K can be
acheived by tiling as 32K * 2K blocks.

My initial solution will cost <$1500. About half that represents a
Nationl Instruments IMAQ-1424 32-bit parallel image acquisition card
with 144 MB of on-board RAM I bought on E-bay. The AD converter
evaluation module is about $100, LVDS interface converter $50, mil-spec
circular connectors about $100, etc. Initially I'll use only 16 (14
really) bits to digitize, but there's room for adding a second
simultaneous 16 bit channel for BSE etc. as the 32 bit 1424 interface
can be configured as two 16 bit channels or "taps", they call 'em.

As I get more familiar with NI and Labview I hope to be able to
transition to a 32 bit high-speed I/O card which can be had on e-bay for
$50-$150. That, or I'll get an NI AD/DA card. At that point I'll sell
the 1424 on ebay. There's a good chance I'll be using both the 1424 and
the 32 bit I/O card together initially, the 32 bit I/O card may drive
the X and Y.

My solution will have a "universal" interface box about 12"*8"*4" with a
few BNC's on it. These will feed into a "personality module" of about
the same size for each scope I have. The personality module will have
the low voltage power supplies and probably the scan coil amps, focus
controls, etc.

About how much are those commercial solutions you mention, do you know?
Of course the webs sites don't have prices.

-Jeff

Somebody reportedly makes a tiny (lipstick size?) SEM column that
sells for a few thousand dollars. It's just a commercial hot electron
gun, a lens element or two, and an electrostatic deflection structure,
which is just 8 round rods arranged in a circle, four at DC and four
driven with deflection voltages. One of them, one electron multiplier,
and a vacuum pump, and there you are. Oh, some electronics, too.

John

 

[Jerry Avins]

Jeff Miller wrote:

...

As it turns out, scan/image rotation is a fairly common feature of high
end scopes. I'm wondering what it would take for me to incorporate such
a feature into my project. I'm not sure whether or not this feature is
limited to digitally controlled scopes: I'd very much like to know.

The trace on my Tektronix scope is rotated (to align it with the
internal graticule) by varying the DC in an axial solenoid that the beam
passes through. I don't know how fat that technique can twist the scan
without distorting it, but if it can carry you to +/-22.5 degrees, that
might be enough.

My preferred way: The deflection currents are analog. Matrixing the
nominal X and Y drives will swing the beam to any angle you want. What
you need to do that is a pair of digital volume controls -- one fed with
X and a digital gain, the other with Y and a different digital gain --
between the DACs and each deflection driver. Computing the gains is a
one-time thing. If you don't like the math, you can use a LUT.

I'll be driving my coils with 16 bit D/A converters: I want to
experiment with ultra-high resolution images blown up to poster size.

The software to rotate the scan with the drive coils uses the same
algorithm used for drawing slanted lines on stepper-motor plotters. Once
thought through, implementation is straightforward.

Of course rotating the scan in multiples of 90 degress is trivial, but
ultimately I'd like arbitrary rotation or at least in 45 degree
increments. I could probably get an idea of the algorithm I'm looking
for by drawing some superimposed squares, mapping point to point and
finding the pattern. For example: if the baseline (non-rotated) image is
generated using 0,0 as the upper-left origin and proceeds right and down
as it does with traditional TV, then for a 45 degree counter-clockwise
rotation I'd load 0,32K in for the origin.

45 degrees is trivial, but there is a sqrt(2) scale factor that you
might need to correct.

I see a shift problem looming. I think I see the solution, though.
Although I'm using 16 bit D/A, software limitations will limit my image
to 32K by 32K pixels. In the example above, then, for the baseline image
I'd load 16K,16K in as the upper-left origin and 48K,48K as the
lower-right end of frame. For 45 degree counter-clockwise I'd load
approximately 11K,32K in for the origin: that's 16K divided by 1.414 for
X and center of range for Y.

But perhaps you're familiar with this algorithm? Are sin, cos, or tan a
necessary part of the algorithm for arbitrary rotation? Do I really need
more that 16 bits of D/A resolution to do this right? If the rotation
angle is very small, for example such that the origin of the first line
is at ~16K,16001K and the end of the first scan line is at ~48K,16K it
seems like I might need as much as 30 or 32 bits of resolution to slowly
jack up the Y value by one pixel over the course of the scan line.

Think Moire. In any case, if your step size exceeds your resolution
limit, the image will be (non alcoholically) pixellated.

I should probably be using signed notation in these examples to make
them more clear.

Clear as is.

Is there a way to do this in the analog domain? In actuall fact I'll
probably have 4 D/A converters total, two MDACS for the scanning and two
more driving the reference of the MDACS to set gain. Normally the gain
dacs would be loaded with the same values to give equal magnification in
x and y: in fact a single DAC could be used to set both gains but I want
to be able calibrate them. At some level it seems like I'm just changing
the "slope" of the scan lines and that perhaps only a pair of analog
multiplications is needed: perhaps the output of the X DAC is simply
multiplied by a term that varies with the degree of rotation and then
multiplied by the Y term. And vice versa?

See above. Great minds and all that.

I can't quite get my head around it, though I think I'm almost there.

I hope I helped.

Anyone know how this type of thing is done in practice, or have any
comments?

-Jeff

Jerry
--
Engineering is the art of making what you want from things you can get.
ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ

 

[Rene Tschaggelar]

John Larkin wrote:

On Wed, 06 Apr 2005 08:11:08 GMT, Jeff Miller
cornheadorama@hotmail.com> wrote:


My solution will have a "universal" interface box about 12"*8"*4" with a
few BNC's on it. These will feed into a "personality module" of about
the same size for each scope I have. The personality module will have
the low voltage power supplies and probably the scan coil amps, focus
controls, etc.

About how much are those commercial solutions you mention, do you know?
Of course the webs sites don't have prices.


Somebody reportedly makes a tiny (lipstick size?) SEM column that
sells for a few thousand dollars. It's just a commercial hot electron
gun, a lens element or two, and an electrostatic deflection structure,
which is just 8 round rods arranged in a circle, four at DC and four
driven with deflection voltages. One of them, one electron multiplier,
and a vacuum pump, and there you are. Oh, some electronics, too.

John,
you're simplifying it a bit. The size of the cathode
is one point that limits the resolution. Best in terms
of resolution would be a cold field emitting cathode,
but this doesn't supply a sufficiently big current.
The cathode is more than just a cathode of an electron
tube.
I admit I forgot the difference between a SEM and normal
EM in terms of the finer technical details.

Rene
--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net

 

[John Larkin]

On Wed, 06 Apr 2005 17:17:39 +0200, Rene Tschaggelar <none@none.net>
wrote:

John Larkin wrote:

On Wed, 06 Apr 2005 08:11:08 GMT, Jeff Miller
cornheadorama@hotmail.com> wrote:


My solution will have a "universal" interface box about 12"*8"*4" with a
few BNC's on it. These will feed into a "personality module" of about
the same size for each scope I have. The personality module will have
the low voltage power supplies and probably the scan coil amps, focus
controls, etc.

About how much are those commercial solutions you mention, do you know?
Of course the webs sites don't have prices.


Somebody reportedly makes a tiny (lipstick size?) SEM column that
sells for a few thousand dollars. It's just a commercial hot electron
gun, a lens element or two, and an electrostatic deflection structure,
which is just 8 round rods arranged in a circle, four at DC and four
driven with deflection voltages. One of them, one electron multiplier,
and a vacuum pump, and there you are. Oh, some electronics, too.

John,
you're simplifying it a bit. The size of the cathode
is one point that limits the resolution. Best in terms
of resolution would be a cold field emitting cathode,
but this doesn't supply a sufficiently big current.
The cathode is more than just a cathode of an electron
tube.
I admit I forgot the difference between a SEM and normal
EM in terms of the finer technical details.

You can buy replacement hot sem cathodes, lanthanum hexaboride tips or
something. They wear out, so are basicly an expendable; I don't think
they're expensive.

The cold-cathode tips only work in ultrahigh vacuum, which wouldn't be
feasible for a low-cost sem.

Classic TEMs don't scan, and have an elaborate and expensive set of
condensing and expanding magnetic lenses.

John



John

 

[Rich Grise]

On Wed, 06 Apr 2005 08:44:24 -0700, John Larkin wrote:

On Wed, 06 Apr 2005 17:17:39 +0200, Rene Tschaggelar <none@none.net
John Larkin wrote:
On Wed, 06 Apr 2005 08:11:08 GMT, Jeff Miller
My solution will have a "universal" interface box about 12"*8"*4" with
a few BNC's on it. These will feed into a "personality module" of about
the same size for each scope I have. The personality module will have
the low voltage power supplies and probably the scan coil amps, focus
controls, etc.
About how much are those commercial solutions you mention, do you know?
Of course the webs sites don't have prices.
Somebody reportedly makes a tiny (lipstick size?) SEM column that sells
for a few thousand dollars. It's just a commercial hot electron gun, a
lens element or two, and an electrostatic deflection structure, which
is just 8 round rods arranged in a circle, four at DC and four driven
with deflection voltages. One of them, one electron multiplier, and a
vacuum pump, and there you are. Oh, some electronics, too.
you're simplifying it a bit. The size of the cathode is one point that
limits the resolution. Best in terms of resolution would be a cold field
emitting cathode, but this doesn't supply a sufficiently big current. The
cathode is more than just a cathode of an electron tube.
I admit I forgot the difference between a SEM and normal EM in terms of
the finer technical details.

You can buy replacement hot sem cathodes, lanthanum hexaboride tips or
something. They wear out, so are basicly an expendable; I don't think
they're expensive.

I was once tasked to design a filament current regulator for a hot FEC.
LaB6 sounds terribly, terribly familiar, albeit this was almost 30 years
ago. I saw one in an optical microscope, and it was hard to get it to
focus properly because the light kept bouncing off the crystal facets.
They were really endeavouring to get a one-atom point. And, of course,
trying to accellerate this one-electron-wide beam to 30 KEV. I learned
a lot about Hall effect current sensors when there's a 5 or 10 KV arc
in the electron gun. When someone came up with TransZorbs, our MOV budget
hit the floor. ;-)

Cheers!
Rich