Chris, Jim, and other photoptical gadgeteers.
To make a small spot at the end of a long throw beam is very difficult. =
To a first order, the F number
of the beam has to be approximately the same as the spot size. Thus to =
get a 5 micron spot takes
an F:5 beam. That is, the diameter of the focusing optic must be 1/5 of =
the beam throw and a very special lens to get the required resolution.
To use a CCD in a practical way, the following is possible:
Take a simple hend held beam projector and use the lens from another one =
to refocus the beam.
As may be understood, the source in a laser diode is extremely small in =
one direction (it is the
same direction in which the beam is large when projected). If another =
identical lens is placed
in front of the projector, the resulting image (which will be about 1/8 =
inch away) will be similar
to that which was originally on the face of the laser chip. =20
This would mean that either the CCD, or the laser projector would need =
to be on the moving element.
If you start with only a laser diode, larger lenses could be used so =
that a mirror could be inserted
in the beam and have only the mirror moving. The f:number relationships =
mentioned above are
still important for a small spot. =20
I'd be glad to to try answer any questions regarding such systems by =
e-mail or phone. They take care in design and choice of compoonents. I =
have a couple separate laser diode collimating lenses which=20
could be made available after I confirm that they will not be wasted. =
They are very small with very
short focal lengths.
George Harris
gjharris@.............
707-751-1936
----- Original Message -----=20
From: ChrisAtUpw@..........
To: psn-l@.................
Sent: Thursday, May 31, 2001 5:38 PM
Subject: Re: Not so simple photoelectrics, or are they?
In a message dated 31/05/01, jmhannon@........ writes:=20
Having looked into building a spectrophotometer using a linerar CCD =
array, I=20
know a bit about how they work. This web site=20
=
http://www.kodak.com/cgi-bin/webCatalog.pl?product=3DKODAK+KLI+Series+Ima=
ge+Sensors=20
&cc=3DUS&lc=3Den=20
shows Kodak offerings.=20
Dear Jim Hannon,=20
Thank you for the reference. I note that the KLI-8811 has 8,800 =
pixels=20
7 x 7 microns and is a single row of cells. This might give ~ +/-12 =
bit=20
accuracy (4096) about the centre zero. The problem that is still =
puzzling me=20
is how to use this clever device to measure small enough angles and =
how to=20
keep it stable? The movement due to the six second seismic background =
is of=20
the order of 1 to 5 microns and you would normally set your A/D to =
give >128=20
counts for this.=20
Assuming that you had a small Lehman with a 1 ft beam, you need =
an=20
angular gain of x560 to get one pixel to represent 1 count. A 260 ft =
optical=20
lever which focussed to a spot of 7 microns might be a bit =
impractical. I=20
suppose that you could use two surface silvered optically flat mirror =
bars=20
and reflect the light 9 times, but I suspect that you would need a =
very=20
expensive laser. The small solid state lasers don't seem to give well =
enough=20
defined or focussed beams. Getting such a device designed, set up, =
aligned=20
and stable sounds like the sort of task that I would prefer to leave =
to=20
others. How do you plan to do it Dave?=20
Regards,=20
Chris Chapman=20
Chris, Jim, and other photoptical=20
gadgeteers.
To make a small spot at the end of a =
long throw=20
beam is very difficult. To a first order, the F =
number
of the beam has to be approximately the =
same as=20
the spot size. Thus to get a 5 micron spot takes
an F:5 beam. That is, the =
diameter of the=20
focusing optic must be 1/5 of the beam throw and a very special lens to =
get the=20
required resolution.
To use a CCD in a practical way, the =
following is=20
possible:
Take a simple hend held beam projector =
and use the=20
lens from another one to refocus the beam.
As may be understood, the source in a =
laser diode=20
is extremely small in one direction (it is the
same direction in which the beam is =
large when=20
projected). If another identical lens is placed
in front of the projector, the =
resulting image=20
(which will be about 1/8 inch away) will be similar
to that which was originally on the =
face of the=20
laser chip.
This would mean that either the CCD, or =
the laser=20
projector would need to be on the moving element.
If you start with only a laser diode, =
larger lenses=20
could be used so that a mirror could be inserted
in the beam and have only the mirror =
moving. =20
The f:number relationships mentioned above are
still important for a small spot. =
I'd be glad to to try answer any =
questions=20
regarding such systems by e-mail or phone. They take care in =
design and choice of compoonents. I have =
a couple=20
separate laser diode collimating lenses which
could be made available after I confirm =
that they=20
will not be wasted. They are very small with very
short focal lengths.
George Harris
gjharris@............. <=
/DIV>
707-751-1936
----- Original Message -----
From:=20
ChrisAtUpw@.......
Sent: Thursday, May 31, 2001 =
5:38=20
PM
Subject: Re: Not so simple=20
photoelectrics, or are they?
In a =
message dated=20
31/05/01, jmhannon@........ =
writes:=20
Having looked into building a spectrophotometer using a =
linerar=20
CCD array, I
know a bit about how they work. This web site=20
http://www.kodak.com/cgi-bin/webCatalog.pl?product=3D=
KODAK+KLI+Series+Image+Sensors=20
&cc=3DUS&lc=3Den
shows Kodak offerings.
Dear Jim Hannon,=20
Thank you for the =
reference. I=20
note that the KLI-8811 has 8,800 pixels
7 x 7 microns and is a =
single row=20
of cells. This might give ~ +/-12 bit
accuracy (4096) about the =
centre=20
zero. The problem that is still puzzling me
is how to use this =
clever=20
device to measure small enough angles and how to
keep it stable? =
The=20
movement due to the six second seismic background is of
the order =
of 1 to=20
5 microns and you would normally set your A/D to give >128 =
counts for=20
this.
Assuming that you =
had a=20
small Lehman with a 1 ft beam, you need an
angular gain of x560 to =
get one=20
pixel to represent 1 count. A 260 ft optical
lever which focussed =
to a=20
spot of 7 microns might be a bit impractical. I
suppose that you =
could use=20
two surface silvered optically flat mirror bars
and reflect the =
light 9=20
times, but I suspect that you would need a very
expensive laser. =
The small=20
solid state lasers don't seem to give well enough
defined or =
focussed beams. Getting such a device designed, set up, aligned =
and stable=20
sounds like the sort of task that I would prefer to leave to =
others. How=20
do you plan to do it Dave?=20
Regards,=20
Chris Chapman=20
Larry Cochrane <cochrane@..............>