Loading ...
Sorry, an error occurred while loading the content.

Re: Freedom by degrees

Expand Messages
  • a.johnw
    A fixed primary and the degrees of freedom of the 2ndry miss an important adjustment. The focal plain of the mirror must be square to the axis of the focuser
    Message 1 of 16 , Jun 2, 2008
    • 0 Attachment
      A fixed primary and the degrees of freedom of the 2ndry miss an
      important adjustment. The focal plain of the mirror must be square to
      the axis of the focuser so that the focal plain of the eyepiece or
      camera etc is also square. This would be even more important if a coma
      corrector was used on the scope and in any case is far more critical
      on faster scopes.

      You could have a fixed main mirror but the focuser would have to be
      adjustable for tilt in 2 directions.

      Noting the comment about offset 2ndry's that goes back to an article
      in sky and telescope some where round the 70's. I personally think it
      introduces more problems than it's worth. Commercial examples I've
      owned bear this out too. Some not only offset the 2ndry but also alter
      it's tilt. That means that the whole thing is very difficult to set up
      correctly especially in respect to ensuring the focal plain is square
      to the focuser. Offsetting just isn't worth it for the small gain it
      gives and probable losses in other parts of the beam.

      Beware of setting up skywatcher reflectors and the solution to a meade
      I own was to re machine the 2ndry holder to 45 degrees. The previous
      owner of that scope finished up only using 2/3rds of the main mirror
      via normal collimation methods with the usual type of laser. That's
      one of the reasons I much prefer the reflection method and the aid of
      something with a small hole in it that fits in the focuser. It's then
      easy to centralize the 2ndry in the focuser and the 2ndry in the tube
      before following the usual collimation techniques.

      John


      --- In atm_free@yahoogroups.com, "Nils Olof Carlin"
      <nilsolof.carlin@...> wrote:
      >
      > Ben,
      >
      > if I understand you correctly, you offset the primary towards the
      focuser
      > side of the tube, to achieve much the same thing as the more (but
      not very)
      > common offsetting the diagonal mirror away from the focuser - namely
      making
      > the optical axis parallel to the tube's center axis. It will be
      offset a
      > short distance, as will the whole light path to be unobstructed, but
      this is
      > not really a problem (if so, it would be vignetting by a too narrow
      outer
      > aperture, but this is similar in both cases).
      >
      > The problem I was discussing was not a primary offset toward the
      focuser,
      > but sideways (as seen from the focuser), so that the focuser axis and
      > primary axis do not even lie in the same plane. In this situation,
      you need
      > to tilt or move either the primary or the focuser to bring the axes
      to a
      > common plane. Only then can you finish the collimation by adjusting the
      > secondary.
      >
      > But my reply was about the comment:
      > "The secondary has up to 6 degrees of freedom that can be used to
      eliminate
      > all primary mirror adjustments."
      >
      > To collimate a newtonian optically, there are 3 operations to be done:
      > centering the focuser axis on the primary, centering the primary's
      axis in
      > the focuser, and centering the secondary to center the cole of fully
      > illuminated FOV within the focuser.
      >
      > One way is to start with the focuser axis as reference. First optically
      > center the secondary (which is offseting it mechanically), next tilt
      the
      > secondary to reflect the focuser axis to the center of the primary,
      and last
      > tilt the primary to center its optical axis in the focuser. This is a
      > well-established way, and fairly straightforward with most telescope
      designs
      > and available tools (as the focuser often is not adjustable).
      >
      > The other way would be starting with a fixed primary, avoiding the
      > complications of collimating bolts to carry a heavy primary at varying
      > altitude. First you would position the secondary to get the cone of
      full
      > illumination on axis - I can't think of an easy way of doing this.
      Then you
      > need to tilt the secondary to center its optical axis at the focal
      plane -
      > as easy with a Cheshire as above. Last, you need to tilt the focuser
      to make
      > its axis reflect to the center of the primary. Easy with a laser,
      but only
      > if the focuser is indeed adjustable in tilt. Thus, the above
      statement seems
      > far from complete.
      >
      > Nils Olof
      >
      >
      > > If I shorten one of the posts and shim the other two which mount my
      > > mirror cell to the tube I can offset the primary. The offset is
      > > toward the side of the focuser because the diagonal intersects a cone
      > > of light at 45 degrees. The ellipse figure of that causes the offset
      > > because of the coverging light cone. The cone is wider where the
      > > diagonal is closer to the primary.
      > >
      > > When I look into the focuser the diagonal should sppear centered and
      > > the mirror also. Does this satisfy the alignment of the two axes?
      > >
      > > I have uploaded a photo of my cell to album bekirwicz.
      >
    • Nils Olof Carlin
      John, ... A section (right) through a cone *is* elliptical (a bit counter-intuitive to me as to others, but the old greeks knew it well), so there s no
      Message 2 of 16 , Jun 4, 2008
      • 0 Attachment
        John,


        > The basic idea comes about because the light cone of the mirror is a
        > larger diameter as it gets nearer and nearer the mirror. The top of
        > the 2ndry is furthest from the mirror and the bottom is nearest. I
        > think it was a bit of a simplistic view but the article claimed that
        > the 2ndry should be moved away from the focuser to counter react the
        > increased beam diameter at the lower parts of the 2ndry mirror
        > especially as the light that was missed by not doing this came from
        > the edges of the mirror. I could see sense in reasoning that the 2ndry
        > should not have the usual elliptical shape but should be a section
        > through a cone but I can't see much can be gained by just offsetting
        > it.

        A section (right) through a cone *is* elliptical (a bit counter-intuitive to
        me as to others, but the old greeks knew it well), so there's no problem. If
        you place the elliptic diagonal mirror so that its *outline* is centered as
        seen in a sight tube in the focuser, there is no way you can miss any light
        if the collimation of the optical axes is correct. However, to center the
        *outline* of the diagonal means you have to offset the geometric center of
        the ellipse *down* the tube relative to the focuser's axis - moving it away
        from the focuser is done for another reason!

        Collimation done and the focuser's and primary's axes are merged into one,
        this optical axis will be offset relative to the cylinder that the diagonal
        was "cut" from.
        This leaves you two basic options: leave the diagonal mirror centered in the
        tube and have the optical axis offset towards the focuser (it is centered on
        the primary mirror, but you will have tilted both the primary and diagonal
        mirrors as needed) - or offset the diagonal away from the focuser and have
        the axis centered.

        The practical consequences are small - typically, the offset is something
        like 1% of the primary's diameter, and if ignored, the tilt of the optical
        axis is perhaps 1:300, or something like 0.2 deg - hardly worth worrying
        about, since you always adjust the correct tilt during collimation (you'll
        never know if it 45.0 or actually 44.9 deg to the tube).

        well, never mind - use a combination sight tube/cheshire and you'll be close
        enough. I kind of found the geometry fascinating enough, but YMMV.

        Nils Olof
      • a.johnw
        ... counter-intuitive to ... problem. If ... centered as ... any light ... center the ... center of ... it away ... into one, ... diagonal ... centered in the
        Message 3 of 16 , Jun 5, 2008
        • 0 Attachment
          --- In atm_free@yahoogroups.com, "Nils Olof Carlin"
          <nilsolof.carlin@...> wrote:
          >
          > John,
          >
          >
          > > The basic idea comes about because the light cone of the mirror is a
          > > larger diameter as it gets nearer and nearer the mirror. The top of
          > > the 2ndry is furthest from the mirror and the bottom is nearest. I
          > > think it was a bit of a simplistic view but the article claimed that
          > > the 2ndry should be moved away from the focuser to counter react the
          > > increased beam diameter at the lower parts of the 2ndry mirror
          > > especially as the light that was missed by not doing this came from
          > > the edges of the mirror. I could see sense in reasoning that the 2ndry
          > > should not have the usual elliptical shape but should be a section
          > > through a cone but I can't see much can be gained by just offsetting
          > > it.
          >
          > A section (right) through a cone *is* elliptical (a bit
          counter-intuitive to
          > me as to others, but the old greeks knew it well), so there's no
          problem. If
          > you place the elliptic diagonal mirror so that its *outline* is
          centered as
          > seen in a sight tube in the focuser, there is no way you can miss
          any light
          > if the collimation of the optical axes is correct. However, to
          center the
          > *outline* of the diagonal means you have to offset the geometric
          center of
          > the ellipse *down* the tube relative to the focuser's axis - moving
          it away
          > from the focuser is done for another reason!
          >
          > Collimation done and the focuser's and primary's axes are merged
          into one,
          > this optical axis will be offset relative to the cylinder that the
          diagonal
          > was "cut" from.
          > This leaves you two basic options: leave the diagonal mirror
          centered in the
          > tube and have the optical axis offset towards the focuser (it is
          centered on
          > the primary mirror, but you will have tilted both the primary and
          diagonal
          > mirrors as needed) - or offset the diagonal away from the focuser
          and have
          > the axis centered.
          >
          > The practical consequences are small - typically, the offset is
          something
          > like 1% of the primary's diameter, and if ignored, the tilt of the
          optical
          > axis is perhaps 1:300, or something like 0.2 deg - hardly worth
          worrying
          > about, since you always adjust the correct tilt during collimation
          (you'll
          > never know if it 45.0 or actually 44.9 deg to the tube).
          >
          > well, never mind - use a combination sight tube/cheshire and you'll
          be close
          > enough. I kind of found the geometry fascinating enough, but YMMV.
          >
          > Nils Olof
          >
          I may be being thick Nils but if what you say is correct - that the
          section through a plain cylinder is the same as that through a cone
          there wouldn't be any sense in ofseting it. Actually I'm sure it isn't
          as that is how conic sections are formed.

          John
        • bekirwicz
          Group, I remember it because it is counter-intuitive. Except that a cylinder may be considered a short section of an infinitely steep cone? The drawings at the
          Message 4 of 16 , Jun 6, 2008
          • 0 Attachment
            Group,

            I remember it because it is counter-intuitive. Except that a cylinder
            may be considered a short section of an infinitely steep cone?

            The drawings at the link should help.
            http://mathworld.wolfram.com/ConicSection.html

            Get some paper, a protractor and a ruler. Some experience with
            mechanical drawing helps.

            Draw a steep cone. Also draw the axis of the cone.
            Intersect the cone perpendicular to the axis of the cone. That's the
            circle. Intersect it at 45 angle for the ellipse. We don't need to
            draw the full length to the mirror. That is way off the paper.
            We're only concerned with the conical section from just below
            (objective end) the diagonal intersection.

            The ellipse misses part of the cone opposite the focuser.

            A small offset from center of the primary will correct this. It
            is probably not detectable to the eye in a telescope. For undersized
            diagonals it may not be meaningful. An oversized diagonal would
            collect the light except that it blocked it! The offset is meaningful
            only for an optimum diagonal.

            Ben Waranowitz

            --- In atm_free@yahoogroups.com, "a.johnw" <a.johnw@...> wrote:
            >
            > --- In atm_free@yahoogroups.com, "Nils Olof Carlin"
            > <nilsolof.carlin@> wrote:
            > >
            > > John,
            > >
            > >
            > > > The basic idea comes about because the light cone of the mirror
            is a
            > > > larger diameter as it gets nearer and nearer the mirror. The
            top of
            > > > the 2ndry is furthest from the mirror and the bottom is
            nearest. I
            > > > think it was a bit of a simplistic view but the article claimed
            that
            > > > the 2ndry should be moved away from the focuser to counter
            react the
            > > > increased beam diameter at the lower parts of the 2ndry mirror
            > > > especially as the light that was missed by not doing this came
            from
            > > > the edges of the mirror. I could see sense in reasoning that
            the 2ndry
            > > > should not have the usual elliptical shape but should be a
            section
            > > > through a cone but I can't see much can be gained by just
            offsetting
            > > > it.
            > >
            > > A section (right) through a cone *is* elliptical (a bit
            > counter-intuitive to
            > > me as to others, but the old greeks knew it well), so there's no
            > problem. If
            > > you place the elliptic diagonal mirror so that its *outline* is
            > centered as
            > > seen in a sight tube in the focuser, there is no way you can miss
            > any light
            > > if the collimation of the optical axes is correct. However, to
            > center the
            > > *outline* of the diagonal means you have to offset the geometric
            > center of
            > > the ellipse *down* the tube relative to the focuser's axis -
            moving
            > it away
            > > from the focuser is done for another reason!
            > >
            > > Collimation done and the focuser's and primary's axes are merged
            > into one,
            > > this optical axis will be offset relative to the cylinder that the
            > diagonal
            > > was "cut" from.
            > > This leaves you two basic options: leave the diagonal mirror
            > centered in the
            > > tube and have the optical axis offset towards the focuser (it is
            > centered on
            > > the primary mirror, but you will have tilted both the primary and
            > diagonal
            > > mirrors as needed) - or offset the diagonal away from the focuser
            > and have
            > > the axis centered.
            > >
            > > The practical consequences are small - typically, the offset is
            > something
            > > like 1% of the primary's diameter, and if ignored, the tilt of the
            > optical
            > > axis is perhaps 1:300, or something like 0.2 deg - hardly worth
            > worrying
            > > about, since you always adjust the correct tilt during collimation
            > (you'll
            > > never know if it 45.0 or actually 44.9 deg to the tube).
            > >
            > > well, never mind - use a combination sight tube/cheshire and
            you'll
            > be close
            > > enough. I kind of found the geometry fascinating enough, but YMMV.
            > >
            > > Nils Olof
            > >
            > I may be being thick Nils but if what you say is correct - that the
            > section through a plain cylinder is the same as that through a cone
            > there wouldn't be any sense in ofseting it. Actually I'm sure it
            isn't
            > as that is how conic sections are formed.
            >
            > John
            >
          • bekirwicz
            Some Offset Newtonian Diagonal links. http://www.fpi-protostar.com/bgreer/collim.htm http://www.rfroyce.com/diagsize.htm
            Message 5 of 16 , Jun 6, 2008
            • 0 Attachment
            • Mel Bartels
              ... A small offset from center of the primary will correct this. It is probably not detectable to the eye in a telescope. For undersized diagonals it may not
              Message 6 of 16 , Jun 6, 2008
              • 0 Attachment
                >>>
                A small offset from center of the primary will correct this. It
                is probably not detectable to the eye in a telescope. For undersized
                diagonals it may not be meaningful. An oversized diagonal would
                collect the light except that it blocked it! The offset is meaningful
                only for an optimum diagonal.
                <<<

                And it is also meaningful for pointing the telescope. Without the offset,
                the mechanical axis and the optical axis are not coincident, leading to
                pointing errors. A fast medium scope (12 inch /4) without offset can error
                up to 1/4 degree in pointing.

                Mel Bartels
              • bekirwicz
                Mel, Maybe I am missing something as I do not see how offsetting the primary or not offsetting it causes pointing error. I guess I ll find out. I want to try
                Message 7 of 16 , Jun 6, 2008
                • 0 Attachment
                  Mel,

                  Maybe I am missing something as I do not see how offsetting the
                  primary or not offsetting it causes pointing error. I guess I'll find
                  out. I want to try the offset for as long as I have a secondary
                  obstruction I want to use all of the secondary.

                  I have not figured the amount of cell offset yet but I guess it's in
                  the ball park of 1mm. At F:5.6 one of the links I earlier posted
                  suggests it's not enough to worry about. I tend to agree but am
                  interested in trying as an ATM. By shifting the cell I will keep the
                  parabola axis still parallel to the Sonotube axis. I don't see where
                  it affects pointing. The focuser axis and the primary axis will
                  intersect and be at a 90 degree angle, or pretty close.

                  I intend to replace the Sonotube someday. I have a dobsonian mount I
                  got free but incomplete. In the next year I will have the time to
                  figure out the final configuration. Thinking of a truss tube to
                  install on a good GEM, like the Orion Atlas GoTo.

                  Ben

                  --- In atm_free@yahoogroups.com, "Mel Bartels" <mbartels@...> wrote:
                  >
                  > >>>
                  > A small offset from center of the primary will correct this. It
                  > is probably not detectable to the eye in a telescope. For
                  undersized
                  > diagonals it may not be meaningful. An oversized diagonal would
                  > collect the light except that it blocked it! The offset is
                  meaningful
                  > only for an optimum diagonal.
                  > <<<
                  >
                  > And it is also meaningful for pointing the telescope. Without the
                  offset,
                  > the mechanical axis and the optical axis are not coincident,
                  leading to
                  > pointing errors. A fast medium scope (12 inch /4) without offset
                  can error
                  > up to 1/4 degree in pointing.
                  >
                  > Mel Bartels
                  >
                • Mel Bartels
                  ... Maybe I am missing something as I do not see how offsetting the primary or not offsetting it causes pointing error. I guess I ll find
                  Message 8 of 16 , Jun 6, 2008
                  • 0 Attachment
                    >>>
                    Maybe I am missing something as I do not see how offsetting the
                    primary or not offsetting it causes pointing error. I guess I'll find
                    <<<

                    If the primary is offset and the secondary is not, then there is no
                    additional mis-alignment error introduced into your pointing, even though
                    the tube assembly axis and optical axis are not coincident (they are
                    parallel in 3D). Did I not quite understand your setup?

                    Best is primary centered and diagonal offset, as this makes optical and tube
                    axis coincident. Then there's no issue if you have a rotatable tube that's
                    slightly off with respect to its cradle or holder.

                    Mel Bartels
                  • Nils Olof Carlin
                    The placement of the diagonal is determined by two independent considerations: The first (and most important) is centering the fully illuminated field within
                    Message 9 of 16 , Jun 7, 2008
                    • 0 Attachment
                      The placement of the diagonal is determined by two independent
                      considerations:

                      The first (and most important) is centering the fully illuminated
                      field within the focuser. This depends on the placement up-down the
                      tube, and can be conveniently done with a sight tube, centering the
                      outline of the mirror in it.

                      The second, independent constraint is centering the optical axis in
                      the tube, or at least making it parallel to the tube. For most dob-type
                      telescopes, this isn't really necessary - you set the finder to point
                      in the same direction. However, if you use computerized aiming and the
                      software assumes that the optical axis is exactly perpendicular to the
                      altitude axis, though it is in reality 1/4 deg off, there will be a
                      possibly significant pointing error. But here, the important factor is
                      *not* having the axis parallel to the tube, but rather perpendicular to
                      the alt axis (or with EQ mounts, the dec axis), so in order to ensure
                      perpendicularity, the placement of the trunnions is even more
                      critical!!

                      To do this, the diagonal mirror (or better the cylinder that it is
                      "cut" from), must be offset away from the optical axis (of the primary,
                      but of course from the combined axis after collimation). Whether you do
                      this by moving the primary towards the focuser side of the tube, or the
                      diagonal mirror away from it, is immaterial.

                      I would welcome a change in terminology - the geometry is difficult
                      enough to understand (as we are witnessing) when using the term
                      "offset" for both. I would prefer to call the first "optical centering"
                      as a better descriptive term, and perhaps "axial centering" for the
                      second.

                      I've said it before: if you do the optical centering, and then
                      collimate the optical axes, there is no way you can lose light, or get
                      a focal plane that is not perpendicular to the focuser.

                      Nils Olof


                      Maybe I am missing something as I do not see how offsetting the
                      primary or not offsetting it causes pointing error. I guess I'll find
                      out. I want to try the offset for as long as I have a secondary
                      obstruction I want to use all of the secondary.

                      I have not figured the amount of cell offset yet but I guess it's in
                      the ball park of 1mm. At F:5.6 one of the links I earlier posted
                      suggests it's not enough to worry about. I tend to agree but am
                      interested in trying as an ATM. By shifting the cell I will keep the
                      parabola axis still parallel to the Sonotube axis. I don't see where
                      it affects pointing. The focuser axis and the primary axis will
                      intersect and be at a 90 degree angle, or pretty close.

                      I intend to replace the Sonotube someday. I have a dobsonian mount I
                      got free but incomplete. In the next year I will have the time to
                      figure out the final configuration. Thinking of a truss tube to
                      install on a good GEM, like the Orion Atlas GoTo.

                      Ben

                      --- In atm_free@yahoogroups.com, "Mel Bartels" <mbartels@...> wrote:
                      >
                      > >>>
                      > A small offset from center of the primary will correct this. It
                      > is probably not detectable to the eye in a telescope. For
                      undersized
                      > diagonals it may not be meaningful. An oversized diagonal would
                      > collect the light except that it blocked it! The offset is
                      meaningful
                      > only for an optimum diagonal.
                      > <<<
                      >
                      > And it is also meaningful for pointing the telescope. Without the
                      offset,
                      > the mechanical axis and the optical axis are not coincident,
                      leading to
                      > pointing errors. A fast medium scope (12 inch /4) without offset
                      can error
                      > up to 1/4 degree in pointing.
                      >
                      > Mel Bartels
                      >
                    • a.johnw
                      ... You can work out the offset needed. There is a cone of light coming of the mirror and the major axis of your 2ndry. You just have to centralize that in the
                      Message 10 of 16 , Jun 7, 2008
                      • 0 Attachment
                        --- In atm_free@yahoogroups.com, "bekirwicz" <b3kirwiz@...> wrote:
                        >
                        > Mel,
                        You can work out the offset needed. There is a cone of light coming of
                        the mirror and the major axis of your 2ndry. You just have to
                        centralize that in the cone off the mirror. The offset depends on the
                        F ratio. In principle it's needed on all f ratios.

                        Me I don't think it's worth it but if you have a scope that uses it
                        then it may well be worth sticking with it especially if it's like
                        that meade that I mentioned. (10 f4.3 sonotube equatorial - it's not
                        the tube that's the problem either.)

                        John

                        >
                        > Maybe I am missing something as I do not see how offsetting the
                        > primary or not offsetting it causes pointing error. I guess I'll find
                        > out. I want to try the offset for as long as I have a secondary
                        > obstruction I want to use all of the secondary.
                        >
                        > I have not figured the amount of cell offset yet but I guess it's in
                        > the ball park of 1mm. At F:5.6 one of the links I earlier posted
                        > suggests it's not enough to worry about. I tend to agree but am
                        > interested in trying as an ATM. By shifting the cell I will keep the
                        > parabola axis still parallel to the Sonotube axis. I don't see where
                        > it affects pointing. The focuser axis and the primary axis will
                        > intersect and be at a 90 degree angle, or pretty close.
                        >
                        > I intend to replace the Sonotube someday. I have a dobsonian mount I
                        > got free but incomplete. In the next year I will have the time to
                        > figure out the final configuration. Thinking of a truss tube to
                        > install on a good GEM, like the Orion Atlas GoTo.
                        >
                        > Ben
                        >
                        > --- In atm_free@yahoogroups.com, "Mel Bartels" <mbartels@> wrote:
                        > >
                        > > >>>
                        > > A small offset from center of the primary will correct this. It
                        > > is probably not detectable to the eye in a telescope. For
                        > undersized
                        > > diagonals it may not be meaningful. An oversized diagonal would
                        > > collect the light except that it blocked it! The offset is
                        > meaningful
                        > > only for an optimum diagonal.
                        > > <<<
                        > >
                        > > And it is also meaningful for pointing the telescope. Without the
                        > offset,
                        > > the mechanical axis and the optical axis are not coincident,
                        > leading to
                        > > pointing errors. A fast medium scope (12 inch /4) without offset
                        > can error
                        > > up to 1/4 degree in pointing.
                        > >
                        > > Mel Bartels
                        > >
                        >
                      • bekirwicz
                        Mel, Thanks. After this suggestion triggered cogitation I will keep the mirror centered. I have to cut a new hole for the diagonal holder anyway. I can mount
                        Message 11 of 16 , Jun 7, 2008
                        • 0 Attachment
                          Mel,

                          Thanks. After this suggestion triggered cogitation I will keep
                          the mirror centered. I have to cut a new hole for the diagonal holder
                          anyway. I can mount the 4-vane spider centered. Elongated holes make
                          it adjustable to the offset. Less work and if I get a new truss tube
                          and primary cell I can sell the present 9-pt unmodified. I'm set.

                          Ben

                          --- In atm_free@yahoogroups.com, "Mel Bartels" <mbartels@...> wrote:
                          >
                          > >>>
                          > Maybe I am missing something as I do not see how offsetting the
                          > primary or not offsetting it causes pointing error. I guess I'll
                          find
                          > <<<
                          >
                          > If the primary is offset and the secondary is not, then there is no
                          > additional mis-alignment error introduced into your pointing, even
                          though
                          > the tube assembly axis and optical axis are not coincident (they are
                          > parallel in 3D). Did I not quite understand your setup?
                          >
                          > Best is primary centered and diagonal offset, as this makes optical
                          and tube
                          > axis coincident. Then there's no issue if you have a rotatable
                          tube that's
                          > slightly off with respect to its cradle or holder.
                          >
                          > Mel Bartels
                          >
                        Your message has been successfully submitted and would be delivered to recipients shortly.