Optics Engineering

[kevc70]

Hello,

I have an application where I need to linearly move a focusing lens (1mm diameter, 37.5 mm long) about the x, y, and z axis, as well as provide rotational freedom within a tube with an ID of 37.5mm (same as focusing lens length). I've looked at all the optics companies websites and they offer combination linear/rotational units, but at very steep prices. I was wondering if anyone knew of an inexpensive means of performing the above, possibly through the use of worm gearing, rack and pinion, etc. There needs to be fine adjustment with this, so that the user can lock in the correct focal length when the focusing lens is in the correct position. Any help on this would be greatly appreciated.

Kevin

 

[lumenharold]

Not to be second guessing you but I am curious why you would want so many degrees of freedom for a focus lens. Typically a set up like you are talking about (as I understand it) would be for a test bench set-up. When we provide a focusing mechanism in a lens assembly we want to be sure the lens moves along the optical axis (z) without rotation and with minimal deviation (x and y). that way the system doesn't need to be re-aligned every time the lens moves.

Most lens adjusters like in a telephoto will have a triple start thread so that the lens will move the full travel with one rotation of the adjuster ring. Will the client be setting the focus once and leaving it or will they need to have adjustment available continuously?

Harold
SW2009 SP4.0 OPW2009 SP2 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[IRstuff]

> Please refrain from double posting

> What resolution and accuracy do you need?

TTFN

FAQ731-376

 

[kevc70]

For our application, the focusing lens (which we buy from Doric Lenses) has to be 25.4 (1") from our LED source. From the focusing lens, the image bounces off a mirror and through a 6' focal length projection lens. If we have to project at a distance at, say 8', then we want to be able to move the focusing lens to allow for this. This movement has to be in the z-axis (to and away from us). Also, if the focusing lens is slightly skewed in an angular fashion, then we need to be able to rotate the lens such that it lines up perfectly with the LED source. To adjust it axially is important as well, so that the lens can again line up with the LED source if it is slightly off the LED center. Hopefully, this would be set at the factory, but there may be an instance where this needs to be set again by the field service person upon installation. The accuracy needed would be 1mm increments for axial movement, and .5 degree for angular. This is where having a fine adjustment is so important. Our requirements specify coarse and fine adjustments. For our projection lens, we are looking for axial movement (up and down from the ceiling), also in a fine adjustment manner. Any suggestions would be greatly appreciated.

Kevin

 

[unclesyd]

You might find what you are looking for at Edmund Optics. They were very forthcoming with technical assistance.

http://www.edmundoptics.com/onlinecatalog/displayprod

 

[lumenharold]

On the one hand you are trying to contain cost by not going with off-the-shelf high precision positioners, and on the other hand you are building several adjustments into your product to compensate for alignment. Seems counter productive to me.

We keep cost down by designing the fixturing (adjustments) out of the product. If this is an installation that won't be re-adjusted after set-up then the lens can be positioned at the factory so that it is not skewed. The key will be keeping the optical center (as opposed to the mechanical center) aligned with the system axis as it moves. That is why adustable lens assemblies travel without rotation. We typically design custom dedicated mounts for optics instead of buying catalog adjusters. The exception being for test benches or fixturing.

To keep the lens on axis, mount and secure it within the lens tube using tangent locating surfaces instead of the O.D.. If the system is super sensitive you might need to align the lens and then glue it in place.

To keep the assembly on axis as it moves you might consider using a slip fit with a sleeve that is split so that it can be locked down after adjustment. We have also had good results using a "V" groove locator that is constructed by having one off axis bore that is slightly larger than the lens mount and another off axis radius that creates two lines of contact. Lock the lens assembly in place with a set screw. The better your pre-alignment and lens positioning the less impact lens adijustment will have on the final alignment.

Harold
SW2009 SP4.0 OPW2009 SP2 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[IRstuff]

I completely agree with Harold.

Nearly everything can be compensated for by design.

While doing so might appear to more expensive, up front, the apparent savings is vaporous, and you'll spend far more in technician and customer time unnecessarily tweaking and adjusting in the field, where you cannot have anywhere near the control and instrumentation you have in the factory.

If the focus point is changing during adjustment, it's a problem in the construction of the entire optical train, so a well-designed assembly process will minimize that; we've designed systems where the repeatability of boresight over the focus range is way better than 50 microradians, and we've even purchased off-the-shelf optics that can achieve better than 150 microradians of boresight retention.

TTFN

FAQ731-376

 

[kevc70]

It may get to the point where we do end up buying an off-the shelf positioner to suit our needs.....it's just that the space envelope with which to work is very small and the positioners from edmunds, thorlabs, and newport are a tad bit oversized for our application.

Kevin

 

[unclesyd]

I was thinking about a tube within a tube with helical slots that would cause the one tube to move in two directions by simply twisting one tube or the other.

Also have you looked at helical focusing elements for cameras. One of many at SK Grimes. There other components used on large format cameras that might be interest, like the rack and pinion focusing mechanism and the element that will allow you to skew, for lack of a better word, the image.

http://www.skgrimes.com/thisweek/1_5/index.htm

 

[kevc70]

hmmmm, I like the idea of using a tube within a tube. maybe having a dowel pin lever mounted on the inside tube that moves along an oblong slot in the outer tube. This slot would be slanted so as to cause the inner tube to move in a rotational fashion. The other tube would have a horizontal slot with another dowel pin lever to cause movement in the z-axis. Is this the concept you were thinking about? What I like about it is that it is inexpensive and easy to manufacture. Now, the next challenge is to incorporate a fine adjustment onto the levers. I was thinking of having a return spring mounted on the other of the dowel lever, and having a set screw on the opposite side that can act against the dowel, with the spring providing resistance and fine adjustment.

Kevin

 

[unclesyd]

There are several way this could work including the description you posted. The one that I've worked with was just two telescoping tubes with the helical grooves in the outer tube and cam on the inner tube. Turing the inner tube cause it to move. I'm trying to recall how the outer tube was fixed in place and the helix angle of the grooves.

Another way would be to use a worm and pawl arrangement, like on the older level wind fishing reels to drive the tube. There would be a little lost motion but not if you were driving in one direction or the lost motion could be compensated for as you mention with a spring.

 

[lumenharold]

Kevin,

Are you sure about those lens dimensions? 1mm diameter (.039") is pretty small to be 25.4mm (1.0") from the LED source. The only lens I could imagine Doric providing in that small of diameter would be a ball lens. By any chance would the diameter be the 37.5mm dimension? Unless you have a collimating optic on the LED it seems like you would lose a lot of the output by squeezing it through that small of an aperture and the output would change as the lens is repositioned.

When IRStuff is asking about resolution and accuracy I assume he is refering to the system sensitivity and that would be driven by the lens specifications.

Positioners like the assembly from SK Grimes (nice link unclesyd!) are the tube in a tube with a thread you are considering. The fixed tube is typically the "C" Mount which will have slots cut into it; the lens tube will be mounted in a threaded tube that has material removed to match the slots in the "C" Mount; the threads are the triple (or more) start threads that I mentioned previously; the female threads are cut into an outer sleeve that when turned will move the lens assembly. Take a look at Schneider optics "C" Mount lenses for an idea on how the lock the focus after adjustment.

The accuracy requirements (system sensitivity) will drive the machining tolerances for the sleeves and tubes, etc.

For extremely sensitive systems we have used guide rods instead of the slots but that requires more space and tight tolerances. To get the data on the triple start thread I use the thread calculator program ME ThreadPal.

Harold
SW2009 SP4.0 OPW2009 SP2 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com