INSTRUCTIONS:

To find K40 conversion information use the "INDEX", "SEARCH" or "LABELS" sections in the sidebar.

Monday, June 6, 2016

K40 Optical Model & Alignment Tools

The K40 optical path model

Donate:

Please consider donating (button to the right of this post).
Your donations help fund additional research, tools and parts that I will return to the community as information and how-to's. 

Background

Optical alignment of the K40 is painstaking. Since the beam is invisible adjustments have to be made incrementally vs continuous. Adjustments consist of putting a target material, sticky note or tape over the lens that is the target of the beam, pulsing the laser, make adjustment etc.etc.
It would be desirable to have a means of seeing the light on its optical path so it can be adjusted dynamically.
I have observed  many attempts at creating an through the optics visible alignment beam such as:
  • Laser pointer mounted on end of laser
  • Beam combiner where a Laser pointer is combined into the same path as the infrared beam

Not much success

Here is a list of challenges I have encountered trying to make alignment visible:
  • I cannot mount a diode on the end of my laser as there is inadequate room.
  • No mater what source you use the diode has to be aligned to be perfectly in the center and perpendicular on exit to whatever mounting is used.
  • The diode's light does not always exit perpendicular to its housing.
  • Beam combiners use up power and are expensive
So far it has been more challenging to align the alignment tool than to just align the mirrors using the standard approach :).

Optical Improvements

Before embarking on the optical path modelling I redesigned the head mount.

Optical Path Modelling

I stepped back and decided to model the optical path so that I had a numerical sense of where things are.

http://dck40.blogspot.com/2013/02/post-by-nycon-how-to-adjust-mirrors-for.html

My goals for modelling are:

  • Model the Opto-mechanical subsytem
  • Make alignment easier and more accurate

The Optics Model

Optical Bench

The thing that bothered me about the typical K40 alignment approaches is that the optical path is aligned a single point at a time though the use of targets. This works "ok" if only a single mirror needs tweaking on an already square and parallel optical path. However its necessary to understand the overall optical path position in 3 planes.

To get the best performance anywhere on the bed:

Looking from the top
  • The x axis movement of the objective lens has to be parallel to the laser beam
  • #1 mirrors has to be at -45 degrees to the lasers beam.
  • #2 mirror has to be 45 degrees to the laser beam
  • #3 mirror has to be mounted perpendicular to and at 45 degrees to the laser beam
Looking from the side:
  • All mirrors must be perpendicular to the same plane, which is parallel with the laser beam
  • All mirror centers must align with the laser beams center
  • The objective lens must be parallel to the plane of the laser beam.
I concluded that a better way to align the optics was to establish a reference plane that was aligned parallel to the gantry's x movement. This would work like an optical bench. By putting optical targets on this plane and firing the laser I could detect and measure the parallelism of the full optical path in two axis and at multiple points at the same time. Then looking at the targets would reveal what needs adjusting.

The idea is to put multiple targets in the path of the laser, fire and mark them all. After firing the laser I can see if the beam is parallel to an axis by observing where the beam hits each of the targets relative to each other. If they are dead on with each other the beam is parallel to the surface the targets are sitting on. If they are offset, the beam is traveling at an angle to the surface they are sitting on.

Requirements and sources of error:
  • The target holders must be identical. That is, the center of each target must be the same distance (height) from the surface they are mounted on. 
  • The surface they are sitting on must aligned to the gantry's x-y axis.
To achieve the above requirements I decided to make a "bench" that is affixed to the gantry.

The bench

The bench is made up of a steel plate, struts and fasteners. The bench surface plate is constructed of a .035" thick sheet of steel that was cut to fit in the K40's bay and around mirror #2. This metal was procured from Home Depot.
The bench is mounted on top of precisely fabricated wood struts that tightly hug the gantry in x and Y. The dimensions of the struts were carefully established to space the bench plate up from the gantry surface making its surface level with the top surface of the #2 mirror mount as well as the head.
With this setup the bench plate, the top surface of the #2 mirror mount and #3 mirror are now on the same and hopefully level plane. This plane becomes the datum that all targets are referenced to. The bench plate is fastened to the struts for stability using fasteners and magnets. See drawings for dimensions.

Optical bench layout, struts are underneath

Struts that straddle the frame in the Y direction,  affixed to the underside of the bench

Strut with hold down magnet, straddling the gantry
The Targets & Holders

The targets and their holders are then designed in reference to the bench plane. The target holders are designed such that the center lines of the targets films are the same height off the bench. The single target image is designed to be used in three places:
  • Free inline positioning on the bench by including a fold and held down by magnets
  • Installed in mirror #2's target frame
  • Cut and folded to mount in front of mirror # 3.

Main Target

The main target is placed on the wall inside the laser cabinet with the target graphics facing inside the cabinet.

Inline Targets

The target films can be placed anywhere on the bench plane by folding them and using magnets to hold them in place.


 # 2 Mirror Targets

The target holder for # 2 mirror is designed to be installed on the mirror mount and features a replaceable target film made from printing the target on card stock.
#2 mirror target holder installed

Target holder assembly

Components of the target holder

Newer targets printed on card stock and installed in holder

#3 Mirror targets

#3 mirrors target installed

Test #1 Results

I set up 4 targets for this test.
  • Target 1: main laser target in the laser compartment
  • Target 2: inline target a few inches from 1 magnet-ed to the bench
  • Target 3: inline target a few inches in front of mirror #2 magnet-ed to the bench
  • Target 4: target on mirror #2

Targets with X axis aligned.

Reading the results.

In the picture above I aligned the x axis of all the targets laying them out from left to right. Progressing left to right we observe that the beam is on the same plane in each target from target 1 though all the targets to #4. Since these targets are sequential, at the same height on the bench and spaced apart on the Yaxis it shows that the laser is parallel to the Y axis of the gantry all the way to mirror #2.

We have 3 data point in alignment and that proves the laser is level with the Y axis of the carriage.

Seems the theory of using a bench has proven to be practical.

What do that results on each target tell us?

Target 1: the laser beam exits just to the upper right of the hole in the frames center. This is not germane to alignment but does give a dimensional reference point relative to the cabinet.

Target 2: the left right (x) position of the spot on the target is not important because the x position of the target was not of interest. Later a X alignment ruler will be added to the bench to align the X. What is important is that the beam mark is directly on the center of the Yaxis.

Target 3: this target can be interpreted like target 2, just further away from target 1. The beam is still on the center of the Y axis meaning that the beam is not pointing up or down as it approaches mirror 3.

Target 4: shows that the beam is still straight on mirror #3 and slightly to the left of center. Mirror #3 should be moved to the the right looking from the front of the machine. I just reinstalled this mirror so its not surprising its off center.


Encouraging results for the first test .......

Final #2 Mirror Adjustments

After just 2 movements using the thumbscrews on mirror #2 mount I achieved these results.

Mirror #3 adjustments

Mirror # two is adjusted to center on mirror #3 then mirror #3  is rotated so that it is perpendicular the the incoming beam

The final proof, is the objective lens perpendicular?

Alignment test of the objective lens

The light going through the objective lens is checked for perpendicular-ism by lowering the bed, pulsing and lowering again multiple times. If the beam mark stays in the same place it is perpendicular. If it moves off center it is not. 

A target was laid on the bed and positioned until the burn was on the center mark. Then the above procedure was run. 


Mirror 3 final alignment. Objective lens (output) parallel-ism.

As you can see the beam stayed very close to the same place (highlighted inside the box on the left photo). With these results its clear that the beam in this machine is parallel and aligned. 

Target Materials

In the examples above I used colored 100lb card stock. Below are some other ideas on materials.

Post Its

(From: Jeff Glee.)
If it is possible to lower the beam power to the single watt range, a Post-it makes a great beam locator. The dye used to color the paper is temperature sensitive, and darkens when hot. Depending on the intensity used, it may be necessary to attach the post-it (I used 3M spray adhesive) to a thin aluminum plate (less than 0.1") to help dissipate excess heat and prevent flames. Different color Post-it's react differently, but the best I have found is the bright pink color. Visually, it is about the same effect as you see on commercial fluorescent plates. I have yet to try it at higher than about 2 watts with a 5 mm beam, but given sufficient thermal dissipation (thicker backing, or copper instead of aluminum), higher intensities should be viewable without burning.

Conclusion

I like that along with nearly perfect visual alignment in 3 axis I now have a record of beam position for multiple points along the optical path.


Enjoy and comment
Maker Don