My reason for building a "Artificial Star" was to enable me to do a little "tweaking" of the front to rear lens element alignment on my Meade 127 ED refractor.  I needed to eliminate problems with coma that was due to mis-alignment of the two lens elements themselves.  Trying to do this while actually star testing on a REAL star involved all the typical problems you might expect.  Problems with the "seeing" conditions, problems trying to get at the lateral lens element adjustment screws while the telescope is aimed upward, etc.  It is the kind of issue that you simply can't correct by normal collimation methods such as a Cheshire or laser collimator.  Something different was needed that would REALLY show how the lens was performing, regardless of the weather conditions between the telescope and the stars.  The artificial star can even be used in the daytime, should you have the need.
        The great thing about this tool is that the results it can give you border on the astounding.  I was able to quite easily eliminate any and all traces of coma from mis-alignment that were present in the Meade.  This resulted in pin-point star images that are as good as the lens is capable of delivering.  In reality, that turns out to be very good, indeed.  Prior to that I had wondered about the proverbial "bang for the buck" actually being there, at least with my particular Meade.  Now I have to say that I am in agreement.  When they are performing to their potential the Meade ED refractors definitely give you an exceptional "bang for the buck".
         This is not to say that I've turned it into a TMB,  Astro-Physics or Takahashi duplicate, but when you consider the price difference you have to be quite satisfied with the performance.
        At this point in time I would also like to make it perfectly clear that I am not the slightest bit qualified to analyze diffraction patterns or comment on "color", under or over correction, zones, turned down edges, etc.  The only thing I can be quite sure of is whether or not the optics are aligned.  (although astigmatism isn't too hard to see either)  To find and correct optical alignment problems was my sole purpose for constructing this "artificial star".  Hopefully I will slowly learn to recognize some of the "finer" optical qualities, or lack thereof, when looking at diffraction patterns.  In the meantime, the artificial star "doctor" has paid for itself several times over by helping its' first sick patient recover from a simple case of lens element "mis-alignmentitis".

The initial search.

        Step one was to try and find out what a artificial star is actually supposed to be and how one is constructed.  You will quickly find that there is a very limited amount of information available on the subject, particularly on the internet.  For that reason I hope this little "write-up" can be of help to someone in need of one of these great tools.
        Praise the Lord for the "few" sources of information on this subject, which I will list at the end.

What is the objective?

        To be able to star-test and tune a telescope without having to worry about seeing conditions and to be able to do it when you want, or when the need is there.  Or, as I told my wife, it gives me something to look at when it's cloudy outside, as it usually is.

What are you really trying to construct?

        Basically, you need a fairly bright, small point of light that is no larger than one half (½)arc second in diameter at the distance you will place it away from your telescope.  (this size will work for almost any telescope and you can always place it further away from the telescope, if necessary, or change they eyepiece)  In my case I have a point of light that is slightly smaller than one half arc second at about 100 feet.  I use it at a distance of about 150 feet most of the time.  I cannot imagine this not working for just about anyone.
        Just for the record, it is recommended by Suiter (Star Testing Astronomical Telescopes) that you use a artificial star of the required size at a distance of at least 20 times the focal length of the telescope.
NOTE:  If your objective is simply to align the optics of a refractor lens, such as a Meade ED APO, or collimate a Newtonian reflector, you do not need to be nearly as selective in the size of the "star" itself.  A bright spot of light will suffice.  For actually "testing" the quality of the optics it is a much different matter.

What components are necessary?

        You will need a light source, a opening or hole of a specific size for the light to shine through, and a ocular (eyepiece).
The light source I used was originally a 1100 mcd white LED.  It worked great.  I then went with a brighter 5600 mcd "Ultra Bright" white LED as the light source.  In both cases they worked great, but I would recommend the brighter LED..........not out of absolute necessity though.
        Also be aware that these LEDs are made for about 4.5 volts.  This means you would need to power them with three (3) "C" or "D" cell batteries.  My unit is set up with three "D" cell batteries that are mounted on top of the unit with a power cord that can plug into a RCA jack located on the back of the artificial star, immediately behind the light source.
        You will also need to do a little soldering if you choose to construct this in the same fashion I did.  Certainly, there are other ways to do it.

So, what size hole will be one half (½) arc second, and at what distance?

        A five inch refractor can resolve 0.9 arc seconds, a six inch refractor can resolve 0.74 arc seconds, and so on.  The larger the scope, the greater the resolving power.  So, to properly test the optics you need to create a artificial star that will be no larger than the resolving power of whatever scope you are testing.  As long as you're going to the trouble to build a artificial star you may as well build one that can test most any optics you would encounter.  I decided to make mine slightly smaller than 1/2 arc second.  AND, the "star" size will be adjustable.  How do we figure out the component sizes?
        Actually, this is quite easily figured.  You first need to be aware that ONE ARC SECOND is .0058 inches at a distance of 100 feet.  Assuming you will use the artificial star at a distance of 100 feet you will need a "spot" of light no larger than .0029 inches in diameter for a ½ arc second or smaller star.  Since you will have a heck of a time finding a 0.0029 inch or smaller drill bit, there has to be a different way to do this.
        There are three components that will play a role in determining the size of your "apparent" star.  I will list the particular sizes and distances I used beside the items that are listed below.

    1.  Diameter of the hole the light source shines through.  (#32 drill - 0.116 in.)
    2.  Distance from the aperture (hole) to the eyepiece or ocular.  (400mm)
    3.  Focal length of the eyepiece or ocular.  (9.7mm)

        The distance from the hole to the eyepiece (400mm) divided by the focal length of the eyepiece (9.7mm) will give a result of  41.237.  You then divide the size of the hole (0.116 inches) by 41.237 (or multiply by the reciprocal, as shown below).  This gives a resultant star size of  0.002813 inches, or slightly less than ½ arc second at 100 feet.  (remember, ONE arc second is .0058 inches at 100 feet)
        Another way to do this would be by dividing the focal length of the eyepiece (9.7mm) by the distance from the hole to the eyepiece (400mm) for a result of  0.02425.
        0.02425 can then be used as a "magnification factor".  Multiply the size of the hole (0.116) times 0.02425 and you also arrive at the same apparent star size of 0.002813 inches.
        You can work forward or backward and with any combination of hole size, distance from the hole to the ocular, and focal length of the ocular, to arrive at the combination of your choosing.  It all depends on how large you want the artificial star unit itself to be.  Mine has a total overall length of about two feet.  I find it to be very handy and easy to work with.

Minor technical note!

        It does not appear to me to make any significant difference, HOWEVER, the eyepiece used in the artificial star should be used backward.  The "eye" portion should be toward the light source, and the "field" end of the eyepiece (the end normally inserted into your telescope) should be aimed toward your telescope.  Notice the direction of the eyepiece in one of the pictures below.
        People who know more about this subject than I can ever hope to scratch the surface of say that it is technically correct, but not of great importance.  It actually will work either way.

What holds all this astro-stuff together as a unit?

        It can probably be as elaborate or simple as you want it to be.  At present it appears to me that 2" diameter PVC pipe is probably the easiest material to work with due to the ease of making it in modular sections.
        One section should hold the light source and wiring to the light source.  The second section is to hold the "hole" or "aperture" of a specific size that the light shines through.  The final section is simply to attain the distance you calculate that you need between the "aperture" and the ocular (eyepiece).  Connecting these sections with PVC 2" couplings is a snap.
        The most difficult part is making a couple of round pieces to fit inside the pipe that can hold the LED (hopefully in a LED holder) and another to put a few inches ahead of it with the desired hole or "aperture" size.  Go slowly and make them fit inside the coupler well so things don't move
around.

LED holder mounted on metal disk

        At the eyepiece end it is easy to simply use a 2" to 1.25" adaptor from any telescope focuser to hold the eyepiece.  Since it will be a little sloppy in the 2" PVC, I wrapped my adaptor with a couple rounds of electrical tape until the fit was like I wanted it.  I also drilled the PVC for three thumb screws around the perimeter to hold the adaptor in place.

Other options.

        You can, or maybe should, also consider adding a baffle or two in the main length of tube to minimize scattered light.  You should also paint the inside of the tube black.  I suppose flat black is probably preferred.  I don't have baffles in mine at this time and don't see any ill effects however, it will probably end up with at least one in there.
        I decided that I wanted to have a "on-off" switch on the unit and the ability to remove the power supply from the unit when not in use.  It was easy to place a switch and a plug at the very rear of the unit, immediately behind the LED holder section.  They are mounted on a PVC end cap that seals the rear of the unit.
        I also decided to attach a "home made" battery pack to the top of the unit.  This was easily done with left-over PVC and two end caps.  (along with three battery holders and a little soldering)
        I also used five 2" pipe clamps in the construction.  One was simply used to give me the ability to mount the unit on a tripod.  (I attached a "quick-change" Bogen plate to the bottom of the pipe clamp)  The other four clamps were used to mount the battery pack on the top of the unit.

Invaluable assistance was received from the following:

        My thanks to the assistance given by Landrum Haddix, the person with the only web site I could find that had a word to say about building a artificial star.  The web site and a few e-mails from him helped me get on the right track with this project.
        http://web.qx.net/lhaddix/astar.htm

        I also want to thank Thomas M. Back, one of the "Grand Masters" of optics for confirming that what I built should in fact function properly.  He also mentioned that a green filter should be used when star testing.  A Wratten number 58 (dark green) is recommended and you can use it at either end of the system.
        I finally became aware of his "Star Test Primer" after I had probably driven him nuts with trivial questions, many of which are already answered there.
        http://voltaire.csun.edu/tmb/tmb4.html
        In addition, Mr. Back supplied me with a copy of the article from a 1992 issue of Sky & Telescope that described Richard Berry's method of constructing a artificial star.  (10 years ago something was printed on the subject)

        Of course there is also information contained in H. R. Suiter's essential book, "Star Testing Astronomical Telescopes".  This book can give you all the technical information you will probably ever need but I found I needed a little more help with the actual construction technique.