Replies

[Charlie Miller]

Greetings,

I'm a new member with an old projector - a home-made 16" diameter version of the Steven B. Smiith 20" drum style. My 16" drum's strength is its dense star field (equivalent to the norton star alas 6.4 magnitude field), but because of its relatively small diameter, a strong design demand is the light source.

I'm writing to ask about experiences (and my observations) about light sources. I've migrated to the "bi-pin" still of bulbs used in Maglight (they have high intensity ad small filaments -- two very good things), but I'm still looking for the optimal. One weakness of the maglite style bulbs (or at least the Rayovac "High intensity" 3.6V, 0.3A bi-pin model I'm using) is that it's small glass envelope precludes distributions of light evenly throughout a solid angular region of 340 degrees (the angular coverage of my device). Old research I did on the Spitz A3 indicated that they used to use a bulb that had two important design parameters: a large glass envelope (thus increasing the distance from the filament to the base -- and thus the angular coverage of the light field) and a filament that seemed to be carefully placed within the center of the spherical shape of the bulb. This latter design aspect prevents secondary images developing as a result of filament light reflected off the inside surface of the glass envelope.

Perhaps I'm mentioning things that are well known; I apologize if this is the case. But I'm still searching for that bulb that has:

(1) the smallest (point-like) filament

(2) appropriate brightness (of course, dependent upon dome size, etc.)

(3) whitish color temperature (this is a toughy for incandescent bulbs, with there yellow-weighted spectra)

Note: The proper combination of items (2) and (3) are important, as if I'm compelled to "dim" the bulb because
it is simply too intense, then I get further away from white color temperature.

(4) large coverage of the star sphere / cylinder (i.e. globe of stars)
(this seems to be a drawback of these small "maglite" type bulbs, darn it! I'm loath to thinking of going to a two-bulb
configuration, due to the demand for precise location and masking, etc.)

(5) filament centered within an ideally spherically shaped glass envelop, to prevent the formation of
"secondary" or ghost images due to the lack of such a geometry.

Anyway, I'd appreciate hearing from any who have looked at this issue. I do find that the mag-lite bubs are pretty good, but
I'm striving for the best -- because it is THE limiting factor for my 16 star cylinder.

Thank you for any/all help out there.

Charlie Miller
Home-builder

P.S. I have emailed with Gare and hope to send him a photo or two of my star cylinder and perhaps a shot of a star field (i'm currently working on re-doing all the condenser lense mounts and all that, however).

P.P.S. Thank you for this oppotunity and venue.

[Nathan Volle]

Greetings Charlie, I am actually quite new to O C too, and have had your question---as have ALL of us with "starballs"---. looking for that perfect light source. I have used a Starlab for 5 years doing presentations---including teaching week-long workshops in high school.

Greetings,

I'm a new member with an old projector - a home-made 16" diameter version of the Steven B. Smiith 20" drum style. My 16" drum's strength is its dense star field (equivalent to the norton star alas 6.4 magnitude field), but because of its relatively small diameter, a strong design demand is the light source.

I'm writing to ask about experiences (and my observations) about light sources. I've migrated to the "bi-pin" still of bulbs used in Maglight (they have high intensity ad small filaments -- two very good things), but I'm still looking for the optimal. One weakness of the maglite style bulbs (or at least the Rayovac "High intensity" 3.6V, 0.3A bi-pin model I'm using) is that it's small glass envelope precludes distributions of light evenly throughout a solid angular region of 340 degrees (the angular coverage of my device). Old research I did on the Spitz A3 indicated that they used to use a bulb that had two important design parameters: a large glass envelope (thus increasing the distance from the filament to the base -- and thus the angular coverage of the light field) and a filament that seemed to be carefully placed within the center of the spherical shape of the bulb. This latter design aspect prevents secondary images developing as a result of filament light reflected off the inside surface of the glass envelope.

Perhaps I'm mentioning things that are well known; I apologize if this is the case. But I'm still searching for that bulb that has:

(1) the smallest (point-like) filament

(2) appropriate brightness (of course, dependent upon dome size, etc.)

(3) whitish color temperature (this is a toughy for incandescent bulbs, with there yellow-weighted spectra)

Note: The proper combination of items (2) and (3) are important, as if I'm compelled to "dim" the bulb because
it is simply too intense, then I get further away from white color temperature.

(4) large coverage of the star sphere / cylinder (i.e. globe of stars)
(this seems to be a drawback of these small "maglite" type bulbs, darn it! I'm loath to thinking of going to a two-bulb
configuration, due to the demand for precise location and masking, etc.)

(5) filament centered within an ideally spherically shaped glass envelop, to prevent the formation of
"secondary" or ghost images due to the lack of such a geometry.

Anyway, I'd appreciate hearing from any who have looked at this issue. I do find that the mag-lite bubs are pretty good, but
I'm striving for the best -- because it is THE limiting factor for my 16 star cylinder.

Thank you for any/all help out there.

Charlie Miller
Home-builder

P.S. I have emailed with Gare and hope to send him a photo or two of my star cylinder and perhaps a shot of a star field (i'm currently working on re-doing all the condenser lense mounts and all that, however).

P.P.S. Thank you for this oppotunity and venue.

After long searching and experimenting, I found, (for my needs), the perfect compromise bulb for brightness and point source quality. It may have been mentioned by someone, sometime, on OC---but I have never seen it. It may be that others have ruled it out for filament size or for something else---but it is DEFINITELY the best compromise I have ever found for the standard Starlab projector and I think would be a good choice for you even if you found something better later.

I would welcome others comments on whether they have tried it----and on any drawbacks they may have found----for their particular use. My guess is that filament size is a problem to some and/or shadowing. For my use, to have enough brightness is a key requirement.

It is called: MAGCHARGER RECHARGEABLE Halogen Replacement Lamp for 6 Volt Rechargeable Flashlight

It is requires 6 volts and draws 1.65 Amps which is 9.9 watts. I get them at Frys Electronics for $4.95 each.
The store sticker says: MAGLITE LR00001 CHARGER REPLACEMENT LAMP.

I hope that might help you---and welcome to O C! Nathan

Attached File(s)

 6V_Nicad_Bulb.JPG ( 33.31k ) Number of downloads: 4

 

[Ken Miller]

Greetings Charlie, I am actually quite new to O C too, and have had your question---as have ALL of us with "starballs"---. looking for that perfect light source. I have used a Starlab for 5 years doing presentations---including teaching week-long workshops in high school.
After long searching and experimenting, I found the perfect compromise bulb for brightness and point source quality. It may have been mentioned by someone, sometime, on OC---but I have never seen it. It may be that others have ruled it out for filament size or for something else---but it is DEFINITELY the best compromise I have ever found for the standard Starlab projector and I think would be a good choice for you even if you found something better later.

I would welcome others comments on whether they have tried it----and on any drawbacks they may have found----for their particular use. My guess is that filament size is a problem to some and/or shadowing. For my use, to have enough brightness is a key requirement.

It is called: MAGCHARGER RECHARGEABLE Halogen Replacement Lamp for 6 Volt Rechargeable Flashlight

It is requires 6 volts and draws 1.65 Amps which is 9.9 watts. I get them at Frys Electronics for $4.95 each.
The store sticker says: MAGLITE LR00001 CHARGER REPLACEMENT LAMP.

I hope that might help you---and welcome to O C! Nathan

I have tried this lamp, and didn't like it for two reasons (I reported on it a couple of years ago). First of all it has a glass envelope that causes a strange refraction and generates secondary star images. Secondly, it is so bright that all the projection defects are really noticable. All of this might be ok if you dim the light down enough that the problems are hard to see. In a room that is not fully darkened, it might work well.

Ken

[Nathan Volle]

That's great feedback Ken. Interestingly with a 6.3 volt transformer and the magnetic dimmer I have been using---I have no complaints using it with the Starlab projector.

I have tried this lamp, and didn't like it for two reasons (I reported on it a couple of years ago). First of all it has a glass envelope that causes a strange refraction and generates secondary star images. Secondly, it is so bright that all the projection defects are really noticable. All of this might be ok if you dim the light down enough that the problems are hard to see. In a room that is not fully darkened, it might work well.

Ken

I do know how aggravating double images can be though, and I suppose with the amount the light is dimmed, combined with the background light level----I must have unknowingly reached some acceptable balance that has outweighed my other subconscious criteria. I'll try the Stinger too and compare. Nathan

[Charlie Miller]

That's great feedback Ken. Interestingly with a 6.3 volt transformer and the magnetic dimmer I have been using---I have no complaints using it with the Starlab projector.



I do know how aggravating double images can be though, and I suppose with the amount the light is dimmed, combined with the background light level----I must have unknowingly reached some acceptable balance that has outweighed my other subconscious criteria. I'll try the Stinger too and compare. Nathan

I've switched over to the "stinger-like" bulb (only mine is marketed by Ray-o-vac) and can confirm that the filament size is somewhat smaller and the star images are brighter than most low-voltage incandescent bulbs that are readily available. There is still that annoying problem of projecting a non-point-like filament. As it describes an arc, the star images across the projected hemisphere vary from near-point-like to rather sub-optimal "straight-on" image of the filament.

So here's a thought of potential minor interest. If your projector has a horizon cut-off "basket" inside the starball or stardrum and that basket is designed to "float" along two orthogonal axes of rotation, then one could arrange the bulb socket such that the best-quality images (i.e., most point-like) appear in a preferred part of the projected sky. For example, the star-rise and star-set areas (east and west) could be given best image-size, or perhaps a region of typical interest- along the ecliptic - could be given top priority. So, if your bulb-holder has some degree of positional adjustment, you could do this.

I wish that my star-drum had a larger diameter (> its 16"); a larger star ball would reduce filament size (for a fixed star-hole diameter). but then, of course, you'll need a brighter filament to maint star brightness.

[Nathan Volle]

Thanks Charlie---If you had details on your Ray-o-vac bulb I would be interested----meaning voltage and current and part number.

From the beginning of my quest to find the right bulb I have been in the continual process of relaxing my expectations. The 6 volt Nicad is still the best I've found for my purposes.

I've switched over to the "stinger-like" bulb (only mine is marketed by Ray-o-vac) and can confirm that the filament size is somewhat smaller and the star images are brighter than most low-voltage incandescent bulbs that are readily available. There is still that annoying problem of projecting a non-point-like filament. As it describes an arc, the star images across the projected hemisphere vary from near-point-like to rather sub-optimal "straight-on" image of the filament.

So here's a thought of potential minor interest. If your projector has a horizon cut-off "basket" inside the starball or stardrum and that basket is designed to "float" along two orthogonal axes of rotation, then one could arrange the bulb socket such that the best-quality images (i.e., most point-like) appear in a preferred part of the projected sky. For example, the star-rise and star-set areas (east and west) could be given best image-size, or perhaps a region of typical interest- along the ecliptic - could be given top priority. So, if your bulb-holder has some degree of positional adjustment, you could do this.

I wish that my star-drum had a larger diameter (> its 16"); a larger star ball would reduce filament size (for a fixed star-hole diameter). but then, of course, you'll need a brighter filament to maint star brightness.

I still want to try anything that might help though, and appreciate your useful thought on adjusting the bulb filament orientation. Also I would be interested in what type lenses you used on your cylinder and how you mounted them. Do the lensed stars project small and round even with overly large filament bulbs? Did you ever project more than one star with one lens? The picture below, (taken without a tripod), is of the Starlab starfield showing cassiopeia etc---Deneb is bright and way above the star cylinder near the top. This gives some idea of the quality, (or lack of), that I have come to tolerate. Any thoughts or advice is welcome! Nathan

Attached File(s)

 Starfield_Cassiopeia_upload.JPG ( 21.57k ) Number of downloads: 4

 

[Charlie Miller]

Thanks Charlie---If you had details on your Ray-o-vac bulb I would be interested----meaning voltage and current and part number.

From the beginning of my quest to find the right bulb I have been in the continual process of relaxing my expectations. The 6 volt Nicad is still the best I've found for my purposes.



I still want to try anything that might help though, and appreciate your useful thought on adjusting the bulb filament orientation. Also I would be interested in what type lenses you used on your cylinder and how you mounted them. Do the lensed stars project small and round even with overly large filament bulbs? Did you ever project more than one star with one lens? The picture below is of the Starlab starfield showing cassiopeia etc---Deneb is bright above the star cylinder. This gives some idea of the quality, (or lack of), that I have come to tolerate. Any thoughts or advice is welcome! Nathan

Dear Nathan....
Greetings. Nice picture and star field. Although it's hard for me to say anything super-critical with that particular photo size.
Here is a rather long reply to your short query, with likely more information than you want...


Lenses...
My 16" star drum will eventually have the brightest 120-140 stars with plano-convex condenser lenses. I'm using the "experimental grade" (i.e., likely chipped, etc.) lenses from Anchor Optical -- an offshoot company of Edmund scientific. I typically use the 8mm diameter, 203 mm focal-length lenses, although I'm also playing with the 265 mm F.L. lenses used by Steven Smith's 20" star drum. Curiously, Ive found that having a careful match of the lens. F.L. to the bulb-to-hole distance does not always produce the best image. For example, the 203mm lens works better for Vega, even though a simple calculation shows that, for my star drum, it's a toss-up (between 203 and 265 mm.).

Grinding lenses...
In some areas, the lenses may have too large of diameters to properly center the lens center over the star-hole center (orion's belt). Putting the lens "off center" will cause distortion of the image and the projected location. Thus, i'd recommend grinding away lenses that present this problem. This can involve wet sanding with optical grit paper or grinding grits. I would not recommend cruder methods of cutting lenses, as they will tend to chip the glass. If you can make some sort of rotating disk (with a lip to contain the abrasive slurry), then you can fixate the lens on a stationary jig that positions the lens over the abrasive disk.


The lense mounting tubing...
I mount each lens in a brass tube that is beveled to assure that light exits each in a plane perpendicular to the lens surface. The tubing is the very thin-walled variety that you can buy at better-stocked hardware stores (hobbyist tubing that is rigid, not like, say, thin-walled copper plumbig tubing). Having thin-walled tubing is important, as sometimes the lens mounts get in the way of nearby stars. You can also find thin-walled stainless-steel (S.S.) tubing, which provides different ranges of inner diameters, if you have the need. You can also find aluminum tubing, but it tends to have a bit thicker walls, a disadvantage.

Of course, S.S. is more expensive and a bit harder to cut. I do all my cutting on a bandsaw (using a 32 TPI bi-metal blade) and then clean up the cuts on a stationary belt sander (80 grit paper) and then do finer de-burring using 400-600 grit wet-or-dry paper on a finisihing sander. One slight variation is to do all the beveling on the sander (not bevel on the band saw)... you might be wasting more material and more of the sanding belt, but its an option. Using a bandsaw to cut the tubing is convenient in that the tubing can be cut without crushing or deforming it during the cut.

Manual cutting of the tubing..
If you use a more manual approach (say, a coping saw or hack saw), you might want to make your own mitre-cut box that features the typical mitre-box shape PLUS two pieces of wood that clamp around the tubing PLUS an inner dowel to fit snugly into the to-be-cut tubing. I made such a box for near nothing when I used to cut the tubing by hand-powered saw. I think my mitre box had 5 degree increments. Of course, all this business of making angled tubing only applies to star-drums, not star-balls.

I've found that you can have a 5-degree error in the tube bevel angle without any noticeable effect. Lens imperfections are a bigger contributor, I suspect.


Lens fixation, etc.
I use a non-permanent glue (basically hobbyist glue that is similar to that used for plastic-model building). That glue sets up in just a few minutes, which is handy. I use that glue for both lens-to-tubing fixation and then gluing the brass fitting to the star drum. One disadvantage is that one can accidentally knock off the brass fittings, but I prefer that to using something "more permanent", particularly given possible long-term needs for cleaning or re-aligning. Also, if needed (due to a sloppy gluing), you can always re-do the gluing).

I advise priming the cut-and-sanded brass fittings with a primer coat to prevent oxidation. That is done expeditiously by sliding a series of the beveled fittings onto a dowel and then spray painting, say 15-20-25 of these fittings all at once, using ~ 3 light passes of the sprayed paint. I've been told by a machinist friend that Krylon (basic grey undercoat) is a good quality paint.

In some cases (say for the star Sirius), you might need a different diameter lens and/or you might find difficulty in getting a close match of the lens diameter to the brass tubing. In that case, I'd suggest buying undersized tubing and cutting an expansion slit along the axis of the tubing, to allow it to stretch to fit. Indeed, this trick can also be used (slitting the tubing) in cases where the presence of a slit would prevent the loss of projecting a nearby star whose light would otherwise be blocked.

Tubing for milky way diffuser plastic
I also use similar approaches for Milky Way diffusers. I used Steven Smiths' guide to milky way apetures as a point of departure. The nice thing about milky way diffusers is that you can put in a lot of holes--of diffierent diameters and "sculpt" the region of projection to better fit observations of the milky way. Also, you are not stuck with a particular tubing diameter, but can tailor that diameter to more exactly fit the diameter of the milky way hole. In practice, this means getting away with using smaller tubing, which in turn, allows you to use more M.W. holes (than say Smith suggested) so as to get a more precise contour of the glow.

You might consider painting the MW tubing a different finished color, just to make the functional distinction that much clearer.


Milky Way and Star Field references.
Some 30 years ago, Steven Smith sent me his recommended hole sizes and positions for his M.W. holes for his 20" drum. I found them to be a little too small (once his sizes were scaled to my drum). Also, I have added more holes. I'm not happy with my milky way yet, as it lacks the precision of the real thing. I anticipate the need to use a lot of holes (with relatively small diamters) to do this sculpting. Norton's Star Atlas is a rather inferior source for MW contours, as it offers only a one-intensity contour. The 3rd edition of the Cambridge Star Atlas (available at Amazon, for example) provides a 2-intensity MW contour, which can give you a first-approximation of how you might want to lay out apertures. Then, of course, there are photos, an even better source. There is also the excellent Beckvar charts (no longer published). Both the Cambridge and the Beckvar charts are superior to Norton's in that they encode magnitude in smaller steps and provide star color temperature (if you wanted to be brave and consider coloring some of your stars). Then again, the very rich (>100,000 star) Hipparcos star catalog is on the Internet, free for the downloading, and it provides all sorts of per-star data. I have made my own star charts based on the hipparcos catalog.... one can even convert magnitude numbers to star-hole drill sizes and encode drill size directly on your own charts. It helps that I use a sophisticated plotting program (Sigmaplot) as part of my work-day computer tools.

Comment on MW and star-field compatibility
The addition of variable-intensity Milky Way greatly improves the overall illusion, adding "depth" and richness. However, I personally have a problem with projectors that have a rather conservative star field (say a limitting magnitude that provides thousands fewer stars than the 6.35-6.5 mag limitting magnitude encounted in dark skies), yet still include the Milky Way. Some of the smaller Spitz projectors are "guilty" of that... it's done because MW projection is relatively easy, compared to adding 1000's more stars. But I've found marrying a limited star field to a MW field intelletually and esthetically problematic

Charlie Miller

[Nathan Volle]

Charlie---How much do the lenses cost and what are the possible sources to get them?

Lenses...
My 16" star drum will eventually have the brightest 120-140 stars with plano-convex condenser lenses. I'm using the "experimental grade" (i.e., likely chipped, etc.) lenses from Anchor Optical -- an offshoot company of Edmund scientific. I typically use the 8mm diameter, 203 mm focal-length lenses, although I'm also playing with the 265 mm F.L. lenses used by Steven Smith's 20" star drum. Curiously, Ive found that having a careful match of the lens. F.L. to the bulb-to-hole distance does not always produce the best image. For example, the 203mm lens works better for Vega, even though a simple calculation shows that, for my star drum, it's a toss-up (between 203 and 265 mm.).


Charlie Miller

Thanks for all the good detail you give in your descriptions. I am considering putting lenses on twenty or so of the stars on my Goto E-5 (Viewlex)projector. With a starball of 12 inches, I am assuming that I would need 150 mm F.L. lenses. Any thought or advice? Nathan

[Charlie Miller]

Charlie---How much do the lenses cost and what are the possible sources to get them?




Thanks for all the good detail you give in your descriptions. I am considering putting lenses on twenty or so of the stars on my Goto E-5 (Viewlex)projector. With a starball of 12 inches, I am assuming that I would need 150 mm F.L. lenses. Any thought or advice? Nathan

Hi Nathan,

I have not scoured the Internet for optical supply businesses, so I revert back to Edmund Scientific, whose lens-selling business now appears at www.anchoroptics.com. They sell different quality levels of lenses; their largest selection of F.L. and diameters appear under the "experimental grade", which is their cheapest grade. Typically, those lenses will have chips that usually don't interfere with functionality. (Every once in awhile, they sneak in a lens that looks pretty good, but whose curvative is WAY OFF... always good to look through each lens to check for that less obvious defect... they have a limited return policy time-window.

Good news for you is that if you go to http://www.anchoroptics.com/catalog/produc..._length&d=d
you'll see that you have multiple choices of lenses around 150 mm (I see two, one at 150 mm f.l. dia=12 mm ... likely too big a diameter... and 154 mm f.l., dia=7 mm... which looks better.) The per-lens price for the latter is $6.50, but that price goes down by 30% when you order more than one.

Oh, there's also the possibility of adding two lenses "in series". In that case, I believe the equation for the effective focal length is 1/f(effective) = 1/f1 + 1/f2, where f1 and f2 are the f.l.'s of the individual lenses. Actually, that equation can be less appropriate if there is a large air gap between the two lenses. that's the only equation that I remember, I'd consult a better authority than my crusty memory if you want to go the compound lense route. At any rate, doing so means that you're using two lenses with much longer fl's.

Hope this helps.

Do you have a picture of your projector? I'm just curious.

[Nathan Volle]

That's what I needed to know Charlie. Thanks! I've got some 7mm X 90mm lenses which interestingly seem to work to a degree. I will try the ones that have the theoretically right F. L. though.

Hi Nathan,

I have not scoured the Internet for optical supply businesses, so I revert back to Edmund Scientific, whose lens-selling business now appears at www.anchoroptics.com. They sell different quality levels of lenses; their largest selection of F.L. and diameters appear under the "experimental grade", which is their cheapest grade. Typically, those lenses will have chips that usually don't interfere with functionality. (Every once in awhile, they sneak in a lens that looks pretty good, but whose curvative is WAY OFF... always good to look through each lens to check for that less obvious defect... they have a limited return policy time-window.

Good news for you is that if you go to http://www.anchoroptics.com/catalog/produc..._length&d=d
you'll see that you have multiple choices of lenses around 150 mm (I see two, one at 150 mm f.l. dia=12 mm ... likely too big a diameter... and 154 mm f.l., dia=7 mm... which looks better.) The per-lens price for the latter is $6.50, but that price goes down by 30% when you order more than one.

Oh, there's also the possibility of adding two lenses "in series". In that case, I believe the equation for the effective focal length is 1/f(effective) = 1/f1 + 1/f2, where f1 and f2 are the f.l.'s of the individual lenses. Actually, that equation can be less appropriate if there is a large air gap between the two lenses. that's the only equation that I remember, I'd consult a better authority than my crusty memory if you want to go the compound lense route. At any rate, doing so means that you're using two lenses with much longer fl's.

Hope this helps.

Do you have a picture of your projector? I'm just curious.

The 7mm X 90mm doubled up changes the F. L. to approximately 40mm. This is the setup I plan to use to make more planet projectors to experiment with. When I treat the LEDs as a point source behind a pinhole in the projector--they don't seen to have enough lumens to get the proper brightness---and also the color is off---a harsh bluish/white rather than a soft yellow/white. I will keep experimenting! Thanks again. Nathan

Attached File(s)

 E_5_DOME_PIC_UPLOAD_1.JPG ( 43.43k ) Number of downloads: 1
 CONTROL_PANEL_1.jpg ( 42.13k ) Number of downloads: 0

 

[Nathan Volle]

A correction / addition of the following:

The 7mm X 90mm doubled up changes the F. L. to approximately 40mm. This setup I plan to use to make more planet tubes to experiment with. So L.E.D. s don't have enough lumens from a point on them.

The 7mm X 90mm doubled up changes the F. L. to approximately 40mm. This is the setup I plan to use to make more planet projectors to experiment with. When I treat the LEDs as a point source behind a pinhole in the projector--they don't seen to have enough lumens to get the proper brightness---and also the color is off---a harsh bluish/white rather than a soft yellow/white. I will keep experimenting! Thanks again. Nathan

[Charlie Miller]

A correction / addition of the following:



The 7mm X 90mm doubled up changes the F. L. to approximately 40mm. With this setup I plan to make more planet tubes to experiment with. I am thinking with Ron that L.E.D.s would be nice to use, but so far I a have found with the L.E.D. s I have tried--- that they don't seem to have enough lumens from the single useful point on them, to project with enough brightness.
I will experiment with other L.E.D. s. Nathan

Nathan, I'm inspired by your photos. The Milky Way effect is interesting. To be an annoying critic, does it seem just a bit too bright relative to the stars? Given the problem with "saturation" of either digitial photos or old-fashion silver-emulsion photos, photos of stars can be over-exposed and it can be difficult to accurately deduce the actual relative brightness of things.

Nathan, I'm surprised that LEDs won't work. A lot of LEDs have built-in lenses to focus light in a preferred direction... that's perfect, because in more traditional projectors, you'd need a condenser lens to do that. Look at the LED's that consume 1-2 Watt or so... goodness, I'd think that would be enough. I'm sure you're probably aware that planet projectors typically don't employ the wasteful "pinhole" effect, but actually project an image (transparency) that "takes up" a larger angle of light from the source.

Anyway, it's fun looking at your photos and hearing about your interest in all of this.