I have long considered edge-lighting to be be most desirable way
to light my slide viewers and I'm pleased that 3D World has
chosen this method for their new viewer.

To satisfy my curiosity, I've disassembled my illuminator,
measured it's performance, tweaked it up a bit, written my notes
up and placed them on the web at:
http://stereo.thurstons.us/content/?page_id=58

For list archives, here is an excerpt from that page:

The illuminator in the 3D World viewer is an edge-lit panel. It
works by shining a light into the polished bottom-edge of a lexan
sheet. Through internal reflections, the light is directed out
one of the flat surfaces of the sheet. It provides a very even
light in a very thin package. I’ve always considered edge
lighting to be the most desirable way to light a slide viewer.
I’m very pleased to see 3D World using the concept in their
newest viewer.

Several years ago, I modified one of the 3D World steal-the-light
(STL) viewers and inserted an edge-lit panel. I used a panel with
a fluorescent light source, and powered it from the wall rather
than from batteries. The result was an extremely thin, very
bright light that hid in the back of a my fixed-focus viewer.

In the intervening years, I have considered installing a battery
pack in my illuminated viewer, but the power demands of the
fluorescent light made that impractical. I have also tried
several times to build an edge-lit panel using LEDs in place of
the fluorescent tube. The result would have been an instant-on
light with a smaller electronics package and lower power
consumption.I was never able to design an LED-lit panel which
provided light sufficiently even for slide viewing.

Thank goodness 3D World has succeeded where I was stymied. They
have designed and built an LED-lit panel suitable for use in a
slide viewer. Because it is instant-on (unlike fluorescent), it
can have a momentary switch. Because it has a momentary switch
and LEDs with low power consumption, it can be battery powered.
Because it’s edge-lit, it is very thin and because it’s very
thin, the final viewer can be very compact.


An ideal viewer light will be light weight, sturdy, inexpensive,
evenly lit, bright, long lasting, and of a daylight color
temperature. All of these attributes can be hard to find in a
single product, but I’ve made some measurements of my lights for
comparison.
Viewer Voltage Current Watts ISO 100 EV
Saturn Viewer 9.51 external .3 2.85 11.5
3D World STL 9.5 external .3 2.85 12.6
3D World Illuminated viewer 4.67 .20 0.93 12.0
3D World Illuminated viewer 3.85 .16 0.62 11.6


From these numbers we can see that the LED illuminator from 3D
World requires only 1/3 the power of my fluorescent lamps yet
produces a similar amount of light. The LED source is much
sturdier than my lamps and is instant-on. With these power
requirements (and battery data sheets from Duracell), I estimate
that a set of three alkaline AA batteries would power the light
at an acceptable brightness for over three hours.
Shortcomings

The first thing anyone will notice is that the illuminator is
positioned very close to the slide in the viewer. A consequence
of this is that any dirt on the illuminator will be in focus
along with the slide being viewed. This really isn’t as big a
deal as some would suggest. I’ve been using edge-lit
illuminators in my viewers for years. Yes, the panels can get
dirty and yes the dirt is in focus with the slide. I simply use
a couple blasts of compressed air through the loading-slot every
few months to blow the dust out. If I kept my viewers in a
drawer or a bag rather than laying on my desk in my dusty office
it would probably be even less of a problem than it is :)

As shipped in my viewer (admittedly as single sample from a
relatively early production), the illuminator was poorly
assembled. Having spent several years working with edge-lighting
in slide viewers, I’ve become rather sensitive to their
construction details. To provide the most even and artifact-free
illumination, the “sandwich” must be assembled carefully and kept
free of dirt. The 3D World illuminator required a few
adjustments to let it perform at it’s optimum.

As shipped, the diffuser (the top layer of the sandwich) was
glued along three of its edges to the lexan (center of the
sandwich). The sandwich was then screwed to the back of the
viewer. There were wrinkles in the diffuser which let dirt get
behind the diffuser where it was very visible and distracting
when viewing slides. The wrinkles also produces slight
unevenness in the light which wasn’t terribly noticeable behind
slides but was still a minor distraction with some images.

To correct these two problems, I removed the sandwich from the
back of the viewer, peeled the diffuser off, removed the dirt,
trimmed a tiny bit off the edges of the diffuser, enlarged the
screw holes, reattached the diffuser and reinstalled the
illuminator. I reattached the diffuser with a strip of mylar
tape around the perimiter. This let me get it smoothly down on
the lexan and the enlarged screw holes let me reattach it to the
back without introducing distortions.

During reassembly, I noticed that the LED strip along the bottom
of the lexan was able to pivot away from square. The evenness of
the light in the panel was very dependent on how precisly square
the LEDs were positioned. I will probably insert a thin strip of
foam below my LED strip to ensure it remains pointed directly
into the lexan sheet so I will get the best possible light for my
images.


The eighteen LEDs in the illuminator are wired in parallel and
attached to a strip of circuit board. Power to the LEDs comes
from a separate circuit and appears to be a regulator. The
momentary micro-switch is also located on this board and is
pressed by means of a pivoting plastic button through the top of
the case. The battery compartment runs the width of the viewer
and accepts thee AA cells in series.

Having seen what the rear compartment of the viewer has to offer,
I suggest it would be:

* trivial to install a jack for external power
* easy to install a rocker switch for constant-on operation
* possible to install a larger or rechargeable battery pack
for longer viewing sessions



I am impressed with the illuminator 3D World has created and
think it represents a fantastic value. It isn’t perfect but you
must consider that it’s thin, robust, has good color temperature,
will run for several hours on its batteries and is inexpensive
enough to be used in a viewer that sells for less than $100. You
must also conisder that it is being produced in large quantities.
My fluorescent panels offer a bit more even light, require at
lest 30 seconds to get up to full brightness, cost about $100
build, are fragile and consume three times more power. I don’t
anticipate building any more of my panels now that 3D World has
produced theirs.
--
_________________________
John Thurston
Juneau, Alaska

At 04:42 PM 9/23/2008, you wrote:

>To satisfy my curiosity, I've disassembled my illuminator,
>measured it's performance, tweaked it up a bit, written my notes
>up and placed them on the web at:
>http://stereo.thurstons.us/content/?page_id=58

Thanks for publishing all of this new information, John!

I, too, had noticed that my viewer's diffuser is not perfectly flat,
against the Lexan layer, but it's not nearly as bad as yours was,
going by the pictures seen on your web page. I don't feel a need to
modify mine, but your solution doesn't sound too very challenging.

I think I prefer a brighter viewer than most people are content with,
so I'm curious if you can offer any suggestions regarding how high a
voltage that array of 18 LED's can handle safely.

I wouldn't want heat to become an issue, nor shorten their life span
significantly, but going by your test results, it looks as if you
boosted the measured ISO 100 EV from 11.6 to 12.0 (a 3.45% gain in
EV) by increasing the supply voltage from 3.85 VDC (at 0.16 Amp) to
4.67 VDC (at 0.20 Amp) - a 21.3% increase in voltage.

It's probably not a linear relationship, but if it is, a 100% gain in
ISO 100 EV (making the viewer twice as bright) would require a 617.4%
gain in voltage, which equates to a 27.6V to achieve 12.6 EV, or
33.5V for 13.0 EV. Such voltages would surely fry something, so it
seems there's not much point in trying to boost the brightness
significantly by playing with higher voltage power supplies.

But here's something useful: If anyone wants to add a jack for an
external power supply, your measurements reveal that a transformer
with a rated output of 300 mA (0.3 Amp) at 4.5 VDC should work just fine.

If anyone wants to add an external power supply, here are some
inexpensive components:

Here's a 4.5V, 500 mA power adapter for only $4.85 (before s&h):
http://www.parts-express.com/pe/showdetl.cfm?partnumber=120-015

Here's a 2.5mm panel-mount coaxial power jack for $1.48 at the same vendor:
http://www.partsexpress.com/pe/showdetl.cfm?&Partnumber=090-479

Here's the 2.5mm coaxial plug for the above power adapter (75 cents):
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=120-086

Your current measurements also tell me that the 3DWorld illuminator
is too efficient (a good thing) to warrant the expense of using
non-rechargeable Lithium AA batteries, as I had suggested
previously. Lithium AA's are not cost effective for loads at or
below 300 mA. Good Alkalines will last just as long. Lithium AA's
are great for heavy loads, like a digital camera, but not here.

Mike Davis

Michael Davis wrote:

> But here's something useful: If anyone wants to add a jack for an
> external power supply, your measurements reveal that a transformer
> with a rated output of 300 mA (0.3 Amp) at 4.5 VDC should work just fine.

I'll get back to some of your other items later, but wanted to
mention that if you are going to add a power jack, you should be
sure to get one which disconnects internal power before
connecting the external power (break before make). None of the
suggested jacks seem to have this.

If you use one of the jacks mentioned by Mike you should also use
a single pole, double throw (SPDT) switch to enable/disable the
external power. I can't recommend simply connecting an external
power source in parallel with the internal battery.

This is, of course, irrelevant if you remove the batteries before
connecting external power :)
_________________________
John Thurston
Juneau, Alaska

If anybody's interested in going to shorter focal length lenses that
drop right into the new 3DWorld viewer, check this out...

I ordered a pair of 31.5mm x 66.0mm FL, coated, achromatic doublets
(experimental Grade 1 surplus) from Anchor Optics for $16.10 each (a
total of $45.76 for the pair with sales tax and shipping).

Here's the page where they can be found (look for Stock # AX73771):
http://www.anchoroptics.com/catalog/product.cfm?id=256

It shows a price of $23.00 each, but when you checkout and update the
quantity to 2 or more, the price drops to $16.10 each.

OK... There's no telling how many Anchor Optics has in stock, so
it's first come, first serve, but before you rush out and get some,
here are the Pros and Cons:

PROS:

Farsighted users who can't live with the Infinity-limited (75mm)
focus travel of the new 3DWorld viewer will find that these lenses
hit Infinity at 66mm instead of 75mm and thus, they allow you to
extend the focus an additional 9mm beyond Infinity without any
further modification to the viewer! I'm now actually focusing these
66.0mm lenses in about the middle of the viewer's focus range. It's
really NICE to be able to go out of focus beyond Infinity, then bring
it back into focus precisely where I want it. Joy!

Magnification increases from 3.33x to 3.79x when focused at
Infinity. This doesn't sound like much, but you will notice the
difference immediately. It's remarkable - the view is much more
immersive. Joy!

The installation couldn't be easier - just unscrew the lens retainer
rings, pop out the 75mm lenses and pop in the 66mm lenses - but make
sure you put the flatter element toward your eye, with the more
rounded element toward the film (same as with the 75mm lenses). Joy!


CONS: (Joy not!)

The 66mm lenses are about 1.5mm thicker than the 75mm lenses, so the
retainer rings won't thread down as far, but they will go on with
enough turns to hold the lenses securely. They are also 0.5mm
smaller in diameter, so you should probably wrap their outside edges
with a couple of turns of mylar tape or even clear tape, just to make
sure they sit square to the lens barrel, but I just dropped mine in
and the retainer rings squared them up as I tightened them. (I'll
get around to it later...)

Pincushion distortion is only tolerable, but it's not that bad, in my
opinion. It will likely be an issue for some people and it will
certainly seem worse for subjects like architecture than for
landscapes. It's a small price to pay, in my opinion, for
supercharging the magnification -AND- making the adjustable focus
viable for people who aren't nearsighted.

Vignetting of the corners is also an issue, but I predict that for
those who are sufficiently farsighted to find themselves unable to
focus the original 75mm lenses at maximum travel, vignetting will not
be a problem with the 66mm lenses because they will be pulling the
lenses farther away from the film than I do to bring them into
focus. But for my eyes, which can just barely focus the 75mm lenses
at maximum travel, the 66mm lenses vignette the corners of the
80x140mm 3DWorld plastic mounts that have 52x52mm windows. This
requires that I move my eyes about to look into all of the corners -
but it's really not that disturbing when you consider how much larger
the virtual image is at the greater magnification. RMM-style
80x132mm mounts with 50x50mm windows are only slightly vignetted at
my focus preference. The 40x50mm, 50x40mm, and 52x23mm panoramic
cardboard mounts are not vignetted at all. In fact, my panoramic
mount (52x23mm) views have never looked better than in this 66mm
modified 3DWorld viewer! Note that I am using John Thurston's (DIY)
adapter for loading RMM-style cardboard mounts into the 3DWorld
viewer: http://stereo.thurstons.us/mountcarrier.htm

These lenses allow you to see dust and film grain at a greater
magnification, too, but it's nothing like what we see with smaller
stereo formats.

These so-called "Experimental Grade 1" surplus lenses are completely
free of scratches, chips, or other blemishes, but I did have to clean
them after unpacking them.

Do I recommend them? Let me put it this way - I've got no intention
of putting the 75mm lenses back in any time soon, but I am a little
bit annoyed by the vignetting seen with 52x52 and 50x50
mounts. Again, if you are farsighted and can't use the 75mm 3DWorld
STL viewer that's fixed-focused at Infinity, the vignetting won't be
as bad for you as it is for me, and I'm pretty sure you will like
these 66mm lenses even more than I do.

By the way, I thought about buying the entire Anchor Optics inventory
for these lenses then selling them on eBay at a markup, but I'm too
nice (or maybe I'm just too lazy)!

:-)

Mike Davis

Excellent save, John! I should leave the making of all such
suggestions to you.

:-)

Thanks,

Mike Davis

At 08:40 PM 9/23/2008, you wrote:

>Michael Davis wrote:
>
> > But here's something useful: If anyone wants to add a jack for an
> > external power supply, your measurements reveal that a transformer
> > with a rated output of 300 mA (0.3 Amp) at 4.5 VDC should work just fine.
>
>I'll get back to some of your other items later, but wanted to
>mention that if you are going to add a power jack, you should be
>sure to get one which disconnects internal power before
>connecting the external power (break before make). None of the
>suggested jacks seem to have this.
>
>If you use one of the jacks mentioned by Mike you should also use
>a single pole, double throw (SPDT) switch to enable/disable the
>external power. I can't recommend simply connecting an external
>power source in parallel with the internal battery.
>
>This is, of course, irrelevant if you remove the batteries before
>connecting external power :)
>_________________________
>John Thurston
>Juneau, Alaska

Thanks Mike,

I am ordering a pair too...

The price is good, except for the shipping charge ($10.87 for something
that can be mailed for $3 tops?)

Being able to switch lenses so easily, is a big plus of this viewer.

By the way, I have asked them to sell me the lenses (only). I have not
pressed the matter, but considering that the lenses are not too heavy, I'd
expect to be able to get a pair of the original lenses for only $10... You
think there is a market from MF enthusiasts who build their own viewers?
(I suspect that this group is getting smaller now that 3DWorld is offering
not one but two, make it 3 if you count the exhibition, viewers).

George

> I ordered a pair of 31.5mm x 66.0mm FL, coated, achromatic doublets
> (experimental Grade 1 surplus) from Anchor Optics for $16.10 each (a
> total of $45.76 for the pair with sales tax and shipping).
>
> Here's the page where they can be found (look for Stock # AX73771):
> http://www.anchoroptics.com/catalog/product.cfm?id=256

Michael Davis wrote:

> I think I prefer a brighter viewer than most people are content with,
> so I'm curious if you can offer any suggestions regarding how high a
> voltage that array of 18 LED's can handle safely.

- snip -

> It's probably not a linear relationship, but if it is, a 100% gain in
> ISO 100 EV (making the viewer twice as bright) would require a 617.4%
> gain in voltage, which equates to a 27.6V to achieve 12.6 EV, or
> 33.5V for 13.0 EV. Such voltages would surely fry something, so it
> seems there's not much point in trying to boost the brightness
> significantly by playing with higher voltage power supplies.

I can safely say that you shouldn't pump more than about 4.5VDC
into this thing. I have a regulated power supply I test with but
it is stepped rather than continuously variable.

At 3VDC input, the light output is only about 11 1/4 ISO 100 EV
At 4.5vdc input it is as previously described
At 6.0vdc it was the same as at 4.5vdc . . .
. . . until it abruptly fell to 0 EV :(

I suspect I've cooked the power regulator. I doubt I'll be able
to correct the fault but next weekend I'll check the pin-outs on
the chips. I may get lucky. If not, I'll be building a new
regulator for my illuminator.

fwiw, with a 3VDC input the light produced was decidedly cooler
than at 4.5vdc.
_________________________
John Thurston
Juneau, Alaska

Michael Davis wrote:

>
> I think I prefer a brighter viewer than most people are content with,
> so I'm curious if you can offer any suggestions regarding how high a
> voltage that array of 18 LED's can handle safely.
>
> I wouldn't want heat to become an issue, nor shorten their life span
> significantly, but going by your test results, it looks as if you
> boosted the measured ISO 100 EV from 11.6 to 12.0 (a 3.45% gain in
> EV) by increasing the supply voltage from 3.85 VDC (at 0.16 Amp) to
> 4.67 VDC (at 0.20 Amp) - a 21.3% increase in voltage.

It has been too long since I was an electrical engineering student,
but I think you are looking at the numbers incorrectly.

First, the Exposure Value (EV) scale is not linear. A difference of 1
EV is one stop, or a doubling of the light output. At ISO 100, you
can convert EV to luminance with the formula:

L = 2^(EV-3)

With that formula it appears that the luminance at 4.67VDC is
approximately 32% greater than at 3.85VDC.

Second, the voltage is not the important measure for the LEDs, the
power is. In John's test he increased not only the voltage, but also
the current, and thus the power (P=VI). At 4.67VDC and 0.20A the LEDs
are getting about 51% more power than at 3.85VDC and 0.16A.

So it appears (if I haven't completely screwed up the math) that a 50%
increase in power gave about a 30% increase in light output. If you
look at luminance/watt it looks like the LEDs are becoming less
efficient at the higher voltage.

--
Brian Reynolds | "It's just like flying a spaceship.
reynolds@panix.com | You push some buttons and see
http://www.panix.com/~reynolds/ | what happens." -- Zapp Brannigan
NAR# 54438 |

George,

At 09:45 PM 9/23/2008, you wrote:

>By the way, I have asked [3DWorld] to sell me the [75mm] lenses (only).
>I have not
>pressed the matter, but considering that the lenses are not too heavy, I'd
>expect to be able to get a pair of the original lenses for only $10... You
>think there is a market from MF enthusiasts who build their own viewers?
>(I suspect that this group is getting smaller now that 3DWorld is offering
>not one but two, make it 3 if you count the exhibition, viewers).

3DWorld's 32x75mm lenses are such great lenses, that at $10.00 or
even $20.00 a pair, there might be a couple of MF3D people interested
in getting them for homemade viewers, but I think you could actually
sell more of them to experimenters outside the MF3D community - as
few as that number may seem. In the end, I don't think you'd be able
to sell very many, either way, but I could be wrong.

I just searched eBay for the word "achromatic" and found only one
auction (Item number: 310084628160) for an unmounted lens - a $12.00
pair of one-inch (25mm) by seven-inch (175mm) achromatic
doublets. If that search is any measure of the size of the market
for unmounted achromats, pursuing 3DWorld to sell just their lenses
is probably not worth the effort.

Mike Davis

John,

At 11:09 PM 9/23/2008, you wrote:

>At 3VDC input, the light output is only about 11 1/4 ISO 100 EV
>At 4.5vdc input it is as previously described
>At 6.0vdc it was the same as at 4.5vdc . . .
>. . . until it abruptly fell to 0 EV :(
>
>I suspect I've cooked the power regulator.

Bummer. I wish I hadn't encouraged experimentation in this
area. Thanks for the warning, for everyone else's sake. External
power should be 4.5 Volts maximum.

Mike Davis

Hi Brian,

Thanks for the correction regarding EV. I'm embarrassed I got that wrong.

But I don't think John increased the current - he's just measuring
the current being pulled by the LED's after increasing the
voltage. John could elaborate on this.

But I think you're right regarding EV. And therefore, an external
power supply that gives us exactly 4.5 Volts (never less) could give
us a (noticeably?) brighter viewer than a set of three rechargeable
AA's running at something like 3.6 or 3.8 volts, when fully charged.

Mike Davis

Michael Davis wrote:

>
> Thanks for the correction regarding EV. I'm embarrassed I got that wrong.
>
> But I don't think John increased the current - he's just measuring
> the current being pulled by the LED's after increasing the
> voltage. John could elaborate on this.

That's poor wording on my part. The important point is that with the
increase voltage came an increase in current draw, and the power did
not just increase proportionally with the increase in voltage.

--
Brian Reynolds | "It's just like flying a spaceship.
reynolds@panix.com | You push some buttons and see
http://www.panix.com/~reynolds/ | what happens." -- Zapp Brannigan
NAR# 54438 |