Fresnel lens tests

it works wonderful with flooder type kind of light… it gives more lux than without it…
i have a home made tube with the lens in the front
at about 12” it will give the flooder light average 3X-4X of the lux output on the light meter.
sadly it didn’t work very well with a thrower ( although it gives brighter rings and darker hotspot)
interestingly it makes my TK75 even bigger flooder with almost the same throw :slight_smile:
try it with the zoomie and it will cancel out the square spot and throw longer

I joined the mounted lens to a Raysoon TD 398, a 18650 and XM-L size Sipik 68 clone, without its bezel and aspheric lens.
A Sipik SK 73 with an XM-L with similar output is on the left. The Fresnel Raysoon is to the right and lower, both on low mode.

I could have preserved some zoom by attaching the mount only to the head, but as it is, it is a fixed thrower. There is not a great deal of spill or stray light. Most of the light appears to be in the hot spot. The light is usable in this form, though not rugged.

Some light is scattered in a pattern with rings, probably by the edges between the segments of the lens, and some hits the mounting cone.
The spot is ringed with yellow and red. Some of this may be the angle dependent color of the led, but most is from the chromatic aberration of the lens. Perhaps this is why lens throwers are not common. The separation of color is probably about the same for a Fresnel lens or a continuous lens, but it does to some extent depend on the material.
The 130 mm diameter and 50 mm focal length correspond to a photographic f number of 0.38 . I got it from here http://www.dhgate.com/product/flat-fresnel-lens-for-stage-light-with-size/153508269.html#ctabBox, but the price has already increased a bit.

I did some tests with a Fresnel lens. I found that the colour was more orange around the edges and more blue with the aspheric.

My understanding, which more or less agrees with my observations, is that the shorter wavelengths, cooler colors, are bent more by a lens, so when blue is in focus, red and yellow are less focused and appear on the edges. If the lens is a bit farther from the led, then the red is clearly focused and the blue is out of focus so one gets blue and green borders. My light might be whiter if I had it focused better, but a zoomy would still have yellow and red around the edges, except at full throw.
My aspherics have too short focal lengths to shed much light on the subject, so to speak.

I have been trying to figure out why it is so bright. Even if I arrange them to have equal spot sizes the Fresnel is clearly brighter. I tried swapping the batteries and saw the same thing. The focal length to diameter ratios are 50/130 = 0.38 for the Fresnel and 12/21 = 0.57 for the Sipik 73 (taking the flat face as the lens center and using the aperture in the bezel). This seems to account for it.

Close enough. It’s called “chromatic abberation”, but it’s also often referred to as red/purple fringing. :slight_smile:

I had polycarbonate eye glasses that had a lot of chromatic aberration. Perhaps this lens is polycarbonate. It didn’t say.
The way to greatly reduce it is to use total internal reflection (TIR) for the outer part of the lens, as Fresnel himself did for lighthouses. TIR like other reflection is non-dispersive, though the border of the TIR zone is color dependent.


One could also get the light collection and focal length even better by curving the lens to wrap it around the source, as with the lighthouses. Unfortunately these things would be very difficult for a hobbyist and we will have to weight for them to be found in a manufactured product.
However, the light collection is already excellent, and the colors are not noticeable when I light up distant trees.

They already are, I use them all the time: fresnel stage light - Google Search

Very interesting lens. Looks quite promising. You have any plans on what you would like to put it in yet?

I am just re-doing it with a less valuable lite body and a bit neater. I could construct a fiber glass or carbon fiber head but the new paper will withstand light use, I think. I don’t really have much use for a thrower.

What I mean is “lenses” that curve around the source and use total internal reflection in the outer segments. I don’t recognize any that are not flat in the images from that Google search. The one I have is intended for stage lighting.

I re-did this Fresnel light with a body that I am willing to commit to it and a somewhat neater paper based mounting. Paper is still the material out of which it is easiest to build a cone shape, but I mounted it to the body with epoxy and used several coats of model airplane dope over two sheets of ordinary paper. The result is usable, though a bit crooked looking.

Here is the new Fresnel flashlight beam spot.
I set the camera for auto exposure taken from the center of the field and daylight white balance, so size and and color can be compared but not brightness.

Next is a Sipik SK68, as the first comparison.

This is an old soft plastic 2D incandescent Eveready with well focused xenon aftermarket bulb, an older standard.

Finally a common XM-L zoomy, a Raysoon TD398, an SK98 clone or scaled up SK68 clone.

And a picture of the four lights.
Of the four, the Fresnel light probably puts the largest part of the source light into the spot, because of the larger ratio of diameter to focal length. The result is a light and inexpensive XM-L thrower, but bulky and not very rugged physically.
The body is that of what I call a Sipik 73 clone, for want of a better name, that I acquired inexpensively, so the whole thing cost less than the Raysoon. I took the whole head off this time. The cooling is by conduction through the body and by air circulating over the bare pill, star and dome, but that should be sufficient for this lightly driven XM-L.

Love your creativity and out of the box thinking. Tying that lens holder into a flashlight body seems like quite a challenge. How close is the emitter to the back of the lens?

EDIT: I just went to the product page. I was thinking the lens was cone shaped and nearly 50mm deep. So I take it the emitter is about 50mm behind the lens. A little short of 2 inches. So it should be more efficient than that 100mm aspheric lens from FT. Do you have any lux and current measurements?

Thanks!
I don’t have any way of making quantitative measurements. I could compare it to another similar light with the aspheric still in place if that would help. There is a bit of scattered light, as shown in the original pictures, but I don’t think much is lost from the spot.
There is chromatic aberration, as there would be with a regular lens but not with a silvered or TIR reflector. It doesn’t show much in this auto-exposure picture, but there is a blue border. When I was locating the source I could adjust the compromise between the blue border and orange at the corners by moving the body closer or farther from the lens. It slipped a bit again on setting and I got too much blue. The orange being at the corners must be due to cross terms of the chromatic aberration interacting with other aberrations, probably the focal plane curvature and the large emitter size.
Added: I realize now that part of the reason the border is blue is that the surface I was looking at it on while adjusting the focus was not far away. Blue is more strongly bent than red by most materials, so it focuses closer up.

If I cover the center of the lens with my hand, the spot becomes smaller.
This is because the outer rings are segments of longer focal length lenses than the inner rings are. This is another reason, besides the worse diffraction limitation, that Fresnel lenses are not often used for cameras and telescopes. I had not realized this until recently. It becomes noticeable, even for flashlights, lighthouses and stage lighting, with very large diameter to focal length ratio, but it is not really a bad quality for these applications. The focal length of the center rings is the same as that of a smooth lens that would fit in the same length.


This is a distant tree, over the balcony and the swimming pool. Nearer tree below round topped farther tree.

My daughter’s boy friend suggested that with a soft plastic [such as vinyl] such a light could be folded and put into a pocket.

focal lenght to diameter ratio is one thing, and important to capture as much light as possible.
there is however another reason why the 50mm focal length lens has a brighter spot. The reason is the longer focal length.
have a look at the ‘thin lens’ lens formula:
1/F=1/s+1/s’

s is the distance from the LED die to the lens, while s’ is the distance between the lens and the image of the LED on the wall (assuming you have the image clearly in focus as seems to be your case).

If you know the distance between the lens and the wall (s’), and the focal length F, you can calculate s as follows:
s=1/(1/F-1/s’) (use excel :wink: )

with s and s’, you can calculate the magnification:
M=-s’/s

Now, for the shorter focal length, you’ll get a bigger magnification that for the shorter focal length, because the lens is closer to the LED, but the distance to the wall stays the same.
If you know the LED die size and multiply it with the magnification, you get the size of the image on the wall.

Since the lens really forms an image of the LED die, reducing the size will get you a brighter image.
if you would find 2 lenses with different focal lengths, but with the same diameter to focal length ratio, the lens with the longest focal length will result in the brightests image since it captures the same amount of light, but projects it in a smaller area.

I hope this is not too technical, and helps you understand more about the things you’re seeing.
Here you can find some more fresnel lenses if anyone is interested: short focal length Fresnel lenses for sale

Best regards,
Johan

Perhaps you didn’t understand my thoughts. If I hold the two lights at different distances from the wall, so that the spots come out about the same size, then I can see the total amount of focused light in the relative brightnesses. The effect of the focal length is compensated by the different distances from the wall, so all I see is the part of the led’s output that is focused by each lens. With similar aspheric lenses, the two lights look similar, so if one is brighter in this test that means that more of the light is captured. That appears to be accounted for the the difference in the ratios of diameter to focal length, which determines (through trigonometry) the solid angle subtended. The led’s pattern must be wide enough (XM-L with dome intact) that substantial light hits that extra outer part of the lens. It also shows that, in this application, a Fresnel lens has comparable efficiency to that of the aspheric, that is the steps between the segments don’t cause much light to be lost.
I fully understand the first order optics, but in this case I have compensated for the difference in angular spread by adjusting the distances, rather than by including it in a calculation.