Second light on 'dustbin' lens

[Ultrapurple]

The saga continues...

Long-time readers will remember that I acquired a large reflector lens some time ago. All I had to go on was the lens mount and the tantalising markings "LW" and "2.5°" - it sounds like a superb long-range LWIR lens.

Fast forward a year or so. With a lot of help from Fraser (disassembling my Thermovision 900 and returning to me the mount and lens, not to mention a great deal of off-list advice) I managed to get some reasonably clear and well focused images from it using a Therm-App Pro as the imaging device.



Here's the whole caboodle - the big, wastepaper basket size lens is the big thing on the right; mounted on it (with Blu-Tack and hope) is the lens 'won' from the Thermovision 900 an on the left is a Therm-App Pro and a phone. Here's a closer view:



So what did I see? Well, first of all, let's look at a quickly snatched view using the ThermApp Pro alone with its 35mm focal length lens:



My first view through the lens was the roofline opposite (marked A in the previous image):



So far so good - this was using the 35mm ThermApp lens, held some way behind the big lens and focused at near-infinity. You can see a screw head, indicating that the image from the big lens is in the same plane as the mount.

After some more fiddling, I settled on using the 13mm ThermApp lens (which is not optimal for the sensor), focused at well beyond infinity, mounted as seen in the second photo. This produced a much-enlarged image compared to earlier:



(By the way, the images through the lens are inverted - for ease of viewing I have rotated them here).

In summary, at the moment I'm getting a low contrast image with a central hotspot and some noticeable geometric distortion - but it IS an image and the magnification is significant. Work will continue, and I am optimistic that I'll get really good pictures in due course.

A quick bit of manipulation on images from a pan gives a panorama that, when I compare it to the image from the 35mm lens alone, indicates a focal length in the region of 150mm. I think the FL is probably significantly more than that, in fact, but that's approximately what I'm seeing at the moment.

[Fraser]

Great work 

One thing to be aware of is that these reflector lenses can have a very high F number so are anything but ‘fast’. F8 is common.

I will look up the details of that lens in my archive for you.

Fraser

[Ultrapurple]

Thanks Fraser

The primary mirror is about 250mm diameter and the FoV is 2.5° (I don't know whether that's H, V or diagonal, so let's assume H).

Assuming (probably wrongly) that it was designed for feeding the equivalent a 35µm 320x240 sensor (image size 11.4 x 8.4mm), a focal length of about 250mm is required for a 2.5° HFoV.

Knowing that f-number is equal to the focal length divided by the diameter of the entrance pupil,

f = 250mm (fl) / 250mm (pupil)

-> f = 1

That's a very wide aperture for a reflector lens! (They're usually around f/8, as Fraser said. Examination suggests that the optical configuration is basically a Cassegrain reflector).

Running other sets of numbers suggest:

50µm 320x240 (=16x12mm sensor) implies FL=370mm, f=1.5

Changing over to a line scan sensor with a (total guess) active line of 25mm implies FL=575mm, f=2.3

That's starting to sound slightly more plausible.

It will be interesting to see what Fraser comes up with!

[Fraser]

@Ultrapurple,

You are right. From what I am finding the thermal imaging reflector lenses are quite fast. AGEMA are shy about providing full technical details for that reflector lens. What I did find was that it is a 2.5 degree x 2.5 degree FOV (camera produced a 1:1 square aspect ratio). It is stated as having a minimum focus distance 20m.

The specs I have for your lens, that was designed for the THV880 and 870 series scanners, are as follows:

2.5 Degrees FOV / 20m minimum focus / 308mm Focal Distance / 0.43mrad Geometrical Resolution (slit response 50% contrast)

I found a datasheet for the earlier AGA telescope and I attach that for your information. That unit is 2 degrees x 2 degrees FOV and has larger Germanium optics. It is a larger lens but maybe useful information all the same.

Inframetrics produced similar telescope lenses so I will look for some data on them in slow time.

Fraser

[Ultrapurple]

Thanks Fraser

308mm FL is exceptionally useful information. From that we can see that

f = 308/250

f = 1.25 or so - a pretty fast design for a reflecting lens!

Actually, given that the entrance pupil is occluded by the secondary mirror the aperture will actually be somewhat less than 1.25. I guess perhaps 1.5 or 1.6 or so. The important thing is that it's in the right ball park for a modern uncooled microbolometer sensor to give good images once it's all set up nicely.

SCORE!

[Ultrapurple]

A further thought. If the lens does end up giving me a 2.5° HFoV then the moon will still only be 1/5 of the sensor wide - about 100 pixels.

But I expect it will be a bit different from that when I eventually get it set up with the correct supplementary lenses to get a clear image on a sensor.

[Fraser]

Ultrapurple,

Having thought about it, the lens does have a very large diameter 1st surface mirror so it would be a good energy collector compared to the smaller diameter mirrors I am used to seeing in visible light reflector lenses. I am not certain of the mirrors efficiency however.

It is a very nice lens 

Fraser

[bap2703]

The 2.5° angle is not a property of the lens itself : it takes into account the sensor size.
A smaller sensor would produce a smaller angle.
You might (or not) be more familiar with this issue in consumer cameras: full frame vs APS-C sensor --> the later produce narrower angle ~ almost as if a longer focal length lens was used.

[Vipitis]

Have you tried to use it as primary lens?

[Ultrapurple]

@bap2703 - yes, I am aware that the angle is only as stated when used with the sensor size it was designed for. I sort-of hinted in that direction in my early calculations of possible focal length and f number. But it's good to see it stated explicitly - thanks.

@Vipitis - It won't work as a pure primary lens, at least as it stands. You may remember my early experiments that found this out. The nice thing is that with the newly installed intermediary lens I am getting magnification comparable to what I saw the first time I got images from it.

This lens is a real learning experience for me. Unfortunately, as always, time is against me and I rarely have enough spare time to drag it out, set it up and experiment - at least, not when the weather is suitable. But that changes fairly quickly. The British have a well-deserved reputation for talking about the weather that completely flummoxes people from places with predictable weather: last time I was in Egypt I chatted with someone who simply couldn't get their head around the notion that Brits can't reliably plan a summer outdoor event because it's perfectly possible there will be rain, a thunderstorm and/or even hail. They were used to knowing that there simply won't be any clouds worth talking about for six months at a stretch.

[Hydron]

Did you find out whether the intermediate lens had a positive or negative optical polarity? Certainly wouldn't hurt if you could get it to work without the lens on Therm-App as well.

As for contrast etc, a naive calculation using .43mrad and 2.5deg FoV suggests a resolution of ~a hundred pixels across at the native sensor size, which sounds significantly lower than i'd expect out of a modern lens (though I might be doing this all wrong!).

[frogg]

Very cool stuff.

You could build yourself your own James Webb Telescope

I personally find stainless steel to be an excellent infrared reflector, even with surface finishes as crude as #4 brushed. Building a nice infrared reflecting telescope really must be the most economical method...

Such reflectors could be stamped/deep-drawn and polished quite economically I would think...